Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 11:28 am
Chocolatebar wrote: Mon Feb 20, 2023 11:16 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am If you don't take equity leverage in the taxable account, you're bucketing, and the cost of bucketing far outweighs the cost of the tax drag. You're forcing yourself into an incredibly inefficent AA. Just look at the difference between these two accounts which is very similar to what you're doing. They both start with 1.4x leverage. But the second one rebalances between the buckets, treating it as one bucket with a low level of leverage, rather than 2 separate buckets (1 with no leverage - taxable, and 1 with high leverage - IRA).
I disagree with this point. As long as I consider the AA of my entire portfolio (which I have), there shouldn't be a problem. Is simply splitting funds into separate buckets where they are held better bad? Why? They are different types of accounts (buckets) which are designed to hold funds differently.

Also - the taxable does have leverage. On the bond side, which is more tax efficient. I also think it's important to stress that the IRA will be ~10% of my entire portfolio this year...
They are separate because you have no ability to maintain your AA if the market goes down. Comeinvest and I discussed this above and agreed that accounts that cannot be leveraged must be treated separately. If you were planning to leverage the taxable if the market went down in order to maintain your AA, then you could treat them as one bucket.
Sure you do. You just have to shift allocations in each bucket to meet the overall target AA. It's only a minor problem in this situation because my Roth is so small, but it should grow quickly as I prioritize contributions to it over my taxable throughout the years.
skierincolorado wrote: Mon Feb 20, 2023 11:28 am
Chocolatebar wrote: Mon Feb 20, 2023 11:16 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am It's really simple if you think about it. If you have a significant tax drag problem, it's because stock and bond markets did very well - in which case UPRO+TYD will destroy NTSX and you will be glad to have this "problem". If markets are more choppy and go up slowly, you won't have any tax drag, and you'll still beat NTSX.
Why can't this situation just happen in my IRA then? I won't have any tax drag problems to worry about there.
Because the cost of adopting very high leverage in one account and no leverage in the other account far outweighs the cost of any tax drag which would only occur if markets did very well. You're making your best case scenario better but making your worst case scenario worse. Usually the goal is to do the opposite. Your worst case scenario is worse because if markets do poorly you will experience extreme volatility decay and you will have received no benefit for doing so - in fact you will have missed out on additional TLH that you could use to offset future gains.
I don't understand this perspective. My overall AA is exactly the same, so it will behave the same regardless of how many accounts it's spread across. It'll actually perform better during a good year because I won't have any tax consequences.

120/60 is 120/60. If most of the stocks are in tax-advantaged and most of the bonds are in taxable - that's a good thing!
skierincolorado wrote: Mon Feb 20, 2023 11:28 am
Chocolatebar wrote: Mon Feb 20, 2023 11:16 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am As comeinvest and I discussed above, this is simply impossible to do. Several people in this thread including myself ran into this exact problem this year. When markets dropped we were unable to maintain our target AA for the whole portfolio because significant sub-accounts could not be leveraged. I was lucky in that I could leverage my 401k even though initially I planned not to (due to an aversion to LETFs). Others were not so lucky, and were not able to maintain their AA, and were not able to fully partake in the market rebound.
I mean, 6.5k/year isn't pocket change. Most years I'd be DCAing into it at the end of every month with my extra money. In the event a big crash happened - I could be more aggressive with the remaining limit for the year and throughout the following year until recovery.
skierincolorado wrote: Mon Feb 20, 2023 10:51 am I agree more leverage is probably worth it but there is a limit past which the volatility decay becomes too extreme. The lifecycle investing authors recommended a cap of 2x. I think a bit higher than that is ok but maybe not all the way to 3x. Since the accounts aren't fungible because you aren't planning to leverage the taxable account, the accounts should be treated as separate buckets. I wouldn't leverage my whole net worth 3x, so I probably wouldn't leverage single account 3x either if is non-fungible with my other accounts. 3x is way beyond the Kelly criterion.

Since 1900, 3x leverage has the same total return as 1x leverage, but with some fun 97% drawdowns mixed in.
I'm looking at my entire portfolio across accounts. Total leverage would barely exceed 1.1x.
Yes but they are separate buckets.
So what? Why should I care about how many accounts my portfolio is spread across? As long as each "bucket" contains what it was best designed for - there shouldn't be any problems.
skierincolorado wrote: Mon Feb 20, 2023 11:28 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am I wouldn't DCA. If the market goes up this year you miss out. Lifecycle investing principles tell us to invest as much as possible in the present by leveraging, and DCA would be the opposite. Worst case is the market tanks but rebounds before your 2024 contrib. I'd just tone it back with AVUV or TYD. Something to reduce the leverage a bit and hopefully with a correlation <1.
I do generally agree with lump summing > DCA. However, in this scenario DCA seems like a no-brainer from a risk management perspective. It practically guarantees my Roth can't be completely wiped out.
Some form of DCA might make sense but the bigger risk is that markets do well before you accumulate wealth. DCAing makes this risk worse. I'd lump some most of it and maybe hold a little in reserve. Which effectively just means the account has lower leverage. The money is there, it's just not physically in the account. Which is functionally the same as what I proposed which was leveraging it 2-2.5x instead of 3x.
DCA only "makes this risk worse" if you assume one outcome (stocks go up). Any other outcome makes DCA the better play. The worst outcome for lump summing is FAR worse than the worst outcome for DCA.

Arguing this point doesn't really matter anyway since most of the 12.5k this year needs to be DCA in some way. I can't just throw 12k in tomorrow. I think 6k is the most I'd be able to lump sum at once.
Last edited by Chocolatebar on Mon Feb 20, 2023 12:37 pm, edited 1 time in total.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 11:37 am
Chocolatebar wrote: Mon Feb 20, 2023 11:24 am
skierincolorado wrote: Mon Feb 20, 2023 11:18 am I mean honestly analyzing tax drag on HFEA from 2010-2020 is kind of comical if you think about it. A .73% drag is the difference between having 1.2M and 1.25M. Which is still infinitely more than 500k (NTSX). If stock market returns were low enough that HFEA returned less than NTSX, there would be no tax drag.
Just because HFEA returned more than NTSX would've over that decade doesn't mean it was better from a risk perspective. HFEA isn't worth implementing at all ever imo because it's too risky.

EDIT - I assume you mean mHFEA actually when you refer to HFEA? If so, I bet NTSX still would've had better risk-adjusted returns when you factor in taxes (especially if you looked at a longer timeline).
Just saw you edit. That's not true. From 2010-present 10k NTSX (90/60) would have final value of 47k. A slightly more leveraged but less tax efficient AA of 130/60 has a final value of 71k.

The CAGR is 16.17% vs 12.57%.

The tax drag on HFEA from 2010-2020 was .73% if making contribs. The tax drag on this much more dialed back version would be less. The difference in CAGR dwarfs the tax drag.

Now what if markets don't do as well? Well then you simply don't have any tax drag to worry about. Even if markets go up modestly, you'll be harvesting the shares with losses. The tax drag only occurs if markets go straight up and there are no shares with losses. If markets go up, you'll be glad to have the leverage. If markets do so-so you'll still be glad to have the leverage, and there won't be tax drag. If markets do poorly, you'll get TLH.

Those of us in this thread for over a year have all experienced this personally. When markets crashed this year, I wasn't worrying about tax drag. I was glad that I was able to maintain my AA and not forced into deleveraging at market bottoms.


Choosing to bucket because of fear of tax drag will make your worst case outcome much worse. It might make your best case outcome better if markets only go up and you experience no volatility decay and avoid tax drag. I'm not worried about my best case. I'm worried about my worst case. In my worst case, tax drag will be zero but volatility decay will be real.
What was the max drawdown of each portfolio? Just curious.

This doesn't really matter since I'm not proposing a 90/60 portfolio. My portfolio will be 120/60.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Mon Feb 20, 2023 12:24 pm
skierincolorado wrote: Mon Feb 20, 2023 11:28 am
Chocolatebar wrote: Mon Feb 20, 2023 11:16 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am If you don't take equity leverage in the taxable account, you're bucketing, and the cost of bucketing far outweighs the cost of the tax drag. You're forcing yourself into an incredibly inefficent AA. Just look at the difference between these two accounts which is very similar to what you're doing. They both start with 1.4x leverage. But the second one rebalances between the buckets, treating it as one bucket with a low level of leverage, rather than 2 separate buckets (1 with no leverage - taxable, and 1 with high leverage - IRA).
I disagree with this point. As long as I consider the AA of my entire portfolio (which I have), there shouldn't be a problem. Is simply splitting funds into separate buckets where they are held better bad? Why? They are different types of accounts (buckets) which are designed to hold funds differently.

Also - the taxable does have leverage. On the bond side, which is more tax efficient. I also think it's important to stress that the IRA will be ~10% of my entire portfolio this year...
They are separate because you have no ability to maintain your AA if the market goes down. Comeinvest and I discussed this above and agreed that accounts that cannot be leveraged must be treated separately. If you were planning to leverage the taxable if the market went down in order to maintain your AA, then you could treat them as one bucket.
Sure you do. You just have to shift allocations in each bucket to meet the overall target AA. It's only a minor problem in this situation because my Roth is so small, but it should grow quickly as I prioritize contributions to it over my taxable throughout the years.
OK I was under the impression that you would not further leveage the taxable in order to maintain the 120/60 target AA. If that is your plan then I think it's great. Although as mentioned in many ways it is more work than holding a box spread or futures which may require no / less rebalancing.

The rebalancing must be frequent when using UPRO because your AA will drift quickly. If the market goes down your AA will immediately drift from target. If it goes back up before rebalancing you've locked in those losses and volatility decay. In my testing, with a 3x fund like UPRO rebalancing monthly is sufficient, but using narrow rebalancing bands is better. You can imagine with a 5x fund (if one existed) the volatility decay on a daily basis would be insurmountable without at least weekly rebalancing.
Chocolatebar wrote: Mon Feb 20, 2023 12:24 pm
skierincolorado wrote: Mon Feb 20, 2023 11:28 am
Chocolatebar wrote: Mon Feb 20, 2023 11:16 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am It's really simple if you think about it. If you have a significant tax drag problem, it's because stock and bond markets did very well - in which case UPRO+TYD will destroy NTSX and you will be glad to have this "problem". If markets are more choppy and go up slowly, you won't have any tax drag, and you'll still beat NTSX.
Why can't this situation just happen in my IRA then? I won't have any tax drag problems to worry about there.
Because the cost of adopting very high leverage in one account and no leverage in the other account far outweighs the cost of any tax drag which would only occur if markets did very well. You're making your best case scenario better but making your worst case scenario worse. Usually the goal is to do the opposite. Your worst case scenario is worse because if markets do poorly you will experience extreme volatility decay and you will have received no benefit for doing so - in fact you will have missed out on additional TLH that you could use to offset future gains.
I don't understand this perspective. My overall AA is exactly the same, so it will behave the same regardless of how many accounts it's spread across. It'll actually perform better during a good year because I won't have any tax consequences.

120/60 is 120/60. If most of the stocks are in tax-advantaged and most of the bonds are in taxable - that's a good thing!
The AA will drift on a daily basis and faster than you might think. 3x leverage is a lot and UPRO experiences huge swings. Like I said, I wasn't aware you were planning to add UPRO (or something else) to your taxable in order to maintain the target AA. If so, that should work as long as you rebalance back to target on a monthly basis.
Chocolatebar wrote: Mon Feb 20, 2023 12:24 pm
skierincolorado wrote: Mon Feb 20, 2023 11:28 am
Chocolatebar wrote: Mon Feb 20, 2023 11:16 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am As comeinvest and I discussed above, this is simply impossible to do. Several people in this thread including myself ran into this exact problem this year. When markets dropped we were unable to maintain our target AA for the whole portfolio because significant sub-accounts could not be leveraged. I was lucky in that I could leverage my 401k even though initially I planned not to (due to an aversion to LETFs). Others were not so lucky, and were not able to maintain their AA, and were not able to fully partake in the market rebound.
I mean, 6.5k/year isn't pocket change. Most years I'd be DCAing into it at the end of every month with my extra money. In the event a big crash happened - I could be more aggressive with the remaining limit for the year and throughout the following year until recovery.
skierincolorado wrote: Mon Feb 20, 2023 10:51 am I agree more leverage is probably worth it but there is a limit past which the volatility decay becomes too extreme. The lifecycle investing authors recommended a cap of 2x. I think a bit higher than that is ok but maybe not all the way to 3x. Since the accounts aren't fungible because you aren't planning to leverage the taxable account, the accounts should be treated as separate buckets. I wouldn't leverage my whole net worth 3x, so I probably wouldn't leverage single account 3x either if is non-fungible with my other accounts. 3x is way beyond the Kelly criterion.

Since 1900, 3x leverage has the same total return as 1x leverage, but with some fun 97% drawdowns mixed in.
I'm looking at my entire portfolio across accounts. Total leverage would barely exceed 1.1x.
Yes but they are separate buckets.
So what? Why should I care about how many accounts my portfolio is spread across? As long as each "bucket" contains what it was best designed for - there shouldn't be any problems.
I feel like this is becoming a terminoligy issue. By buckets I mean buckets that you do not rebalance across. They have separate AAs that do not affect each other. This is what I thought you were planning. If you plan to increase leverage in taxable if the IRA tanks, then there is no issue. But it should be done regularly - like monthly.

Chocolatebar wrote: Mon Feb 20, 2023 12:24 pm
skierincolorado wrote: Mon Feb 20, 2023 11:28 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am I wouldn't DCA. If the market goes up this year you miss out. Lifecycle investing principles tell us to invest as much as possible in the present by leveraging, and DCA would be the opposite. Worst case is the market tanks but rebounds before your 2024 contrib. I'd just tone it back with AVUV or TYD. Something to reduce the leverage a bit and hopefully with a correlation <1.
I do generally agree with lump summing > DCA. However, in this scenario DCA seems like a no-brainer from a risk management perspective. It practically guarantees my Roth can't be completely wiped out.
Some form of DCA might make sense but the bigger risk is that markets do well before you accumulate wealth. DCAing makes this risk worse. I'd lump some most of it and maybe hold a little in reserve. Which effectively just means the account has lower leverage. The money is there, it's just not physically in the account. Which is functionally the same as what I proposed which was leveraging it 2-2.5x instead of 3x.
DCA only "makes this risk worse" if you assume one outcome (stocks go up). Any other outcome makes DCA the better play. The worst outcome for lump summing is FAR worse than the worst outcome for DCA.

Arguing this point doesn't really matter anyway since most of the 12.5k this year needs to be DCA in some way. I can't just throw 12k in tomorrow. I think 6k is the most I'd be able to lump sum at once.
Yes it's the same problem with regular DCA without leverage. But your contributions are effectively DCA already. The optimum balance is probably lump sum half or 2/3 or whatever you are able and DCA the rest. But if you actually have a plan to maintain your overall AA, it's a moot point.


The issues with LETFs are complex and subtle. I get that the tax drag during a massive bull market is obvious and plain to see. The volatility decay that occurs by not maintaining your target AA is subtle and hard to see but far more costly. It makes your worst case outcomes even worse. Even maintaining a fixed AA you will experience volatility decay and be ignoring lifecycle investing principles. With bucketing you make the problem even worse. More volatilty decay and more lifecycle risk because you are selling stocks every time the market drops.

Personally, I am barely comfortable with the use of LETFs at all and I have a detailed plan to maintain my target AA across accounts and fight the volatility decay. To fight the volatility decay effectively (and follow lifecycle investing principles) I don't rebalance to a static AA - the AA is dynamic and increases leverage as my wealth decreases. Despite this detailed plan to fight the volatility decay, I still experience volatility decay within UPRO between rebalancings.


There's three general ways to run leverage

1. Lifecycle style where leverage increases as wealth decreases. This experiences the least volatility decay and reduces the sequence of return risk from contributions.

2. Static where a target AA is maintained at all times. Some volatility decay and lots of sequence of return risk. Hopefully the market does well after you've made most of your lifetime contributions.

3. Separate buckets with separate AAs where overall leverage will decrease as the market drops. Massive volatility decay and massive sequence of return risk.


I thought what you were proposing is #3. If it's actually #2 and you have a detailed plan to maintain your target AA on a monthly basis, then some of my previous replies/posts are off base.
Last edited by skierincolorado on Mon Feb 20, 2023 1:50 pm, edited 1 time in total.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 1:30 pm
Chocolatebar wrote: Mon Feb 20, 2023 12:24 pm
skierincolorado wrote: Mon Feb 20, 2023 11:28 am
Chocolatebar wrote: Mon Feb 20, 2023 11:16 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am If you don't take equity leverage in the taxable account, you're bucketing, and the cost of bucketing far outweighs the cost of the tax drag. You're forcing yourself into an incredibly inefficent AA. Just look at the difference between these two accounts which is very similar to what you're doing. They both start with 1.4x leverage. But the second one rebalances between the buckets, treating it as one bucket with a low level of leverage, rather than 2 separate buckets (1 with no leverage - taxable, and 1 with high leverage - IRA).
I disagree with this point. As long as I consider the AA of my entire portfolio (which I have), there shouldn't be a problem. Is simply splitting funds into separate buckets where they are held better bad? Why? They are different types of accounts (buckets) which are designed to hold funds differently.

Also - the taxable does have leverage. On the bond side, which is more tax efficient. I also think it's important to stress that the IRA will be ~10% of my entire portfolio this year...
They are separate because you have no ability to maintain your AA if the market goes down. Comeinvest and I discussed this above and agreed that accounts that cannot be leveraged must be treated separately. If you were planning to leverage the taxable if the market went down in order to maintain your AA, then you could treat them as one bucket.
Sure you do. You just have to shift allocations in each bucket to meet the overall target AA. It's only a minor problem in this situation because my Roth is so small, but it should grow quickly as I prioritize contributions to it over my taxable throughout the years.
OK I was under the impression that you would not further leveage the taxable in order to maintain the 120/60 target AA. If that is your plan then I think it's great. Although as mentioned in many ways it is more work than holding a box spread or futures which may require no / less rebalancing.

The rebalancing must be frequent when using UPRO because your AA will drift quickly. If the market goes down your AA will immediately drift from target. If it goes back up before rebalancing you've locked in those losses and volatility decay. In my testing, with a 3x fund like UPRO rebalancing monthly is sufficient, but using narrow rebalancing bands is better. You can imagine with a 5x fund (if one existed) the volatility decay on a daily basis would be insurmountable without at least weekly rebalancing.
No, your assumption is correct. I refuse to have stock leverage in my taxable when it's more effectively held in a tax-advantaged account.
skierincolorado wrote: Mon Feb 20, 2023 1:30 pm
Chocolatebar wrote: Mon Feb 20, 2023 12:24 pm
skierincolorado wrote: Mon Feb 20, 2023 11:28 am
Chocolatebar wrote: Mon Feb 20, 2023 11:16 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am It's really simple if you think about it. If you have a significant tax drag problem, it's because stock and bond markets did very well - in which case UPRO+TYD will destroy NTSX and you will be glad to have this "problem". If markets are more choppy and go up slowly, you won't have any tax drag, and you'll still beat NTSX.
Why can't this situation just happen in my IRA then? I won't have any tax drag problems to worry about there.
Because the cost of adopting very high leverage in one account and no leverage in the other account far outweighs the cost of any tax drag which would only occur if markets did very well. You're making your best case scenario better but making your worst case scenario worse. Usually the goal is to do the opposite. Your worst case scenario is worse because if markets do poorly you will experience extreme volatility decay and you will have received no benefit for doing so - in fact you will have missed out on additional TLH that you could use to offset future gains.
I don't understand this perspective. My overall AA is exactly the same, so it will behave the same regardless of how many accounts it's spread across. It'll actually perform better during a good year because I won't have any tax consequences.

120/60 is 120/60. If most of the stocks are in tax-advantaged and most of the bonds are in taxable - that's a good thing!
The AA will drift on a daily basis and faster than you might think. 3x leverage is a lot and UPRO experiences huge swings. Like I said, I wasn't aware you were planning to add UPRO (or something else) to your taxable in order to maintain the target AA. If so, that should work as long as you rebalance back to target on a monthly basis.
It would be 10% of my total portfolio. Less than that throughout most of this year...
skierincolorado wrote: Mon Feb 20, 2023 1:30 pm
Chocolatebar wrote: Mon Feb 20, 2023 12:24 pm
skierincolorado wrote: Mon Feb 20, 2023 11:28 am
Chocolatebar wrote: Mon Feb 20, 2023 11:16 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am As comeinvest and I discussed above, this is simply impossible to do. Several people in this thread including myself ran into this exact problem this year. When markets dropped we were unable to maintain our target AA for the whole portfolio because significant sub-accounts could not be leveraged. I was lucky in that I could leverage my 401k even though initially I planned not to (due to an aversion to LETFs). Others were not so lucky, and were not able to maintain their AA, and were not able to fully partake in the market rebound.
I mean, 6.5k/year isn't pocket change. Most years I'd be DCAing into it at the end of every month with my extra money. In the event a big crash happened - I could be more aggressive with the remaining limit for the year and throughout the following year until recovery.
skierincolorado wrote: Mon Feb 20, 2023 10:51 am I agree more leverage is probably worth it but there is a limit past which the volatility decay becomes too extreme. The lifecycle investing authors recommended a cap of 2x. I think a bit higher than that is ok but maybe not all the way to 3x. Since the accounts aren't fungible because you aren't planning to leverage the taxable account, the accounts should be treated as separate buckets. I wouldn't leverage my whole net worth 3x, so I probably wouldn't leverage single account 3x either if is non-fungible with my other accounts. 3x is way beyond the Kelly criterion.

Since 1900, 3x leverage has the same total return as 1x leverage, but with some fun 97% drawdowns mixed in.
I'm looking at my entire portfolio across accounts. Total leverage would barely exceed 1.1x.
Yes but they are separate buckets.
So what? Why should I care about how many accounts my portfolio is spread across? As long as each "bucket" contains what it was best designed for - there shouldn't be any problems.
I feel like this is becoming a terminoligy issue. By buckets I mean buckets that you do not rebalance across. They have separate AAs that do not affect each other. This is what I thought you were planning. If you plan to increase leverage in taxable if the IRA tanks, then there is no issue. But it should be done regularly - like monthly.
So if the entire portfolio I'm proposing was in my taxable - that would be fine? But, because I want to have the least tax-friendly 10% of it in a tax-advantaged account - then bad?
skierincolorado wrote: Mon Feb 20, 2023 1:30 pm
Chocolatebar wrote: Mon Feb 20, 2023 12:24 pm
skierincolorado wrote: Mon Feb 20, 2023 11:28 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am
skierincolorado wrote: Mon Feb 20, 2023 10:51 am I wouldn't DCA. If the market goes up this year you miss out. Lifecycle investing principles tell us to invest as much as possible in the present by leveraging, and DCA would be the opposite. Worst case is the market tanks but rebounds before your 2024 contrib. I'd just tone it back with AVUV or TYD. Something to reduce the leverage a bit and hopefully with a correlation <1.
I do generally agree with lump summing > DCA. However, in this scenario DCA seems like a no-brainer from a risk management perspective. It practically guarantees my Roth can't be completely wiped out.
Some form of DCA might make sense but the bigger risk is that markets do well before you accumulate wealth. DCAing makes this risk worse. I'd lump some most of it and maybe hold a little in reserve. Which effectively just means the account has lower leverage. The money is there, it's just not physically in the account. Which is functionally the same as what I proposed which was leveraging it 2-2.5x instead of 3x.
DCA only "makes this risk worse" if you assume one outcome (stocks go up). Any other outcome makes DCA the better play. The worst outcome for lump summing is FAR worse than the worst outcome for DCA.

Arguing this point doesn't really matter anyway since most of the 12.5k this year needs to be DCA in some way. I can't just throw 12k in tomorrow. I think 6k is the most I'd be able to lump sum at once.
Yes it's the same problem with regular DCA without leverage. But your contributions are effectively DCA already. The optimum balance is probably lump sum half or 2/3 or whatever you are able and DCA the rest. But if you actually have a plan to maintain your overall AA, it's a moot point.

The issues with LETFs are complex and subtle. I get that the tax drag during a massive bull market is obvious and plain to see. The volatility decay that occurs by not maintaining your target AA is subtle and hard to see but far more costly. It makes your worst case outcomes even worse. Even maintaining a fixed AA you will experience volatility decay and be ignoring lifecycle investing principles. With bucketing you make the problem even worse. More volatilty decay and more lifecycle risk because you are selling stocks every time the market drops.
I just want to be clear - is the main reason you don't like the portfolio because the roth is so small (~10%)? If each account had the same amount of money in it, would you still say I shouldn't only leverage stocks in one and bonds in the other, even though it would be just as easy to rebalance as if they were all in one account?
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Mon Feb 20, 2023 1:49 pm No, your assumption is correct. I refuse to have stock leverage in my taxable when it's more effectively held in a tax-advantaged account.



It would be 10% of my total portfolio. Less than that throughout most of this year...


So if the entire portfolio I'm proposing was in my taxable - that would be fine? But, because I want to have the least tax-friendly 10% of it in a tax-advantaged account - then bad?


I just want to be clear - is the main reason you don't like the portfolio because the roth is so small (~10%)? If each account had the same amount of money in it, would you still say I shouldn't only leverage stocks in one and bonds in the other, even though it would be just as easy to rebalance as if they were all in one account?
You're basically saying that since it's only 10%, it doesn't matter that it's incredibly inefficient and experiences extreme volatility decay and does the opposite of what lifecycle investing tells us to do. Well the upside is also proportionally small. The AA you proposed is 111/54. Because you are bucketing with two independent AAs, your overall portfolio AA will decrease if the market drops. 3 months from now it would not surprise me if your overall portfolio AA had drifted to 95/57 because UPRO tanks on you.

You could run these two buckets forever and probably the ending total value will be worse than just going 100% SPY in both accounts. And experienced far more risk along the way. If you don't see how, you don't understand volatility decay. 3x leverage has the same total return as 1x leverage since 1900. It's waaaaaaaaaaaay past the Kelly Criterion. Yes 10% is a small amount and maybe you don't intend to run them like this forever, but that's not a justification for running such a bad portfolio design in the interim. The potential benefit is also quite small and you've opened yourself up to the risk that 5 years from now the value in your IRA is still near zero because of extreme volatility decay from owning UPRO.

And you're opening yourself up to all of this extreme volatility decay and sequence risk to protext against a problem that only exists if markets do phenomenally well (tax drag).

I mean part of me agrees it's only 10% and so it dosn't really matter. The benefits of all this complexity and fees and minor amounts of leverage will be quite small and so are the risks and costs. But I'm trying to demonstrate the principles because the principles will apply equally whether you have a lot of leverage or a little. Most of us have a lot more leverage than 111/54. But the same principles apply. The people who bucketed last year got hammered and missed on the rebound. The people who didn't bucket participated fully in the rebound and got to do some TLH at the bottom. Following the principles of lifecycle investing and avoiding volatility decay made a bad market environment tolerable. If the market does phenomanlly well the next 10 years (which I don't expect that it will) yeah I'll have a little bit of tax drag. But if the markets do mediocre or poorly I won't have any tax drag and my performance will be vastly superior to bucketing because I will have much less volatility decay and sequence of return risk.

The only scenario where bucketing into tax-efficient buckets wins, is if markets go nearly straight up. That's not an eventuality you should be concerned about hedging against. It's like saying I'm not going to grow profitable apple trees on my farm because one day I might be rich and want to put a swimming pool there. Your odds of getting rich in the first place will be increased by planting the apple trees (ie avoiding volatility decay).


Ultimately I feel like you are not grasping the hidden cost of volatility decay. It is far greater than any *potential* tax drag. This principle applies equally whether you have a lot of leverage or a little leverage.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 2:11 pm
Chocolatebar wrote: Mon Feb 20, 2023 1:49 pm No, your assumption is correct. I refuse to have stock leverage in my taxable when it's more effectively held in a tax-advantaged account.

It would be 10% of my total portfolio. Less than that throughout most of this year...


So if the entire portfolio I'm proposing was in my taxable - that would be fine? But, because I want to have the least tax-friendly 10% of it in a tax-advantaged account - then bad?


I just want to be clear - is the main reason you don't like the portfolio because the roth is so small (~10%)? If each account had the same amount of money in it, would you still say I shouldn't only leverage stocks in one and bonds in the other, even though it would be just as easy to rebalance as if they were all in one account?
You're basically saying that since it's only 10%, it doesn't matter that it's incredibly inefficient and experiences extreme volatility decay and does the opposite of what lifecycle investing tells us to do. Well the upside is also proportionally small. The AA you proposed is 111/54. Because you are bucketing with two independent AAs, your overall portfolio AA will decrease if the market drops. 3 months from now it would not surprise me if your overall portfolio AA had drifted to 95/57 because UPRO tanks on you.
It isn't inefficient. It's extremely tax-efficient. If it was all in one account it would be fine, but because it's split between two in a tax optimal way, that's somehow bad. If UPRO tanks in 3 months then it'll be a great day for DCA...
skierincolorado wrote: Mon Feb 20, 2023 2:11 pm You could run these two buckets forever and probably the ending total value will be worse than just going 100% SPY in both accounts. And experienced far more risk along the way. If you don't see how, you don't understand volatility decay. 3x leverage has the same total return as 1x leverage since 1900. It's waaaaaaaaaaaay past the Kelly Criterion. Yes 10% is a small amount and maybe you don't intend to run them like this forever, but that's not a justification for running such a bad portfolio design in the interim. The potential benefit is also quite small and you've opened yourself up to the risk that 5 years from now the value in your IRA is still near zero because of extreme volatility decay from owning UPRO.

And you're opening yourself up to all of this extreme volatility decay and sequence risk to protext against a problem that only exists if markets do phenomenally well (tax drag).

I mean part of me agrees it's only 10% and so it dosn't really matter. The benefits of all this complexity and fees and minor amounts of leverage will be quite small and so are the risks and costs. But I'm trying to demonstrate the principles because the principles will apply equally whether you have a lot of leverage or a little. Most of us have a lot more leverage than 111/54. But the same principles apply. The people who bucketed last year got hammered and missed on the rebound. The people who didn't bucket participated fully in the rebound and got to do some TLH at the bottom. Following the principles of lifecycle investing and avoiding volatility decay made a bad market environment tolerable. If the market does phenomanlly well the next 10 years (which I don't expect that it will) yeah I'll have a little bit of tax drag. But if the markets do mediocre or poorly I won't have any tax drag and my performance will be vastly superior to bucketing because I will have much less volatility decay and sequence of return risk.
12.5k should be PLENTY to maintain my target AA this year.

Which would you rather see me do - 12.5k SPY in my Roth, or 12.5k DCA throughout the year UPRO...
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Mon Feb 20, 2023 2:22 pm

It isn't inefficient. It's extremely tax-efficient. If it was all in one account it would be fine, but because it's split between two in a tax optimal way, that's somehow bad. If UPRO tanks in 3 months then it'll be a great day for DCA...



12.5k should be PLENTY to maintain my target AA this year.

Which would you rather see me do - 12.5k SPY in my Roth, or 12.5k DCA throughout the year UPRO...
Probably the 12.5k SPY. UPRO is incredibly inefficient due to volatility decay. It's way past the Kelley Criterion. I generally consider it junk to own unless you have some plan to rebalance back into it if the market tanks. The DCA helps, but at the cost of having less initial investment. 40% SPY 60% UPRO could work especially if you shift more to UPRO if the market drops. Unless shifting towards UPRO in a market drop is somehow harder than DCA?

It's a false choice of course. You could simply hold a more efficient AA in taxable that won't experience the absurd volatility decay. If the market does extremely well there would be some minimal tax drag. Since the leverage level is so low, the drag would probably be under 0.2% even in an extreme bull market. It was a mere .73% in one of the strongest markets in history 2010-2020 on a fully leveraged regular HFEA. I'm also guessing you would opt for higher leverage than 111/54 if you weren't constraining yourself only to the IRA.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 2:11 pm The only scenario where bucketing into tax-efficient buckets wins, is if markets go nearly straight up. That's not an eventuality you should be concerned about hedging against. It's like saying I'm not going to grow profitable apple trees on my farm because one day I might be rich and want to put a swimming pool there. Your odds of getting rich in the first place will be increased by planting the apple trees (ie avoiding volatility decay). [/b]
But I really have two fields - one where I can grow apple trees tax free and the other where I can't.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 2:30 pm
Chocolatebar wrote: Mon Feb 20, 2023 2:22 pm

It isn't inefficient. It's extremely tax-efficient. If it was all in one account it would be fine, but because it's split between two in a tax optimal way, that's somehow bad. If UPRO tanks in 3 months then it'll be a great day for DCA...



12.5k should be PLENTY to maintain my target AA this year.

Which would you rather see me do - 12.5k SPY in my Roth, or 12.5k DCA throughout the year UPRO...
Probably the 12.5k SPY. UPRO is incredibly inefficient due to volatility decay. It's way past the Kelley Criterion. I generally consider it junk to own unless you have some plan to rebalance back into it if the market tanks. The DCA helps, but at the cost of having less initial investment. 40% SPY 60% UPRO could work especially if you shift more to UPRO if the market drops. Unless shifting towards UPRO in a market drop is somehow harder than DCA?

It's a false choice of course. You could simply hold a more efficient AA in taxable that won't experience the absurd volatility decay. If the market does extremely well there would be some minimal tax drag. Since the leverage level is so low, the drag would probably be under 0.2% even in an extreme bull market. It was a mere .73% in one of the strongest markets in history 2010-2020 on a fully leveraged regular HFEA. I'm also guessing you would opt for higher leverage than 111/54 if you weren't constraining yourself only to the IRA.
But If I included the 10% UPRO in my taxable instead that would be fine, right?
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Mon Feb 20, 2023 2:57 pm
skierincolorado wrote: Mon Feb 20, 2023 2:30 pm
Chocolatebar wrote: Mon Feb 20, 2023 2:22 pm

It isn't inefficient. It's extremely tax-efficient. If it was all in one account it would be fine, but because it's split between two in a tax optimal way, that's somehow bad. If UPRO tanks in 3 months then it'll be a great day for DCA...



12.5k should be PLENTY to maintain my target AA this year.

Which would you rather see me do - 12.5k SPY in my Roth, or 12.5k DCA throughout the year UPRO...
Probably the 12.5k SPY. UPRO is incredibly inefficient due to volatility decay. It's way past the Kelley Criterion. I generally consider it junk to own unless you have some plan to rebalance back into it if the market tanks. The DCA helps, but at the cost of having less initial investment. 40% SPY 60% UPRO could work especially if you shift more to UPRO if the market drops. Unless shifting towards UPRO in a market drop is somehow harder than DCA?

It's a false choice of course. You could simply hold a more efficient AA in taxable that won't experience the absurd volatility decay. If the market does extremely well there would be some minimal tax drag. Since the leverage level is so low, the drag would probably be under 0.2% even in an extreme bull market. It was a mere .73% in one of the strongest markets in history 2010-2020 on a fully leveraged regular HFEA. I'm also guessing you would opt for higher leverage than 111/54 if you weren't constraining yourself only to the IRA.
But If I included the 10% UPRO in my taxable instead that would be fine, right?
Yes if you had a plan to rebalance back to 10% when the market drops (using m1 auto rebalancing) or preferably higher than 10% per lifecycle investing.
Last edited by skierincolorado on Mon Feb 20, 2023 3:23 pm, edited 1 time in total.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Mon Feb 20, 2023 2:56 pm
skierincolorado wrote: Mon Feb 20, 2023 2:11 pm The only scenario where bucketing into tax-efficient buckets wins, is if markets go nearly straight up. That's not an eventuality you should be concerned about hedging against. It's like saying I'm not going to grow profitable apple trees on my farm because one day I might be rich and want to put a swimming pool there. Your odds of getting rich in the first place will be increased by planting the apple trees (ie avoiding volatility decay). [/b]
But I really have two fields - one where I can grow apple trees tax free and the other where I can't.
No. In the analogy not planting apples = volatility decay. You can change the analogy but it doesn't correspond to the choice at hand. You can either choose volatility decay or no volatility decay. If you choose volatility decay, you can avoid *potential* tax drag, but at what cost?

Generally speaking it just doesn't seem like you understand how bad the volatility decay of UPRO is.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 3:19 pm
Chocolatebar wrote: Mon Feb 20, 2023 2:56 pm
skierincolorado wrote: Mon Feb 20, 2023 2:11 pm The only scenario where bucketing into tax-efficient buckets wins, is if markets go nearly straight up. That's not an eventuality you should be concerned about hedging against. It's like saying I'm not going to grow profitable apple trees on my farm because one day I might be rich and want to put a swimming pool there. Your odds of getting rich in the first place will be increased by planting the apple trees (ie avoiding volatility decay). [/b]
But I really have two fields - one where I can grow apple trees tax free and the other where I can't.
No. In the analogy not planting apples = volatility decay. You can change the analogy but it doesn't correspond to the choice at hand. You can either choose volatility decay or no volatility decay. If you choose volatility decay, you can avoid *potential* tax drag, but at what cost?

Generally speaking it just doesn't seem like you understand how bad the volatility decay of UPRO is.
I understand how bad it is on the way down, but on the way up it works in your favor. It's a double edged sword. Luckily, the market is expected to move up over the long-term.

https://www.optimizedportfolio.com/how- ... he-market/
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 3:15 pm
Chocolatebar wrote: Mon Feb 20, 2023 2:57 pm
skierincolorado wrote: Mon Feb 20, 2023 2:30 pm
Chocolatebar wrote: Mon Feb 20, 2023 2:22 pm

It isn't inefficient. It's extremely tax-efficient. If it was all in one account it would be fine, but because it's split between two in a tax optimal way, that's somehow bad. If UPRO tanks in 3 months then it'll be a great day for DCA...



12.5k should be PLENTY to maintain my target AA this year.

Which would you rather see me do - 12.5k SPY in my Roth, or 12.5k DCA throughout the year UPRO...
Probably the 12.5k SPY. UPRO is incredibly inefficient due to volatility decay. It's way past the Kelley Criterion. I generally consider it junk to own unless you have some plan to rebalance back into it if the market tanks. The DCA helps, but at the cost of having less initial investment. 40% SPY 60% UPRO could work especially if you shift more to UPRO if the market drops. Unless shifting towards UPRO in a market drop is somehow harder than DCA?

It's a false choice of course. You could simply hold a more efficient AA in taxable that won't experience the absurd volatility decay. If the market does extremely well there would be some minimal tax drag. Since the leverage level is so low, the drag would probably be under 0.2% even in an extreme bull market. It was a mere .73% in one of the strongest markets in history 2010-2020 on a fully leveraged regular HFEA. I'm also guessing you would opt for higher leverage than 111/54 if you weren't constraining yourself only to the IRA.
But If I included the 10% UPRO in my taxable instead that would be fine, right?
Yes if you had a plan to rebalance back to 10% when the market drops (using m1 auto rebalancing) or preferably higher than 10% per lifecycle investing.
Kind of wish this was its own thread since it's a pretty good debate about how much it's worth prioritizing tax efficiency when using leverage. I'm sure others here wish it was too (sorry everyone). I really do get your point, but I strongly believe leveraged stocks should never be held in a taxable account when they can just as easily be held in tax-advantaged ones. I understand they can't "just as easily" today in my case because the problem is it may not be as easy to rebalance in the event of a major market crash (the account is small).

For the sake of compromise, let's say I want to keep my taxable just as I've pitched - 90/60, but in the event of a historic market crash, I'll at least be open to adding equity leverage in hopes we're near the bottom of it. In the meantime, I want my Roth to be as aggressive as possible because I want my equity leverage to be there long-term. What's the most aggressive strategy (90-100% US stocks I assume (maybe some bonds?)) that you would approve I allocate 12.5k this year to my Roth (6k lump - 6.5k dca second half of year)?

I think you would agree this is reasonable?
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Mon Feb 20, 2023 3:25 pm
skierincolorado wrote: Mon Feb 20, 2023 3:19 pm
Chocolatebar wrote: Mon Feb 20, 2023 2:56 pm
skierincolorado wrote: Mon Feb 20, 2023 2:11 pm The only scenario where bucketing into tax-efficient buckets wins, is if markets go nearly straight up. That's not an eventuality you should be concerned about hedging against. It's like saying I'm not going to grow profitable apple trees on my farm because one day I might be rich and want to put a swimming pool there. Your odds of getting rich in the first place will be increased by planting the apple trees (ie avoiding volatility decay). [/b]
But I really have two fields - one where I can grow apple trees tax free and the other where I can't.
No. In the analogy not planting apples = volatility decay. You can change the analogy but it doesn't correspond to the choice at hand. You can either choose volatility decay or no volatility decay. If you choose volatility decay, you can avoid *potential* tax drag, but at what cost?

Generally speaking it just doesn't seem like you understand how bad the volatility decay of UPRO is.
I understand how bad it is on the way down, but on the way up it works in your favor. It's a double edged sword. Luckily, the market is expected to move up over the long-term.

https://www.optimizedportfolio.com/how- ... he-market/
In really exceptionally great markets, sure. Most of the time the Kelly Criterion is way below 3x. I'm reposting the backtests that illustrate this.

Both portfolios start with 1.4x. One of them does what I refer to as "bucketing" - no rebalancing - it's really two separate buckets one with a 100% SPY allocaiton and another with a 100% UPRO allocation. The SPY bucket is 80% of your total wealth, and the UPRO bucket is 20% of wealth. But the other portfolio treats it as one bucket and rebalances such that the 1.4x is maintained (at least on a monthly basis).

This former is exactly what you are proposing (but 20% UPRO instead of 10%). The former literally exactly simulates what you are proposing vs the latter is exactly what I am suggesting.

The one I am suggesting returns 7% more after just 3 years. It probably aslo experienced near zero net tax drag, especially if contributions were being made. That comes out to over 2% per year. The volatility drag = 2%. The tax drag is near zero. You're paying a 2% volatility drag in your implementation to avoid a *potential* 0.73% tax drag that would only occur 1) if you were highly leveraged (you're not) and 2) if the market is consistently very strong and all of your share lots have gains (unlikely). You're paying a 2% volatility drag that occurs in normal and poor markets, and makes your "worst case" scenario even worse, to avoid a tax drag only occurs in the "best case" scenario and is still smaller than the volatility drag.

I really don't see how somebody could rationally make the choice to make their worst case CAGR 2.1% worse just so they could make their best case CAGR .4% better. Perhaps my commentary has been too vague. Hopefully quantifying the "volatility drag" and giving it a name so that we can directly compare it to the tax drag makes the choice clearer. You are choosing to experience 2% volatility drag in most normal markets to avoid an 0.4% tax drag in the best markets.

The volatility decay of UPRO is a massive drag on the portfolio. 2% per year in this example. Unless you fight the volatility decay by rebalancing into it when it goes down.


https://www.portfoliovisualizer.com/bac ... tion2_1=80

https://www.portfoliovisualizer.com/bac ... tion2_1=80
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 4:23 pm Both portfolios start with 1.4x. One of them does what I refer to as "bucketing" - no rebalancing - it's really two separate buckets one with a 100% SPY allocaiton and another with a 100% UPRO allocation. The SPY bucket is 80% of your total wealth, and the UPRO bucket is 20% of wealth. But the other portfolio treats it as one bucket and rebalances such that the 1.4x is maintained (at least on a monthly basis).

This former is exactly what you are proposing (but 20% UPRO instead of 10%). The former literally exactly simulates what you are proposing vs the latter is exactly what I am suggesting.

The one I am suggesting returns 7% more after just 3 years. It probably aslo experienced near zero net tax drag, especially if contributions were being made. That comes out to over 2% per year. The volatility drag = 2%. The tax drag is near zero. You're paying a 2% volatility drag in your implementation to avoid a *potential* 0.73% tax drag that would only occur 1) if you were highly leveraged (you're not) and 2) if the market is consistently very strong and all of your share lots have gains (unlikely). You're paying a 2% volatility drag that occurs in normal and poor markets, and makes your "worst case" scenario even worse, to avoid a tax drag only occurs in the "best case" scenario and is still smaller than the volatility drag.

I really don't see how somebody could rationally make the choice to make their worst case CAGR 2.1% worse just so they could make their best case CAGR .4% better. Perhaps my commentary has been too vague. Hopefully quantifying the "volatility drag" and giving it a name so that we can directly compare it to the tax drag makes the choice clearer. You are choosing to experience 2% volatility drag in most normal markets to avoid an 0.4% tax drag in the best markets.

The volatility decay of UPRO is a massive drag on the portfolio. 2% per year in this example. Unless you fight the volatility decay by rebalancing into it when it goes down.


https://www.portfoliovisualizer.com/bac ... tion2_1=80

https://www.portfoliovisualizer.com/bac ... tion2_1=80
I don't think it's fair to say my proposed implementation doesn't have any rebalancing though. Obviously the Roth would if I lump summed 12.5k UPRO into it and it immediately crashed, but that was never going to be the case. I DO get the point you're making though, so thank you for taking the time to show that.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Mon Feb 20, 2023 3:43 pm
skierincolorado wrote: Mon Feb 20, 2023 3:15 pm
Chocolatebar wrote: Mon Feb 20, 2023 2:57 pm
skierincolorado wrote: Mon Feb 20, 2023 2:30 pm
Chocolatebar wrote: Mon Feb 20, 2023 2:22 pm

It isn't inefficient. It's extremely tax-efficient. If it was all in one account it would be fine, but because it's split between two in a tax optimal way, that's somehow bad. If UPRO tanks in 3 months then it'll be a great day for DCA...



12.5k should be PLENTY to maintain my target AA this year.

Which would you rather see me do - 12.5k SPY in my Roth, or 12.5k DCA throughout the year UPRO...
Probably the 12.5k SPY. UPRO is incredibly inefficient due to volatility decay. It's way past the Kelley Criterion. I generally consider it junk to own unless you have some plan to rebalance back into it if the market tanks. The DCA helps, but at the cost of having less initial investment. 40% SPY 60% UPRO could work especially if you shift more to UPRO if the market drops. Unless shifting towards UPRO in a market drop is somehow harder than DCA?

It's a false choice of course. You could simply hold a more efficient AA in taxable that won't experience the absurd volatility decay. If the market does extremely well there would be some minimal tax drag. Since the leverage level is so low, the drag would probably be under 0.2% even in an extreme bull market. It was a mere .73% in one of the strongest markets in history 2010-2020 on a fully leveraged regular HFEA. I'm also guessing you would opt for higher leverage than 111/54 if you weren't constraining yourself only to the IRA.
But If I included the 10% UPRO in my taxable instead that would be fine, right?
Yes if you had a plan to rebalance back to 10% when the market drops (using m1 auto rebalancing) or preferably higher than 10% per lifecycle investing.
Kind of wish this was its own thread since it's a pretty good debate about how much it's worth prioritizing tax efficiency when using leverage. I'm sure others here wish it was too (sorry everyone). I really do get your point, but I strongly believe leveraged stocks should never be held in a taxable account when they can just as easily be held in tax-advantaged ones. I understand they can't "just as easily" today in my case because the problem is it may not be as easy to rebalance in the event of a major market crash (the account is small).

For the sake of compromise, let's say I want to keep my taxable just as I've pitched - 90/60, but in the event of a historic market crash, I'll at least be open to adding equity leverage in hopes we're near the bottom of it. In the meantime, I want my Roth to be as aggressive as possible because I want my equity leverage to be there long-term. What's the most aggressive strategy (90-100% US stocks I assume (maybe some bonds?)) that you would approve I allocate 12.5k this year to my Roth (6k lump - 6.5k dca second half of year)?

I think you would agree this is reasonable?
It's a solid improvement. I'd want to do it in more than a historic market crash, any sort of medium sized downturn. It just doesn't make any sense to be selling shares of stock (or share equivalents) when the market goes down. It locks in losses that are unrecoverable and those losses show up as a signficant volatility drag (-2% CAGR) in the backtests.

Comeinvest and I agreed if the unleveragable account is small relative to the leverageable account AND total leverage is low enough that you can achieve your overall target AA even in a severe crash using only the leverageable account, THEN it can make sense to calculate and target an AA for the whole portfolio, but implement the leverage only in the leverageable account. If either of those conditions are not met, then it doesn't work.

Targeting an AA for your whole portfolio but leveraging only in one account also requires a dynamic rebalancing strategy that increases/decreases leverage in the IRA as needed to maintain the total portfolio AA. It's not easy.

I would be good with only holding the leverage in the IRA if the IRA was at least double in size to taxable AND the total portfolio AA was less than 1.2x equity leverage. For example, 100k in taxable and 200k in IRA. If the target is 1.2x leverage then the IRA would be 1.3x leveraged. 1.3x leverage in the IRA is low enough that we would be able to lever it up to 2x or more if needed in a major crash to maintain the overall target AA of 1.2x. In 1929 and 2008 this would have been necessary but we would (just barely) have been able to maintain the target AA.

But you are no where close to have twice as much money in your IRA as taxable. Thus you must choose between *possible* tax drag of up to .73% for highly leveraged portfolios in very strong bull markets, OR volatility drag of 2% in all but the best of markets. I choose the latter. So far I've experienced no tax drag (actually significant tax boost in the form of TLH to buffer my losses). And I've avoided a ton of volatility decay (9% or so).
Last edited by skierincolorado on Mon Feb 20, 2023 4:58 pm, edited 3 times in total.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 4:46 pm But you are no where close to have twice as much money in your IRA as taxable.
And ironically I never will be without equity leverage. If only I could leverage my 401k...sigh...
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Mon Feb 20, 2023 4:51 pm
skierincolorado wrote: Mon Feb 20, 2023 4:46 pm But you are no where close to have twice as much money in your IRA as taxable.
And ironically I never will be without equity leverage.
Leveraging past the Kelley criterion won't help you get there, it will make the disparity worse.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 4:46 pm Thus you must choose between *possible* tax drag of up to .73% for highly leveraged portfolios in very strong bull markets, OR volatility drag of 2% in all but the best of markets.
But only one of these strategies could lead to my long-term desired allocation...

The volatility drag could be a short-term problem/price. The tax drag is a long-term/"forever" problem...
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Mon Feb 20, 2023 4:56 pm
skierincolorado wrote: Mon Feb 20, 2023 4:46 pm Thus you must choose between *possible* tax drag of up to .73% for highly leveraged portfolios in very strong bull markets, OR volatility drag of 2% in all but the best of markets.
But only one of these strategies could lead to my long-term desired allocation...

The volatility drag could be a short-term problem/price. The tax drag is a long-term/"forever" problem...
Putting the cart before the horse. You have to get rich first to have a tax drag problem. Experiencing massive (-2.1%) volatility decay isn't going to help you get rich. These LETFs are not to be trifled with. The 2010-present performance is deceptive. Eventually you would be putting all new contributions towards the unleveaged funds in your portfolio. This would help bring the leverage back to a lower AA as you age. You're right eventually you migh have to sell significant gains at 15% cap gains tax rates. But probably not for a couple decades and by that point avoiding the 2.1%/year volatility drag will have coumpounded many times over and pay for those taxes many times over.

Personally I estimate I have avoided nearly 30k dollars of volatility decay in just 2 years and harvested 30k of losses. I expect I'll continue avoiding losses from volaitlity decay. I hope I see the day that I see some tax drag. I see virtually no chance that I ever face a tax cost that supersedes the combined savings from eliminated volatility decay and TLH.
Last edited by skierincolorado on Mon Feb 20, 2023 10:59 pm, edited 3 times in total.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 4:46 pm Targeting an AA for your whole portfolio but leveraging only in one account also requires a dynamic rebalancing strategy that increases/decreases leverage in the IRA as needed to maintain the total portfolio AA. It's not easy.
Again - technically I would be leveraging both accounts. Unless you don't count bond leverage.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Mon Feb 20, 2023 5:07 pm
skierincolorado wrote: Mon Feb 20, 2023 4:46 pm Targeting an AA for your whole portfolio but leveraging only in one account also requires a dynamic rebalancing strategy that increases/decreases leverage in the IRA as needed to maintain the total portfolio AA. It's not easy.
Again - technically I would be leveraging both accounts. Unless you don't count bond leverage.
I'm just looking at the equity leverage to keep things simple, you have no equity leverage in the taxable
parval
Posts: 153
Joined: Tue Oct 22, 2019 9:23 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by parval »

Folks w/ ZN/ZF (or other 1256 contracts), are you all just manually entering every transaction in your tax forms?

Just noticed Turbotax imports rest of IB docs but won't auto populate these.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

parval wrote: Mon Feb 20, 2023 10:02 pm Folks w/ ZN/ZF (or other 1256 contracts), are you all just manually entering every transaction in your tax forms?

Just noticed Turbotax imports rest of IB docs but won't auto populate these.
I just entered the summary info
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Mon Feb 20, 2023 4:46 pm I would be good with only holding the leverage in the IRA if the IRA was at least double in size to taxable AND the total portfolio AA was less than 1.2x equity leverage. For example, 100k in taxable and 200k in IRA. If the target is 1.2x leverage then the IRA would be 1.3x leveraged. 1.3x leverage in the IRA is low enough that we would be able to lever it up to 2x or more if needed in a major crash to maintain the overall target AA of 1.2x. In 1929 and 2008 this would have been necessary but we would (just barely) have been able to maintain the target AA.

But you are no where close to have twice as much money in your IRA as taxable. Thus you must choose between *possible* tax drag of up to .73% for highly leveraged portfolios in very strong bull markets, OR volatility drag of 2% in all but the best of markets. I choose the latter. So far I've experienced no tax drag (actually significant tax boost in the form of TLH to buffer my losses). And I've avoided a ton of volatility decay (9% or so).
After reassessing everything I've decided you're right. The only thing I disagree with above is that the Roth needs to be twice as large as the taxable. Wouldn't it be fine if they had roughly the same amount? Assuming you were aiming for leverage <1.5x. I think the above assumes you wouldn't have any equity in taxable.

EDIT - I think I see you mean ONLY if there was a severe market crash like 1929 or 2008. Honestly, that's so uncommon that I don't think it's reasonable to require tax-leveraged 2x taxable. Not saying you shouldn't mitigate/have a plan for it (seems like even then the strategy would still hold up as long as you shift leverage to taxable), but holding yourself back every year from a tax efficiency perspective in fear of an imminent market crash that you could still handle seems like a tough pill to swallow. I realized that if the market does crash though that it makes perfect sense to leverage equity in taxable since doing so will harvest losses.

Anyways, I've put together a new portfolio that I'm sure you'll agree is an improvement:

roth: (12k) (~10%)
TYA - 33 (3.3)
UPRO - 67 (6.7)

taxable: (109k) (~90%)
UPRO - 15 (13.5)
AVUV - 25 (22.5)
VEA - 10 (9)
AVDV - 10 (9)
VWO - 10 (9)
AVES - 10 (9)
TYA - 20 (18)

AA: 87/36/54 US/INT/BND

It really doesn't make any sense to avoid UPRO in my taxable at this time since I currently have 20k of losses harvested (17k after taxes). The only thing I'm unsure of is whether the international diversification (VEA, AVDV, VWO, AVES) is worth a higher ER than VXUS would run. I also need to figure out what to use for bonds until TYA reaches 50mil AUM. I think that's what M1 waits to see before they add a new fund to the platform. I'm not against using TYD or EDV for a little bit until that happens. I believe you prefer the former?
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Wed Feb 22, 2023 5:26 am
skierincolorado wrote: Mon Feb 20, 2023 4:46 pm I would be good with only holding the leverage in the IRA if the IRA was at least double in size to taxable AND the total portfolio AA was less than 1.2x equity leverage. For example, 100k in taxable and 200k in IRA. If the target is 1.2x leverage then the IRA would be 1.3x leveraged. 1.3x leverage in the IRA is low enough that we would be able to lever it up to 2x or more if needed in a major crash to maintain the overall target AA of 1.2x. In 1929 and 2008 this would have been necessary but we would (just barely) have been able to maintain the target AA.

But you are no where close to have twice as much money in your IRA as taxable. Thus you must choose between *possible* tax drag of up to .73% for highly leveraged portfolios in very strong bull markets, OR volatility drag of 2% in all but the best of markets. I choose the latter. So far I've experienced no tax drag (actually significant tax boost in the form of TLH to buffer my losses). And I've avoided a ton of volatility decay (9% or so).
After reassessing everything I've decided you're right. The only thing I disagree with above is that the Roth needs to be twice as large as the taxable. Wouldn't it be fine if they had roughly the same amount? Assuming you were aiming for leverage <1.5x. I think the above assumes you wouldn't have any equity in taxable.

EDIT - I think I see you mean ONLY if there was a severe market crash like 1929 or 2008. Honestly, that's so uncommon that I don't think it's reasonable to require tax-leveraged 2x taxable. Not saying you shouldn't mitigate/have a plan for it (seems like even then the strategy would still hold up as long as you shift leverage to taxable), but holding yourself back every year from a tax efficiency perspective in fear of an imminent market crash that you could still handle seems like a tough pill to swallow. I realized that if the market does crash though that it makes perfect sense to leverage equity in taxable since doing so will harvest losses.

Anyways, I've put together a new portfolio that I'm sure you'll agree is an improvement:

roth: (12k) (~10%)
TYA - 33 (3.3)
UPRO - 67 (6.7)

taxable: (109k) (~90%)
UPRO - 15 (13.5)
AVUV - 25 (22.5)
VEA - 10 (9)
AVDV - 10 (9)
VWO - 10 (9)
AVES - 10 (9)
TYA - 20 (18)

AA: 87/36/54 US/INT/BND

It really doesn't make any sense to avoid UPRO in my taxable at this time since I currently have 20k of losses harvested (17k after taxes). The only thing I'm unsure of is whether the international diversification (VEA, AVDV, VWO, AVES) is worth a higher ER than VXUS would run. I also need to figure out what to use for bonds until TYA reaches 50mil AUM. I think that's what M1 waits to see before they add a new fund to the platform. I'm not against using TYD or EDV for a little bit until that happens. I believe you prefer the former?
Yeah I think your edit is correct. Twice as much in the IRA with 1.2x leverage is only necessary out of possibility of 2008 or 1929. Somebody that had the option to move to taxable in such a scenario could be a bit more leveraged. But equal amounts with 1.5x leverage wouldn't work. With 100k in both and 1.5x leverage, one would start with 2x in the IRA. They would almost certainly be forced to leverage the taxable at some point. It also might be making the tax efficiency worse at that point, because in a choppy market you would be realizing gains in taxable and losses in the IRA (the opposite of the goal). Making our "worst case" outcome worse. If one had leveraged the accounts equally from the start, one would be harvesting losses in choppy markets that went sideways or slowly upwards - ameliorating our worst case outcome. For someone who had the *option* to leverage the IRA (in a 2008 scenario) but chose not to I think equal account sizes with 1.2x leverage would be OK. For someone who has no option at all, I think the IRA has to be twice as large.

You do see what I mean about tax drag being near zero (or getting TLH) unless the market is very strong right? When you are contributing you have lots of different entry points. M1 will be selling the lots with the highest basis. In a choppy market that goes sideways this means you'll be getting TLH. If it goes slowly upwards, you'll be near zero tax drag. Only if it goes straight upwards (like 2010-2020) do you get the 0.73% tax drag in that simulation - and that simulation was making smaller contributions than you will be.

You are right - I like the new portfolio (assuming you intend to maintain the AA on a monthly basis (87/36/54). One thing to consider is that because the IRA is still more leveraged, maintaining your overall target AA will be a PAIN and require maintaining a spreadsheet. Right now I calculate the AA in taxable as 109/60 stocks/bonds. A simple and small modification that would greatly simplify your life but possibly incur a small tax drag if the market does very well, would be to increase the AA in taxable from 109/60 to 123/54 (87/36/54). That's not much different from what it is now, but would simplify your life greatly in terms of maintaining the AA - you will have to do nothing.

You are probably right that your accounts are similar enough you should be able to maintain your target AA. And I see why you're doing it, you're giving a bit more leverage to the IRA for tax efficiency purposes but not so much that it's going to be killed by volatility decay. It's smart and I agree with it. But you're causing yourself a lot more work in order to maintain the overall AA and you may find that the work isn't worth it. It's going to require a spreadsheet and manually rebalancing across the accounts. Having the same AA in both would be less work but comes down to personal preference. I'm comfortable that the AAs are close enough in the two accounts that you won't kill the IRA with volatility decay and you can increase the leverage in the taxable as necessary to ensure that overall you are not being hurt by volatility decay.

One question would be, if the market goes down, which account would you increase leverage in first in order to maintain your overall AA of 87/36/54?

Although I think it's overall good that you're worrying less about tax drag which is small and only occurs if the market does well, and more about volatility drag which is big and occurs all the time, I actually think one of your stated reasons might be off. The 20k in harvested losses - we want to hold onto them for as long as possible. Eventually even the most tax efficient strategy will eliminate those - so we will be paying taxes on our investments one day. Whether that's 3 years from now or 30 years from now - we don't know. I suppose taxes paid 30 years from now have less present value - so your reasoning might be partially correct. With your low level of leverage, your starting losses, and if you did some TLH in the future, you might be able to avoid investment taxes for a very long time.

Even with the fees, historically TYD is more efficient than EDV. The one thing you'll have to check out is how big the bid/ask spreads are on TYA if it is ever allowed. I think for buy and hold it's probably OK but worth looking into. Occasional rebalancing should be fine too but really frequent rebalancing with a wide bid ask would be problematic.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Wed Feb 22, 2023 11:20 am You do see what I mean about tax drag being near zero (or getting TLH) unless the market is very strong right? When you are contributing you have lots of different entry points. M1 will be selling the lots with the highest basis. In a choppy market that goes sideways this means you'll be getting TLH. If it goes slowly upwards, you'll be near zero tax drag. Only if it goes straight upwards (like 2010-2020) do you get the 0.73% tax drag in that simulation - and that simulation was making smaller contributions than you will be.
I do. I'm not afraid to leverage my taxable now, so thanks for teaching me that. I believe as long as you're prioritizing leverage in tax-advantage first in a way that's easy to manage - it's perfectly fine to add more/as little as needed leverage to taxable to maintain the target AA or go slightly past it in the event of a market downturn. You'll likely be harvesting losses at that point, so why should someone be concerned about capital gains?

I do know implementing/managing this approach is harder though if there is a great difference between account values, but I still think it's worth leveraging tax-advantaged more than taxable generally.
skierincolorado wrote: Wed Feb 22, 2023 11:20 am You are right - I like the new portfolio (assuming you intend to maintain the AA on a monthly basis (87/36/54). One thing to consider is that because the IRA is still more leveraged, maintaining your overall target AA will be a PAIN and require maintaining a spreadsheet. Right now I calculate the AA in taxable as 109/60 stocks/bonds. A simple and small modification that would greatly simplify your life but possibly incur a small tax drag if the market does very well, would be to increase the AA in taxable from 109/60 to 123/54 (87/36/54). That's not much different from what it is now, but would simplify your life greatly in terms of maintaining the AA - you will have to do nothing.
Yeah I'll always opt for simplicity (especially with a newborn). Did you have specific allocations in mind for that?
skierincolorado wrote: Wed Feb 22, 2023 11:20 am One question would be, if the market goes down, which account would you increase leverage in first in order to maintain your overall AA of 87/36/54?
Always tax-advantaged first, but NEVER past 2x. After that - we shift attention to taxable.
skierincolorado wrote: Wed Feb 22, 2023 11:20 am Although I think it's overall good that you're worrying less about tax drag which is small and only occurs if the market does well, and more about volatility drag which is big and occurs all the time, I actually think one of your stated reasons might be off. The 20k in harvested losses - we want to hold onto them for as long as possible. Eventually even the most tax efficient strategy will eliminate those - so we will be paying taxes on our investments one day. Whether that's 3 years from now or 30 years from now - we don't know. I suppose taxes paid 30 years from now have less present value - so your reasoning might be partially correct. With your low level of leverage, your starting losses, and if you did some TLH in the future, you might be able to avoid investment taxes for a very long time.
Yeah, I understand we'd prefer to always have losses available to report at tax time. I was just making the point that when you still have losses in your back pocket, it makes even less sense to worry about tax drag and more reason to deploy equity leverage in taxable.
skierincolorado wrote: Wed Feb 22, 2023 11:20 am Even with the fees, historically TYD is more efficient than EDV. The one thing you'll have to check out is how big the bid/ask spreads are on TYA if it is ever allowed. I think for buy and hold it's probably OK but worth looking into. Occasional rebalancing should be fine too but really frequent rebalancing with a wide bid ask would be problematic.
Yeah, I'll roll with TYD to start.
comeinvest
Posts: 2669
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

The effects of monetary tightening

Image
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

Chocolatebar wrote: Wed Feb 22, 2023 11:56 am
skierincolorado wrote: Wed Feb 22, 2023 11:20 am One question would be, if the market goes down, which account would you increase leverage in first in order to maintain your overall AA of 87/36/54?
Always tax-advantaged first, but NEVER past 2x. After that - we shift attention to taxable.
After thinking about this, I think I should tone down the leverage in the Roth to start. That way we can bring it up to 2 if needed. It doesn't make sense to start at 2.

roth: (12k) (10%)
TYD - 50 (5)
UPRO - 50 (5)

Need to think about bringing the leverage up a little in the taxable now.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

comeinvest wrote: Wed Feb 22, 2023 3:23 pm The effects of monetary tightening

Image
I increased my exposure and sold a ZF and replaced it with 4 SOFR. I do like the way it looks now and it's interesting to see the instant rates at each part of the curve.

I also did some back of the napkin math and the first 8 quarters sofr contracts have an average yield nearly identical to the 2 year treasury (4.69%).
Last edited by skierincolorado on Wed Feb 22, 2023 10:40 pm, edited 1 time in total.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Wed Feb 22, 2023 4:09 pm
Chocolatebar wrote: Wed Feb 22, 2023 11:56 am
skierincolorado wrote: Wed Feb 22, 2023 11:20 am One question would be, if the market goes down, which account would you increase leverage in first in order to maintain your overall AA of 87/36/54?
Always tax-advantaged first, but NEVER past 2x. After that - we shift attention to taxable.
After thinking about this, I think I should tone down the leverage in the Roth to start. That way we can bring it up to 2 if needed. It doesn't make sense to start at 2.

roth: (12k) (10%)
TYD - 50 (5)
UPRO - 50 (5)

Need to think about bringing the leverage up a little in the taxable now.
Even with your original, the leverage in the taxable (1.05x) was already quite close to the overall target (1.23x). With your edit it would be even closer (maybe 1.12x vs 1.23x). But because the IRA is still more leveraged, you'll have to be constantly manually modulating the AA in each account in order to maintain the target AA. Slightly skewing the leverage to the IRA is probably a good idea, as long as the target overall AA is maintained. The 2x in your IRA to start was definitely too much because basically you shouldn't take it any higher, so if the market goes down you have to lever up the taxable. Levering up the taxable at market lows might actually be less tax efficient because you'd be siphoning more of your gains into there. The IRA gets hit by volatility which is made up for by bigger gains in taxable where you're fighting the volatility decay of the IRA. Might actually be less tax efficient. Starting with 1.5x in the IRA might work, because for smaller market drops you could just lever up the IRA (up to a limit). You'd be surprised how quickly you'd have to lever the IRA up given how small it is relative to taxable. One 5% dip in the market, and the IRA is already close to 2x.

I'd save all the hassle for a gain that may not even be there. The AA that keeps your original target would be across all accounts:

20.5/22.5/9/9/9/21 UPRO/AVUV/VEA/AVDV/VWO/AVES/TYA

The taxable account only increases leverage slightly from what you posted originally, but it saves all the hassle of trying to preserve a target overall AA across 2 accounts with different starting AAs.

Or you could bring the Roth down to ~1.4x equity leverage, taxable up to ~1.18x and just keep them as separate fixed allocations for the time being. At that point the leverage in the IRA might be close enough to the leverage in your taxable that it's not experiencing too much additional volatility decay. If you wanted to go through the trouble of leveraging it up when the market drops in order to maintain your overall target AA, you could choose to do so. But you also wouldn't have to worry about it if you didn't get to it. Your AA would become very slightly less leveraged in a market drop, but it wouldn't be a big deal and wouldn't enhance your volatility decay too much. In a big market drop (like >30%) your overall AA might drift from 1.23x equity leverage to like 1.15, and if you noticed that, you could(should) up the leverage in the IRA first and taxable if necessary. But at least your overall AA won't be wildly swinging on every 10% market move requiring frequent manual rebalancing.

So yeah just set the leverages either close enough in each account that your overall portfolio AA isn't going to drift substantially in a market crash (1.18x in taxable and 1.4x in IRA), or just set them to be the same (1.23x in both). Or just be prepared to have to manually rebalance frequently if the starting leverages are substantially different in each account.

I think you have a good plan at this point. If you do get to it in the future there is a lot of value in lifecycle investing and it would be good to try to implement that (with LETFs or futures/boxes). Lifecycle investing is important. As much time is spent on this whole modern portfolio diversification optimizatoin stuff, the most important part is arguably lifecycle investing which is most often neglected. Even if you don't get a solid plan in place, I'd consider having it in the back of your mind that you should increase leverage if the market drops (a hard thing to do psychologically without a firm plan). You should also decrease leverage as the market goes up and/or as you get wealthier. Of course, theoretically you are starting out underleveraged (lifecycle investing would probably tell you to be between 1.5x and 2x depending on how much you expect to save in the future) so that might not be for a while.
comeinvest
Posts: 2669
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

skierincolorado wrote: Wed Feb 22, 2023 9:51 pm
comeinvest wrote: Wed Feb 22, 2023 3:23 pm The effects of monetary tightening

Image
I increased my exposure and sold a ZF and replaced it with 4 SOFR. I do like the way it looks now and it's interesting to see the instant rates at each part of the curve.

I also did some back of the napkin math and the first 8 quarters sofr contracts have an average yield nearly identical to the 2 year treasury (4.69%).
I think it's difficult to do a performance attribution, though, of the term premium earned from the SOFR futures vs for example the /ZT future (something I would like to do for verification of this implementation of duration exposure). Because you would have to match the durations which are constantly changing for the /ZT future; then you would have to adjust for the slightly different equivalent maturities because the performance can greatly depend on that as the screenshot shows; and then the term premia earned over a timeframe will be minuscule compared to fluctuations from rate changes.

I almost forgot to roll the treasury futures that have a deadline Monday morning; what a hassle. I'm tempted to convert the /ZF and /ZN to a SOFR strip one day.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Wed Feb 22, 2023 10:15 pm Or you could bring the Roth down to ~1.4x equity leverage, taxable up to ~1.18x and just keep them as separate fixed allocations for the time being. At that point the leverage in the IRA might be close enough to the leverage in your taxable that it's not experiencing too much additional volatility decay. If you wanted to go through the trouble of leveraging it up when the market drops in order to maintain your overall target AA, you could choose to do so. But you also wouldn't have to worry about it if you didn't get to it. Your AA would become very slightly less leveraged in a market drop, but it wouldn't be a big deal and wouldn't enhance your volatility decay too much. In a big market drop (like >30%) your overall AA might drift from 1.23x equity leverage to like 1.15, and if you noticed that, you could(should) up the leverage in the IRA first and taxable if necessary. But at least your overall AA won't be wildly swinging on every 10% market move requiring frequent manual rebalancing.

So yeah just set the leverages either close enough in each account that your overall portfolio AA isn't going to drift substantially in a market crash (1.18x in taxable and 1.4x in IRA), or just set them to be the same (1.23x in both). Or just be prepared to have to manually rebalance frequently if the starting leverages are substantially different in each account.
How does this look:
roth: (12k) (~10%)
tyd - 53 (5.3) x 3 = 16
upro - 47 (4.7) x 3 = 14

taxable: (109k) (~90%)
upro - 17 (15) x 3 = 45
avuv - 25 (22.5)
vea - 10 (9)
avdv - 10 (9)
vwo - 10 (9)
aves - 10 (9)
tyd - 18 (16) x 3 = 48

roth leverage: 1.4x
taxable leverage: 1.16x
total leverage: 1.175x

I definitely want to keep the leverage in my roth slightly higher at all times if possible in hopes it can get closer to my taxable balance over time. I think this would be a good start. Another thing I realized is that if I don't buy a house in a few months (looking extremely unlikely atm), then I may have >40k available to yeet into my taxable. I'm sure that'll require me to recalc the allocations. I really like the targets for the small caps and internationals. It'll just depend on what my roth looks like by that point. I'm not concerned with making adjustments to it at this point as long as I make sure to keep the leverage in roth only slightly higher than taxable.

EDIT - Do you think I need to rebalance this monthly or would quarterly be fine/preferred? I want to rebalance as little as possible, but obviously will keep an eye on things and make adjustments when significant corrections or rallies happen.
skierincolorado wrote: Wed Feb 22, 2023 10:15 pm I think you have a good plan at this point. If you do get to it in the future there is a lot of value in lifecycle investing and it would be good to try to implement that (with LETFs or futures/boxes). Lifecycle investing is important. As much time is spent on this whole modern portfolio diversification optimizatoin stuff, the most important part is arguably lifecycle investing which is most often neglected. Even if you don't get a solid plan in place, I'd consider having it in the back of your mind that you should increase leverage if the market drops (a hard thing to do psychologically without a firm plan). You should also decrease leverage as the market goes up and/or as you get wealthier. Of course, theoretically you are starting out underleveraged (lifecycle investing would probably tell you to be between 1.5x and 2x depending on how much you expect to save in the future) so that might not be for a while.
Couple things I've been meaning to collect your thoughts on. I know you don't like LETFs, but were you surprised to see in the backtests that they typically provided better outcomes for individuals with only a 20 year timeline compared to other leverage methods?

Also I know you are a big fan of Lifecycle Investing. I am too, but I don't think it's perfect/reasonable for most people. I think you agree it doesn't make sense for most people, but I also think for most of the people it does make sense for - there are still things to consider:
  • It basically says we should account for future contributions now - I understand that we'll all likely have future contributions to commit to our retirement funds, but what happens in extreme circumstances where you or a loved one become too ill to work?
  • Also life changes and our goals do too. For example, I'm only comfortable deploying this strategy because I have a job I love and believe I can do it for 30 years, but what if I suddenly don't/can't? What if I end up going through a period of hating my job (like I have before in life) and struggling to find one that makes me happy?
I know the strategy might have answers to these and I may have missed them, but I think they are basically unmeasurable risks. In my mind, that's why it seems reasonable to me for leverage to remain fluid throughout one's lifetime. I know you are ideally supposed to deleverage as time goes on which naturally accounts for these things later in life, but starting out it's impossible to say when one may truly need to draw from their retirement money.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Thu Feb 23, 2023 7:34 am

How does this look:
roth: (12k) (~10%)
tyd - 53 (5.3) x 3 = 16
upro - 47 (4.7) x 3 = 14

taxable: (109k) (~90%)
upro - 17 (15) x 3 = 45
avuv - 25 (22.5)
vea - 10 (9)
avdv - 10 (9)
vwo - 10 (9)
aves - 10 (9)
tyd - 18 (16) x 3 = 48

roth leverage: 1.4x
taxable leverage: 1.16x
total leverage: 1.175x

I definitely want to keep the leverage in my roth slightly higher at all times if possible in hopes it can get closer to my taxable balance over time. I think this would be a good start. Another thing I realized is that if I don't buy a house in a few months (looking extremely unlikely atm), then I may have >40k available to yeet into my taxable. I'm sure that'll require me to recalc the allocations. I really like the targets for the small caps and internationals. It'll just depend on what my roth looks like by that point. I'm not concerned with making adjustments to it at this point as long as I make sure to keep the leverage in roth only slightly higher than taxable.
They're probably close enough to run separately with M1 rebalancing each account internally. I liked the other one more though because the overall leverage was 1.23x instead of 1.175x and just guessing based on your age and wealth but lifecycle investing would probably say 1.5x+ overall.

Also from a lifecycle investing perspective, the market has dropped so much just in the time we've been having this conversation, you should be adding .05x or more to the leverage based on that alone.
Chocolatebar wrote: Thu Feb 23, 2023 7:34 am EDIT - Do you think I need to rebalance this monthly or would quarterly be fine/preferred? I want to rebalance as little as possible, but obviously will keep an eye on things and make adjustments when significant corrections or rallies happen.
If rebalancing the accounts internally, I think that unless the market makes major moves you probably wouldn't need to rebalance on any sort of schedule, especially if you bump the taxable up a hair more because at that point the leverages in the two accounts will not be too different and you won't see your overall AA drift unless the market moves a lot. Just let M1 rebalance the accounts internally. If the market drops a lot, do what you have to do get the leverage back to your original target (or preferably even higher per lifecycle investing).


Chocolatebar wrote: Thu Feb 23, 2023 7:34 am

Couple things I've been meaning to collect your thoughts on. I know you don't like LETFs, but were you surprised to see in the backtests that they typically provided better outcomes for individuals with only a 20 year timeline compared to other leverage methods?
Sorry I'm not sure what you mean by this. Other methods of leverage are hard to test, but in my attempts to do so other methods do better because they experience less volatility decay
Chocolatebar wrote: Thu Feb 23, 2023 7:34 am Also I know you are a big fan of Lifecycle Investing. I am too, but I don't think it's perfect/reasonable for most people. I think you agree it doesn't make sense for most people, but I also think for most of the people it does make sense for - there are still things to consider:
  • It basically says we should account for future contributions now - I understand that we'll all likely have future contributions to commit to our retirement funds, but what happens in extreme circumstances where you or a loved one become too ill to work?
  • Also life changes and our goals do too. For example, I'm only comfortable deploying this strategy because I have a job I love and believe I can do it for 30 years, but what if I suddenly don't/can't? What if I end up going through a period of hating my job (like I have before in life) and struggling to find one that makes me happy?
I know the strategy might have answers to these and I may have missed them, but I think they are basically unmeasurable risks. In my mind, that's why it seems reasonable to me for leverage to remain fluid throughout one's lifetime. I know you are ideally supposed to deleverage as time goes on which naturally accounts for these things later in life, but starting out it's impossible to say when one may truly need to draw from their retirement money.
The answer to the first question is life and disability insurance.

I think for the second question, you have to factor in these risks. Personally I feel that I am very likely to continue accumulating wealth for a while longer barring injury or death. The lowest I would stop is probably 1M. I can reduce my sequence of return risk and lengthen my investment horizon by investing more of that 1M today. Even with leverage, I only have about half that exposure to equities today. By the time I get to 1M I would expect I would be mostly de-leveraged, unless my income increases substantially and I plan to continue working a while longer.

I actually think for all people lifecycle investing makes sense. That's the whole point. And anybody that uses a target date retirement fund is essentially using lifecycle investing (but with an arbitrary upper bound of 1x leverage). Having a glidepath is applicable for anybody who is saving and contributing for retirement (glidepaths are probably also partially justified by investment horizon). Most people do accumulate wealth and investments over their life. These people would have benefited from spreading these investments across time to reduce risk and lengthen their investment horizon. You should only invest and leverage money that you are confident will stay invested. It should only be leveraged if you are confident it will be added to over time. For people with median income or higher, those tenets likely are true the large majority of the time. Just think of all the 20 and 30 somethings that are at least making the matching contributions to their 401k.
Last edited by skierincolorado on Thu Feb 23, 2023 12:58 pm, edited 1 time in total.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

comeinvest wrote: Thu Feb 23, 2023 12:15 am
skierincolorado wrote: Wed Feb 22, 2023 9:51 pm
comeinvest wrote: Wed Feb 22, 2023 3:23 pm The effects of monetary tightening

Image
I increased my exposure and sold a ZF and replaced it with 4 SOFR. I do like the way it looks now and it's interesting to see the instant rates at each part of the curve.

I also did some back of the napkin math and the first 8 quarters sofr contracts have an average yield nearly identical to the 2 year treasury (4.69%).
I think it's difficult to do a performance attribution, though, of the term premium earned from the SOFR futures vs for example the /ZT future (something I would like to do for verification of this implementation of duration exposure). Because you would have to match the durations which are constantly changing for the /ZT future; then you would have to adjust for the slightly different equivalent maturities because the performance can greatly depend on that as the screenshot shows; and then the term premia earned over a timeframe will be minuscule compared to fluctuations from rate changes.

I almost forgot to roll the treasury futures that have a deadline Monday morning; what a hassle. I'm tempted to convert the /ZF and /ZN to a SOFR strip one day.
The duration for the SOFRs is constantly changing too as we add new SOFR contracts though. It's not identical but pretty dang close. Like the cheapest to deliver security is getting younger as our SOFRs get younger, then the CTD probably switches when the ZT roles, but we also add a SOFR at that time as well (except I'm only doing every other). Just ballparking but these slight differences and adjustments shouldn't cause the theoretical yield to be more than like .05% different from just taking a simple average of the first 8 contracts.

I also read something that said when the SOFRs first came out there was arbitrage with eurodollars, but once the contract became more widely used the arbitrage was eliminated and they traded in tight alignment.

Admittedly I'm not 100% confident. If I was, I would also be very tempted to change everything to SOFR. I did manage to eliminate 1 roll with my new design.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Thu Feb 23, 2023 12:39 pm They're probably close enough to run separately with M1 rebalancing each account internally. I liked the other one more though because the overall leverage was 1.23x instead of 1.175x and just guessing based on your age and wealth but lifecycle investing would probably say 1.5x+ overall.

Also from a lifecycle investing perspective, the market has dropped so much just in the time we've been having this conversation, you should be adding .05x or more to the leverage based on that alone.
Chocolatebar wrote: Thu Feb 23, 2023 7:34 am EDIT - Do you think I need to rebalance this monthly or would quarterly be fine/preferred? I want to rebalance as little as possible, but obviously will keep an eye on things and make adjustments when significant corrections or rallies happen.
If rebalancing the accounts internally, I think that unless the market makes major moves you probably wouldn't need to rebalance on any sort of schedule, especially if you bump the taxable up a hair more because at that point the leverages in the two accounts will not be too different and you won't see your overall AA drift unless the market moves a lot. Just let M1 rebalance the accounts internally. If the market drops a lot, do what you have to do get the leverage back to your original target (or preferably even higher per lifecycle investing).
I really like these allocations, so I actually think what I'll do is start with them and let my leverage rise over time to ~1.5x. If the market goes down, I'll change allocations to get closer, but I do think as my Roth gets bigger, it'll be easier to maintain more leverage without sacrificing international diversification. Although, if I keep doing annual lump sums of 40k+ into my taxable after failing to buy a home, my Roth may never catch up haha.
skierincolorado wrote: Thu Feb 23, 2023 12:39 pm
Chocolatebar wrote: Thu Feb 23, 2023 7:34 am Couple things I've been meaning to collect your thoughts on. I know you don't like LETFs, but were you surprised to see in the backtests that they typically provided better outcomes for individuals with only a 20 year timeline compared to other leverage methods?
Sorry I'm not sure what you mean by this. Other methods of leverage are hard to test, but in my attempts to do so other methods do better because they experience less volatility decay
Oh sorry I should've been specific. I was referring to the "SIMPLE" doc in the OP. It states:
Analyzing the results for a shorter investment horizon of 20 years, we see that the benefits of leverage are significantly reduced. While strategies with a constant leveraged blend of equities and treasuries still increase median outcomes, they no longer produce better 10th percentile outcomes than a traditional glidepath strategy. For a 20-year horizon, these strategies can be classified as high risk, high reward.

The lifecycle strategy, on the other hand, does increase both the 10th and 50th percentiles relative to the traditional glidepath, regardless of the method of leverage used. In fact, leveraged ETFs perform the best in these backtests, likely due to the daily resetting of leverage reducing risk over a shorter timespan.

For investors with a financial situation similar to this scenario, the lifecycle strategy using leveraged ETFs appears to be the optimal choice, with the added perk of being simple to implement.
skierincolorado wrote: Thu Feb 23, 2023 12:39 pm
Chocolatebar wrote: Thu Feb 23, 2023 7:34 am Also I know you are a big fan of Lifecycle Investing. I am too, but I don't think it's perfect/reasonable for most people. I think you agree it doesn't make sense for most people, but I also think for most of the people it does make sense for - there are still things to consider:
  • It basically says we should account for future contributions now - I understand that we'll all likely have future contributions to commit to our retirement funds, but what happens in extreme circumstances where you or a loved one become too ill to work?
  • Also life changes and our goals do too. For example, I'm only comfortable deploying this strategy because I have a job I love and believe I can do it for 30 years, but what if I suddenly don't/can't? What if I end up going through a period of hating my job (like I have before in life) and struggling to find one that makes me happy?
I know the strategy might have answers to these and I may have missed them, but I think they are basically unmeasurable risks. In my mind, that's why it seems reasonable to me for leverage to remain fluid throughout one's lifetime. I know you are ideally supposed to deleverage as time goes on which naturally accounts for these things later in life, but starting out it's impossible to say when one may truly need to draw from their retirement money.
The answer to the first question is life and disability insurance.

I think for the second question, you have to factor in these risks. Personally I feel that I am very likely to continue accumulating wealth for a while longer barring injury or death. The lowest I would stop is probably 1M. I can reduce my sequence of return risk and lengthen my investment horizon by investing more of that 1M today. Even with leverage, I only have about half that exposure to equities today. By the time I get to 1M I would expect I would be mostly de-leveraged, unless my income increases substantially and I plan to continue working a while longer.

I actually think for all people lifecycle investing makes sense. That's the whole point. And anybody that uses a target date retirement fund is essentially using lifecycle investing (but with an arbitrary upper bound of 1x leverage). Having a glidepath is applicable for anybody who is saving and contributing for retirement (glidepaths are probably also partially justified by investment horizon). Most people do accumulate wealth and investments over their life. These people would have benefited from spreading these investments across time to reduce risk and lengthen their investment horizon. You should only invest and leverage money that you are confident will stay invested. It should only be leveraged if you are confident it will be added to over time. For people with median income or higher, those tenets likely are true the large majority of the time. Just think of all the 20 and 30 somethings that are at least making the matching contributions to their 401k.
Yeah, I think as long as I have to work in my field to pay the bills, I can justify leverage. I know a lot of people though who aren't fortunate enough to be able to consistently contribute towards retirement. Unexpected expenses come up and they need to pause contributions indefinitely. Do you think Lifecycle Investing still makes sense for them? I feel like one of the requirements is to be able to consistently contribute. For years I haven't even been able to make consistent contributions because I've been saving for a house. I'm sure there will be years where things will come up (or I'll invest in something outside of the stock market like real estate) and I won't be able to contribute to my taxable at all!
comeinvest
Posts: 2669
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

skierincolorado wrote: Thu Feb 23, 2023 12:56 pm
comeinvest wrote: Thu Feb 23, 2023 12:15 am
skierincolorado wrote: Wed Feb 22, 2023 9:51 pm
comeinvest wrote: Wed Feb 22, 2023 3:23 pm The effects of monetary tightening

Image
I increased my exposure and sold a ZF and replaced it with 4 SOFR. I do like the way it looks now and it's interesting to see the instant rates at each part of the curve.

I also did some back of the napkin math and the first 8 quarters sofr contracts have an average yield nearly identical to the 2 year treasury (4.69%).
I think it's difficult to do a performance attribution, though, of the term premium earned from the SOFR futures vs for example the /ZT future (something I would like to do for verification of this implementation of duration exposure). Because you would have to match the durations which are constantly changing for the /ZT future; then you would have to adjust for the slightly different equivalent maturities because the performance can greatly depend on that as the screenshot shows; and then the term premia earned over a timeframe will be minuscule compared to fluctuations from rate changes.

I almost forgot to roll the treasury futures that have a deadline Monday morning; what a hassle. I'm tempted to convert the /ZF and /ZN to a SOFR strip one day.
The duration for the SOFRs is constantly changing too as we add new SOFR contracts though. It's not identical but pretty dang close. Like the cheapest to deliver security is getting younger as our SOFRs get younger, then the CTD probably switches when the ZT roles, but we also add a SOFR at that time as well (except I'm only doing every other). Just ballparking but these slight differences and adjustments shouldn't cause the theoretical yield to be more than like .05% different from just taking a simple average of the first 8 contracts.

I also read something that said when the SOFRs first came out there was arbitrage with eurodollars, but once the contract became more widely used the arbitrage was eliminated and they traded in tight alignment.

Admittedly I'm not 100% confident. If I was, I would also be very tempted to change everything to SOFR. I did manage to eliminate 1 roll with my new design.
I was more thinking of comparing the performance of the two implementations over a 3-months period or so, not over longer timeframes on a portfolio level which would be even more complicated. The time periods of SOFR futures 3-months average calculation period to determine the settlement price, and treasury futures' CTD maturities, are not perfectly aligned. That introduces an error, especially with fluctuating forward rates on the forward timeline like we are seeing now. You could do some extrapolation, but there will be a margin of error. Also, the CTD and by implication the treasury futures will be affected by coupon payments, which will tilt the exposures to instantaneous forward rates on the timeline. You could adjust for that manually, but it's a lot of work. If there was for example a systematic bias resulting in a 0.1% p.a. difference in implementation efficiency of duration risk exposure between the two implementations, and if you were to compare the performance over 3 months for example, you would have to detect a 0.025% performance difference by doing a lot of detailed math and cancelling out noise.
Another way would be comparing performance and risk (measured volatility or downside risk) of both methods over longer timeframes; but we don't have historical futures data.

Of course the bias could be either direction, if we suspect a systematic bias. Last quarter there was some erratic behavior of treasury futures calendar rolls.

What you read may have been about the Eurodollar/SOFR spread that is now fixed by law to 0.26161% yield difference for existing legal contracts referring to Eurodollar. But that doesn't validate the assumption that the SOFR curve and by implication the roll-down returns will behave the same as historically the Eurodollar curve and Eurodollar futures, and that there won't be any systematic bias. The spread has fluctuated in the past. TED spread was the spread between 3-mo. LIBOR and T-bills; not referring to futures. SOFR vs Eurodollar is one thing; another is SOFR vs treasuries, and a third potential source of bias, or differences in implied cost, could be futures vs underlyings.

If you base your strategy on the two Eurodollar futures papers that were referenced in this thread: ED futures were forward rates on future LIBOR rates; who says they didn't reflect an additional risk-based term premium, i.e. an overestimation of realized TED spreads?

Thinking of it, your approach of comparing yields has some appeal, but is based on several assumptions. For example, that the SOFR spot rates (not sure if I use the correct terminology) is identical to the ultra-short-term treasury yields - which is not the case. SOFR and treasury curves are slightly different, which means that the roll-down glide path and by implication the returns over a period are different.

I read some papers that Eurodollar and SOFR futures are relatively accurate hedges for interest rate risk, but cannot locate them right now.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Thu Feb 23, 2023 1:28 pm
I really like these allocations, so I actually think what I'll do is start with them and let my leverage rise over time to ~1.5x. If the market goes down, I'll change allocations to get closer, but I do think as my Roth gets bigger, it'll be easier to maintain more leverage without sacrificing international diversification. Although, if I keep doing annual lump sums of 40k+ into my taxable after failing to buy a home, my Roth may never catch up haha.
I seriously doubt the IRA would ever catch up, even without contributions to taxable and the 6.5k to the IRA. If the taxable grows by 6% real and the IRA grows by 7% real plus 6500, I calculate that it will take 26 years for the taxable to catch up. And that assumes zero contributions to taxable.
Chocolatebar wrote: Thu Feb 23, 2023 1:28 pm
Oh sorry I should've been specific. I was referring to the "SIMPLE" doc in the OP. It states:

Analyzing the results for a shorter investment horizon of 20 years, we see that the benefits of leverage are significantly reduced. While strategies with a constant leveraged blend of equities and treasuries still increase median outcomes, they no longer produce better 10th percentile outcomes than a traditional glidepath strategy. For a 20-year horizon, these strategies can be classified as high risk, high reward.

The lifecycle strategy, on the other hand, does increase both the 10th and 50th percentiles relative to the traditional glidepath, regardless of the method of leverage used. In fact, leveraged ETFs perform the best in these backtests, likely due to the daily resetting of leverage reducing risk over a shorter timespan.

For investors with a financial situation similar to this scenario, the lifecycle strategy using leveraged ETFs appears to be the optimal choice, with the added perk of being simple to implement.
The LETFs win in the spreadsheet because the rebalancing is annual and the level of leverage is quite high in most of the sims. The non-LETF sims get extremely high volatility decay because they are rebalancing back to target only once per year. If we're comparing a 2x LETF and 2x non-LETF sim, in a bear market year the 2x non-LETF will actually get the full 2x bear market return. So if the market goes down 25%, the non-LETF version goes down 50%. But the LETF version would only go down 35-40%. The non-LETF version is now 3x leveraged, and the sim will rebalance it back to 2x. So this is actually making the volatility decay much worse in the non-LETF versions when the leverage is high. In the lifecycle sims the leverae is capped at 2x, and it runs at 2x for much of the sim. So through that whole period, the leverage is often bouncing up to 2.5x++ mid-year and not shown by the sim, and then rebalanced back to 2x by the sim at the start of the year. If the leverage is less than 2x, and it's lifecycle investing (not fixed allocation) then it's not as bad because the sim should be increasing the leverage in a bear market (up to cap of 2x if necessary), so the annual rebalancing isn't as bad.

That took me a while to work out by looking through some specific scenarios. Basically the annual rebalancing of high leverage ratios is problematic and causes severe volatility decay. It needs to be monthly.

There's some other issues with the sims as well you should be aware of 1) there's not enough data (cohorts) because it starts in 1955 and with 35 working years, that only gives us 30 cohorts, which is too small of a sample size 2) we need to consider the 1st or at least 5th percentile, but there's not really enough data to do that reliably 3) the optimistic cost of leverage is accurate for futures 4) the fixed allocation sims look great because they just run high levels of leverage for a really long time which worked out in most of these scenarios, but on a larger data set looking at the 1st or 5th percentile they would look very risky 5) especially for the 35 yr sim running leverage for longer, rather than glidepathing, looks good because the end dates are all between 1990 and 2020 and stocks and bonds had a big run up in the 80s and 90s that make holding leverage through the 80s and 90s look good. If we had looked at a cohort retiring in 1965, holding leverage up until the year of retire would not look nearly so good.

Chocolatebar wrote: Thu Feb 23, 2023 1:28 pm
Yeah, I think as long as I have to work in my field to pay the bills, I can justify leverage. I know a lot of people though who aren't fortunate enough to be able to consistently contribute towards retirement. Unexpected expenses come up and they need to pause contributions indefinitely. Do you think Lifecycle Investing still makes sense for them? I feel like one of the requirements is to be able to consistently contribute. For years I haven't even been able to make consistent contributions because I've been saving for a house. I'm sure there will be years where things will come up (or I'll invest in something outside of the stock market like real estate) and I won't be able to contribute to my taxable at all!
I don't think you really have to consistently contribute. You just can't spend the money all of a sudden. Not only should you not leverage if that's the case, you shouldn't invest in anything but shorter duration bonds. Like half this country never really accumulates significant wealth, so this isn't for them. But for the other half of people who do accumulate wealth as they age, some form of lifecycle investing makes sense. Really just on the money that's never going to be touched until retirement (401ks). The money that might get spent shouldn't be invested at all (other than short term bonds). A more efficient way would be to invest all your money in stocks (no leverage) and then when you spend all the money in taxable, leverage up the 401k. I mean people just don't have the foresight to plan all this. Like I have zero expectation most people could do this. But if people were more informed, theoretically it would make sense, even if their spending and earning habits didn't change, at least for the half of people who are likely to accumulate significant wealth over their life (even if there are some bumps along the way).

I guess what I'm getting at is that it's a pretty durable fact that people with above median income get wealthier with age. Like look at median wealth charts by age. The median net worth in late 20s is 7.5k. The median net worth in late 40s is 164k. The median net worth in early 60s is 230k.

30% of people in their early 60s have more than 550k net worth. 20% have over a million, compared to just 1% of 30 year olds. Yeah a bunch of it is home equity, but the general principal holds. Human capital is converted to capital with time. People generally aren't lazy - they have jobs and work long careers.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Fri Feb 24, 2023 12:40 am
Chocolatebar wrote: Thu Feb 23, 2023 1:28 pm
I really like these allocations, so I actually think what I'll do is start with them and let my leverage rise over time to ~1.5x. If the market goes down, I'll change allocations to get closer, but I do think as my Roth gets bigger, it'll be easier to maintain more leverage without sacrificing international diversification. Although, if I keep doing annual lump sums of 40k+ into my taxable after failing to buy a home, my Roth may never catch up haha.
I seriously doubt the IRA would ever catch up, even without contributions to taxable and the 6.5k to the IRA. If the taxable grows by 6% real and the IRA grows by 7% real plus 6500, I calculate that it will take 26 years for the taxable to catch up. And that assumes zero contributions to taxable.
I agree that it'll probably never catch up, but my hope is that it'll at least be comparable 20 years from now. Although things could change if I ever begin making too much to contribute to the Roth or I roll my 401k into it. I should actually look into whether my employer would let me just do an "in-service" rollover. That would improve my prospects tremendously.
skierincolorado wrote: Fri Feb 24, 2023 12:40 am The LETFs win in the spreadsheet because the rebalancing is annual and the level of leverage is quite high in most of the sims. The non-LETF sims get extremely high volatility decay because they are rebalancing back to target only once per year. If we're comparing a 2x LETF and 2x non-LETF sim, in a bear market year the 2x non-LETF will actually get the full 2x bear market return. So if the market goes down 25%, the non-LETF version goes down 50%. But the LETF version would only go down 35-40%. The non-LETF version is now 3x leveraged, and the sim will rebalance it back to 2x. So this is actually making the volatility decay much worse in the non-LETF versions when the leverage is high. In the lifecycle sims the leverae is capped at 2x, and it runs at 2x for much of the sim. So through that whole period, the leverage is often bouncing up to 2.5x++ mid-year and not shown by the sim, and then rebalanced back to 2x by the sim at the start of the year. If the leverage is less than 2x, and it's lifecycle investing (not fixed allocation) then it's not as bad because the sim should be increasing the leverage in a bear market (up to cap of 2x if necessary), so the annual rebalancing isn't as bad.

That took me a while to work out by looking through some specific scenarios. Basically the annual rebalancing of high leverage ratios is problematic and causes severe volatility decay. It needs to be monthly.

There's some other issues with the sims as well you should be aware of 1) there's not enough data (cohorts) because it starts in 1955 and with 35 working years, that only gives us 30 cohorts, which is too small of a sample size 2) we need to consider the 1st or at least 5th percentile, but there's not really enough data to do that reliably 3) the optimistic cost of leverage is accurate for futures 4) the fixed allocation sims look great because they just run high levels of leverage for a really long time which worked out in most of these scenarios, but on a larger data set looking at the 1st or 5th percentile they would look very risky 5) especially for the 35 yr sim running leverage for longer, rather than glidepathing, looks good because the end dates are all between 1990 and 2020 and stocks and bonds had a big run up in the 80s and 90s that make holding leverage through the 80s and 90s look good. If we had looked at a cohort retiring in 1965, holding leverage up until the year of retire would not look nearly so good.
Yeah, those are all great points!
skierincolorado wrote: Fri Feb 24, 2023 12:40 am
Chocolatebar wrote: Thu Feb 23, 2023 1:28 pm Yeah, I think as long as I have to work in my field to pay the bills, I can justify leverage. I know a lot of people though who aren't fortunate enough to be able to consistently contribute towards retirement. Unexpected expenses come up and they need to pause contributions indefinitely. Do you think Lifecycle Investing still makes sense for them? I feel like one of the requirements is to be able to consistently contribute. For years I haven't even been able to make consistent contributions because I've been saving for a house. I'm sure there will be years where things will come up (or I'll invest in something outside of the stock market like real estate) and I won't be able to contribute to my taxable at all!
I don't think you really have to consistently contribute. You just can't spend the money all of a sudden. Not only should you not leverage if that's the case, you shouldn't invest in anything but shorter duration bonds. Like half this country never really accumulates significant wealth, so this isn't for them. But for the other half of people who do accumulate wealth as they age, some form of lifecycle investing makes sense. Really just on the money that's never going to be touched until retirement (401ks). The money that might get spent shouldn't be invested at all (other than short term bonds). A more efficient way would be to invest all your money in stocks (no leverage) and then when you spend all the money in taxable, leverage up the 401k. I mean people just don't have the foresight to plan all this. Like I have zero expectation most people could do this. But if people were more informed, theoretically it would make sense, even if their spending and earning habits didn't change, at least for the half of people who are likely to accumulate significant wealth over their life (even if there are some bumps along the way).

I guess what I'm getting at is that it's a pretty durable fact that people with above median income get wealthier with age. Like look at median wealth charts by age. The median net worth in late 20s is 7.5k. The median net worth in late 40s is 164k. The median net worth in early 60s is 230k.

30% of people in their early 60s have more than 550k net worth. 20% have over a million, compared to just 1% of 30 year olds. Yeah a bunch of it is home equity, but the general principal holds. Human capital is converted to capital with time. People generally aren't lazy - they have jobs and work long careers.
That's a better way to look at it. Whenever I have extra money in my checking at the end of the month I ask myself: "When do I expect to need this?" If the answer is retirement - it goes to my retirement accounts and stays there until retirement. Crazy things do happen in life though and I know plenty of people who have unfortunately had to withdraw from their retirement funds early, but I think that's why it's super important to have an emergency fund.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Fri Feb 24, 2023 5:12 am
That's a better way to look at it. Whenever I have extra money in my checking at the end of the month I ask myself: "When do I expect to need this?" If the answer is retirement - it goes to my retirement accounts and stays there until retirement. Crazy things do happen in life though and I know plenty of people who have unfortunately had to withdraw from their retirement funds early, but I think that's why it's super important to have an emergency fund.
I wish I had known how to leverage earlier because when we made our house purchase it took a big chunk out of my savings. I had stupidly invested all but 10k in stocks so had to sell like 35k for the down payment near the covid market low. We had kind of given up buying a house so I had reinvested it. I knew having to sell was bad at the time but didn't know how to leverage to maintain my exposure. I even looked at margin rates at my broker but they were way too high. Ideally instead of not investing the down payment I would have invested it but with a plan to up my leverage when I sold. Or you can not invest the down payment and have higher leverage on the rest of your portfolio. As long as the down payment is earning a competitive rate on cash they are functionally identical (the rate on cash probably won't ever be quite as high as the borrow rate implicit in letfs or futures unless you do some bank bonuses).

120/60/30/-80 stocks/bonds/savings acct/borrowing is exactly the same as 100/50/-50 srocks/bonds/borrowing. The cash nets out. Except the latter is usually a hair better because borrowing rates are usually a hair higher than savings account rates. Depending on how big the spread between borrow rates and savings rates is, the taxes on selling the invested down payment could cancel any benefit. You'd probably save about $1000 per year by investing your down payment to reduce your borrowing, but then possibly have a tax bill that cancels that benefit.

I like to think of how much I've borrowed. Right now you've borrowed 20k to buy stock and 60k to buy bonds. Of course money is fungible and you can't really say whether the money was borrowed to buy stock or bond, but I like to think of it this way in comparison to the alternative of being 100% equities without any borrowing. Thinking of it this way is helpful to the psychology of it I think. Ignoring the bonds which are relatively safe, you've borrowed 20k to buy stock. That's probably what a years worth of contributions? So you've pre-bought one year of contributions. If you did not change the number of shares of stock you own, you would pay back the loan in one year.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Fri Feb 24, 2023 9:28 am
Chocolatebar wrote: Fri Feb 24, 2023 5:12 am
That's a better way to look at it. Whenever I have extra money in my checking at the end of the month I ask myself: "When do I expect to need this?" If the answer is retirement - it goes to my retirement accounts and stays there until retirement. Crazy things do happen in life though and I know plenty of people who have unfortunately had to withdraw from their retirement funds early, but I think that's why it's super important to have an emergency fund.
I wish I had known how to leverage earlier because when we made our house purchase it took a big chunk out of my savings. I had stupidly invested all but 10k in stocks so had to sell like 35k for the down payment near the covid market low. We had kind of given up buying a house so I had reinvested it. I knew having to sell was bad at the time but didn't know how to leverage to maintain my exposure. I even looked at margin rates at my broker but they were way too high. Ideally instead of not investing the down payment I would have invested it but with a plan to up my leverage when I sold. Or you can not invest the down payment and have higher leverage on the rest of your portfolio. As long as the down payment is earning a competitive rate on cash they are functionally identical (the rate on cash probably won't ever be quite as high as the borrow rate implicit in letfs or futures unless you do some bank bonuses).

120/60/30/-80 stocks/bonds/savings acct/borrowing is exactly the same as 100/50/-50 srocks/bonds/borrowing. The cash nets out. Except the latter is usually a hair better because borrowing rates are usually a hair higher than savings account rates. Depending on how big the spread between borrow rates and savings rates is, the taxes on selling the invested down payment could cancel any benefit. You'd probably save about $1000 per year by investing your down payment to reduce your borrowing, but then possibly have a tax bill that cancels that benefit.

I like to think of how much I've borrowed. Right now you've borrowed 20k to buy stock and 60k to buy bonds. Of course money is fungible and you can't really say whether the money was borrowed to buy stock or bond, but I like to think of it this way in comparison to the alternative of being 100% equities without any borrowing. Thinking of it this way is helpful to the psychology of it I think. Ignoring the bonds which are relatively safe, you've borrowed 20k to buy stock. That's probably what a years worth of contributions? So you've pre-bought one year of contributions. If you did not change the number of shares of stock you own, you would pay back the loan in one year.
I just consider my down payment and my emergency fund the same. I'll never be comfortable not having cash available personally. I really want real estate to be a significant part of my portfolio, so I can justify having cash available as always being in a position to buy property. It's just another form of leverage (and diversification) if you think about it. It's probably even the safest form of leverage.

BTW it turns out I can transfer 11k from my 401k to my Roth. Need to talk to someone next week to figure out if it's worth paying taxes on it or if it would be better to move it another way tax-free over time.
comeinvest
Posts: 2669
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

/ZF and /ZN implied financing cost dropped today by ca. 0.1% p.a. compared to yesterday for the next 3-4 months, if my math is right. Unfortunately I bought the calendar rolls yesterday.
All the while the Dec 2022 and the Mar 2023 SOFR which cover the period of the next 4 months, stayed almost unchanged or slightly dropped indicating unchanged or slightly increasing interest rates.
I have a hard time reconciling this, or is my logic wrong?
The directions of the entire movements over the last 2 weeks are even more in conflict, or not?

Image

Image

Image

Image
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Fri Feb 24, 2023 10:29 am
skierincolorado wrote: Fri Feb 24, 2023 9:28 am
Chocolatebar wrote: Fri Feb 24, 2023 5:12 am
That's a better way to look at it. Whenever I have extra money in my checking at the end of the month I ask myself: "When do I expect to need this?" If the answer is retirement - it goes to my retirement accounts and stays there until retirement. Crazy things do happen in life though and I know plenty of people who have unfortunately had to withdraw from their retirement funds early, but I think that's why it's super important to have an emergency fund.
I wish I had known how to leverage earlier because when we made our house purchase it took a big chunk out of my savings. I had stupidly invested all but 10k in stocks so had to sell like 35k for the down payment near the covid market low. We had kind of given up buying a house so I had reinvested it. I knew having to sell was bad at the time but didn't know how to leverage to maintain my exposure. I even looked at margin rates at my broker but they were way too high. Ideally instead of not investing the down payment I would have invested it but with a plan to up my leverage when I sold. Or you can not invest the down payment and have higher leverage on the rest of your portfolio. As long as the down payment is earning a competitive rate on cash they are functionally identical (the rate on cash probably won't ever be quite as high as the borrow rate implicit in letfs or futures unless you do some bank bonuses).

120/60/30/-80 stocks/bonds/savings acct/borrowing is exactly the same as 100/50/-50 srocks/bonds/borrowing. The cash nets out. Except the latter is usually a hair better because borrowing rates are usually a hair higher than savings account rates. Depending on how big the spread between borrow rates and savings rates is, the taxes on selling the invested down payment could cancel any benefit. You'd probably save about $1000 per year by investing your down payment to reduce your borrowing, but then possibly have a tax bill that cancels that benefit.

I like to think of how much I've borrowed. Right now you've borrowed 20k to buy stock and 60k to buy bonds. Of course money is fungible and you can't really say whether the money was borrowed to buy stock or bond, but I like to think of it this way in comparison to the alternative of being 100% equities without any borrowing. Thinking of it this way is helpful to the psychology of it I think. Ignoring the bonds which are relatively safe, you've borrowed 20k to buy stock. That's probably what a years worth of contributions? So you've pre-bought one year of contributions. If you did not change the number of shares of stock you own, you would pay back the loan in one year.
I just consider my down payment and my emergency fund the same. I'll never be comfortable not having cash available personally. I really want real estate to be a significant part of my portfolio, so I can justify having cash available as always being in a position to buy property. It's just another form of leverage (and diversification) if you think about it. It's probably even the safest form of leverage.

BTW it turns out I can transfer 11k from my 401k to my Roth. Need to talk to someone next week to figure out if it's worth paying taxes on it or if it would be better to move it another way tax-free over time.
Having 100k invested + 50k borrowed&invested + 50k cash is exactly the same as having 150k invested. I cannot think of a single scenario in which the outcome would differ. The only difference is that the latter captures the spread between cash and borrowing rates but also may incur a tax cost if sold. I don't keep an emergency fund for this reason. Money is fungible. Borrowed money is fungible with cash.

If the 401k is all pre-tax, I wouldn't transfer to Roth unless you would contribute to Roth instead of pre-tax if given the choice. It all comes down to your current marginal tax rate vs your future marginal tax rate. Generally most people with high savings rates should be contributing at least some money to Roth unless they are at peak-career in a high tax bracket. Otherwise they risk excessive RMDs in retirement. Your annual Roth IRA might be enough already. Personally I do about 19k/year pretax and 10k Roth (I have access to Roth 401k and HSA).
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

comeinvest wrote: Fri Feb 24, 2023 11:11 am /ZF and /ZN implied financing cost dropped today by ca. 0.1% p.a. compared to yesterday for the next 3-4 months, if my math is right. Unfortunately I bought the calendar rolls yesterday.
All the while the Dec 2022 and the Mar 2023 SOFR which cover the period of the next 4 months, stayed almost unchanged or slightly dropped indicating unchanged or slightly increasing interest rates.
I have a hard time reconciling this, or is my logic wrong?
The directions of the entire movements over the last 2 weeks are even more in conflict, or not?

Image

Image

Image

Image
I don't think the spread is really the same as the implied financing right? They're different securities. The June contract probably has a longer dated maturity. The roll price would decrease when the 0-4 year spread flattens. Which is what we've been seeing this month and today. The part of the curve that the June contract includes went up in rates more. So the June contract will decrease in price, and make the roll cheaper.
comeinvest
Posts: 2669
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

skierincolorado wrote: Fri Feb 24, 2023 12:32 pm
comeinvest wrote: Fri Feb 24, 2023 11:11 am /ZF and /ZN implied financing cost dropped today by ca. 0.1% p.a. compared to yesterday for the next 3-4 months, if my math is right. Unfortunately I bought the calendar rolls yesterday.
All the while the Dec 2022 and the Mar 2023 SOFR which cover the period of the next 4 months, stayed almost unchanged or slightly dropped indicating unchanged or slightly increasing interest rates.
I have a hard time reconciling this, or is my logic wrong?
The directions of the entire movements over the last 2 weeks are even more in conflict, or not?

Image

Image

Image

Image
I don't think the spread is really the same as the implied financing right? They're different securities. The June contract probably has a longer dated maturity. The roll price would decrease when the 0-4 year spread flattens. Which is what we've been seeing this month and today. The part of the curve that the June contract includes went up in rates more. So the June contract will decrease in price, and make the roll cheaper.
The CTD of the March 2023 /ZN contract matures 10/31/2029. The CTD of the June 2023 /ZN contract matures 12/31/2029. https://www.cmegroup.com/tools-informat ... ytics.html That's a 2 months difference in almost 7 years. That wouldn't move the needle, right? The rest of the calendar roll price is the implied financing cost over the next 3 months, right?
The 7y treasury yield to maturity is ca. 4.1%. The forward rate for 4Q 2029 is ca. 3.6%.

Image

But we don't have to speculate. The two underlying CTDs moved by almost exactly the same amount today, making me believe that we had a change in implied financing cost of ca. 0.1% in just one day:

Image

If unemployed physicist was still here, he could provide the absolute implied financing cost :)

The two CTDs over the last 2 weeks: (IB has an option to create a chart of yields, but it looks like I don't have the required data permissions.)

Image
Last edited by comeinvest on Fri Feb 24, 2023 2:04 pm, edited 4 times in total.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Fri Feb 24, 2023 12:02 pm
Chocolatebar wrote: Fri Feb 24, 2023 10:29 am
skierincolorado wrote: Fri Feb 24, 2023 9:28 am
Chocolatebar wrote: Fri Feb 24, 2023 5:12 am
That's a better way to look at it. Whenever I have extra money in my checking at the end of the month I ask myself: "When do I expect to need this?" If the answer is retirement - it goes to my retirement accounts and stays there until retirement. Crazy things do happen in life though and I know plenty of people who have unfortunately had to withdraw from their retirement funds early, but I think that's why it's super important to have an emergency fund.
I wish I had known how to leverage earlier because when we made our house purchase it took a big chunk out of my savings. I had stupidly invested all but 10k in stocks so had to sell like 35k for the down payment near the covid market low. We had kind of given up buying a house so I had reinvested it. I knew having to sell was bad at the time but didn't know how to leverage to maintain my exposure. I even looked at margin rates at my broker but they were way too high. Ideally instead of not investing the down payment I would have invested it but with a plan to up my leverage when I sold. Or you can not invest the down payment and have higher leverage on the rest of your portfolio. As long as the down payment is earning a competitive rate on cash they are functionally identical (the rate on cash probably won't ever be quite as high as the borrow rate implicit in letfs or futures unless you do some bank bonuses).

120/60/30/-80 stocks/bonds/savings acct/borrowing is exactly the same as 100/50/-50 srocks/bonds/borrowing. The cash nets out. Except the latter is usually a hair better because borrowing rates are usually a hair higher than savings account rates. Depending on how big the spread between borrow rates and savings rates is, the taxes on selling the invested down payment could cancel any benefit. You'd probably save about $1000 per year by investing your down payment to reduce your borrowing, but then possibly have a tax bill that cancels that benefit.

I like to think of how much I've borrowed. Right now you've borrowed 20k to buy stock and 60k to buy bonds. Of course money is fungible and you can't really say whether the money was borrowed to buy stock or bond, but I like to think of it this way in comparison to the alternative of being 100% equities without any borrowing. Thinking of it this way is helpful to the psychology of it I think. Ignoring the bonds which are relatively safe, you've borrowed 20k to buy stock. That's probably what a years worth of contributions? So you've pre-bought one year of contributions. If you did not change the number of shares of stock you own, you would pay back the loan in one year.
I just consider my down payment and my emergency fund the same. I'll never be comfortable not having cash available personally. I really want real estate to be a significant part of my portfolio, so I can justify having cash available as always being in a position to buy property. It's just another form of leverage (and diversification) if you think about it. It's probably even the safest form of leverage.

BTW it turns out I can transfer 11k from my 401k to my Roth. Need to talk to someone next week to figure out if it's worth paying taxes on it or if it would be better to move it another way tax-free over time.
Having 100k invested + 50k borrowed&invested + 50k cash is exactly the same as having 150k invested. I cannot think of a single scenario in which the outcome would differ. The only difference is that the latter captures the spread between cash and borrowing rates but also may incur a tax cost if sold. I don't keep an emergency fund for this reason. Money is fungible. Borrowed money is fungible with cash.

If the 401k is all pre-tax, I wouldn't transfer to Roth unless you would contribute to Roth instead of pre-tax if given the choice. It all comes down to your current marginal tax rate vs your future marginal tax rate. Generally most people with high savings rates should be contributing at least some money to Roth unless they are at peak-career in a high tax bracket. Otherwise they risk excessive RMDs in retirement. Your annual Roth IRA might be enough already. Personally I do about 19k/year pretax and 10k Roth (I have access to Roth 401k and HSA).
What would happen though in a 2008-style crash? What if the market crashed and you lost your job in the same period? You'd have no emergency fund to cover your expenses for the next few months+ right?

EDIT - I do feel pretty confident that my income will rise over time. Leaning towards not going through the effort of rolling over the funds since they are all pre-tax and it would be a bit of a pain to move them. I actually would rather contribute to Roth over pre-tax if I had to choose 1. I should probably think about whether it's worth maxing out my 401k each year like I have been at this point. I suppose it beats taxable. Sucks that I can't leverage any of it though.
Last edited by Chocolatebar on Fri Feb 24, 2023 2:31 pm, edited 1 time in total.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Chocolatebar wrote: Fri Feb 24, 2023 1:18 pm
skierincolorado wrote: Fri Feb 24, 2023 12:02 pm
Chocolatebar wrote: Fri Feb 24, 2023 10:29 am
skierincolorado wrote: Fri Feb 24, 2023 9:28 am
Chocolatebar wrote: Fri Feb 24, 2023 5:12 am
That's a better way to look at it. Whenever I have extra money in my checking at the end of the month I ask myself: "When do I expect to need this?" If the answer is retirement - it goes to my retirement accounts and stays there until retirement. Crazy things do happen in life though and I know plenty of people who have unfortunately had to withdraw from their retirement funds early, but I think that's why it's super important to have an emergency fund.
I wish I had known how to leverage earlier because when we made our house purchase it took a big chunk out of my savings. I had stupidly invested all but 10k in stocks so had to sell like 35k for the down payment near the covid market low. We had kind of given up buying a house so I had reinvested it. I knew having to sell was bad at the time but didn't know how to leverage to maintain my exposure. I even looked at margin rates at my broker but they were way too high. Ideally instead of not investing the down payment I would have invested it but with a plan to up my leverage when I sold. Or you can not invest the down payment and have higher leverage on the rest of your portfolio. As long as the down payment is earning a competitive rate on cash they are functionally identical (the rate on cash probably won't ever be quite as high as the borrow rate implicit in letfs or futures unless you do some bank bonuses).

120/60/30/-80 stocks/bonds/savings acct/borrowing is exactly the same as 100/50/-50 srocks/bonds/borrowing. The cash nets out. Except the latter is usually a hair better because borrowing rates are usually a hair higher than savings account rates. Depending on how big the spread between borrow rates and savings rates is, the taxes on selling the invested down payment could cancel any benefit. You'd probably save about $1000 per year by investing your down payment to reduce your borrowing, but then possibly have a tax bill that cancels that benefit.

I like to think of how much I've borrowed. Right now you've borrowed 20k to buy stock and 60k to buy bonds. Of course money is fungible and you can't really say whether the money was borrowed to buy stock or bond, but I like to think of it this way in comparison to the alternative of being 100% equities without any borrowing. Thinking of it this way is helpful to the psychology of it I think. Ignoring the bonds which are relatively safe, you've borrowed 20k to buy stock. That's probably what a years worth of contributions? So you've pre-bought one year of contributions. If you did not change the number of shares of stock you own, you would pay back the loan in one year.
I just consider my down payment and my emergency fund the same. I'll never be comfortable not having cash available personally. I really want real estate to be a significant part of my portfolio, so I can justify having cash available as always being in a position to buy property. It's just another form of leverage (and diversification) if you think about it. It's probably even the safest form of leverage.

BTW it turns out I can transfer 11k from my 401k to my Roth. Need to talk to someone next week to figure out if it's worth paying taxes on it or if it would be better to move it another way tax-free over time.
Having 100k invested + 50k borrowed&invested + 50k cash is exactly the same as having 150k invested. I cannot think of a single scenario in which the outcome would differ. The only difference is that the latter captures the spread between cash and borrowing rates but also may incur a tax cost if sold. I don't keep an emergency fund for this reason. Money is fungible. Borrowed money is fungible with cash.

If the 401k is all pre-tax, I wouldn't transfer to Roth unless you would contribute to Roth instead of pre-tax if given the choice. It all comes down to your current marginal tax rate vs your future marginal tax rate. Generally most people with high savings rates should be contributing at least some money to Roth unless they are at peak-career in a high tax bracket. Otherwise they risk excessive RMDs in retirement. Your annual Roth IRA might be enough already. Personally I do about 19k/year pretax and 10k Roth (I have access to Roth 401k and HSA).
What would happen though in a 2008-style crash? What if the market crashed and you lost your job in the same period? You'd have no emergency fund to cover your expenses for the next few months+ right?

EDIT - I do feel pretty confident that my income will rise over time. Leaning towards not going through the effort of rolling over the funds since they are all pre-tax and it would be a bit of a pain to move them. I actually would rather contribute to Roth over pre-tax if I had to choose 1. I should probably think about whether it's worth maxing out my 401k each year like I have been at this point. I suppose it beats taxable. Sucks that I can't leverage any of it though.

By borrowing less today, I would simply increase the borrowing in an emergency or planned future expense. If I had 50k in cash today I could simply dump it into my brokerage to pay off 50k in loans. If I lost my job, I would simply withdraw the 50k again putting me right back in the exact same situation as if I had never put the money into my brokerage account in the first place. And by exact, I mean literally exact. It's like it never happened and the money had been sitting in my bank account the whole time. Money is fungible. It's important to have mental clarity of these concepts to maximize efficiencies.

If you believe you should be contributing to Roth based on your current vs future marginal tax rate, then you should convert the 11k from 401k to the Roth ira. That is functionally exactly equivalent to contributing to a roth directly.

Whether you really should be contributing more to Roth than pretax I would be careful. Do you really expect your marginal tax rate in retirement to be higher than at present? If most of your money is Roth, you will have no taxable income in retirement and will have recognized income at higher rates today than you would have in retirement.

Personally I underestimated how much my income would increase and probably should have been doing more Roth 401k when I was young.
comeinvest
Posts: 2669
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

skierincolorado wrote: Wed Feb 22, 2023 10:15 pm 20.5/22.5/9/9/9/21 UPRO/AVUV/VEA/AVDV/VWO/AVES/TYA
TYA now has a competitor: TUA. (Not quite the same, because you won't get the same duration exposure; but supposedly you would outperform ITT).

TUA says it tries to match the duration of ITT, and also to outperform ITT. Their current positions indicate ca. 600% STT; I'm a bit confused because that wouldn't match the 7-10y benchmark (/ZT duration 1.9y; 7y treasury duration 5.8y; 10y treasury duration 8y).

In any case, monitoring this ETF will be another way of validating the thesis of tilting the duration risk to short maturities.
Last edited by comeinvest on Fri Feb 24, 2023 4:27 pm, edited 1 time in total.
Chocolatebar
Posts: 192
Joined: Tue Feb 14, 2023 10:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Chocolatebar »

skierincolorado wrote: Fri Feb 24, 2023 3:19 pm By borrowing less today, I would simply increase the borrowing in an emergency or planned future expense. If I had 50k in cash today I could simply dump it into my brokerage to pay off 50k in loans. If I lost my job, I would simply withdraw the 50k again putting me right back in the exact same situation as if I had never put the money into my brokerage account in the first place. And by exact, I mean literally exact. It's like it never happened and the money had been sitting in my bank account the whole time. Money is fungible. It's important to have mental clarity of these concepts to maximize efficiencies.
Forgive my ignorance (as you have pretty much done throughout this entire thread lol), but how can you be sure you'll be able to withdraw 50k in the wake of a historic market crash (assuming you have to borrow against assets that just lost value)? I know fear holds us back, but I feel like I can't put a price tag on the peace of mind a 3 month emergency fund gives me.
skierincolorado wrote: Fri Feb 24, 2023 3:19 pm If you believe you should be contributing to Roth based on your current vs future marginal tax rate, then you should convert the 11k from 401k to the Roth ira. That is functionally exactly equivalent to contributing to a roth directly.

Whether you really should be contributing more to Roth than pretax I would be careful. Do you really expect your marginal tax rate in retirement to be higher than at present? If most of your money is Roth, you will have no taxable income in retirement and will have recognized income at higher rates today than you would have in retirement.

Personally I underestimated how much my income would increase and probably should have been doing more Roth 401k when I was young.
I can't make a prediction on whether tax rates will be higher or lower in the future, but I'm confident as long as I have my current job for 20+ years and I invest in real estate that my income will be higher in retirement.
comeinvest
Posts: 2669
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

Nothing new to the astute followers of this thread, and I can't remember if this was posted before, but here is a 1990-2018 study. I'm posting 3 charts from it. Notice the historically large divergence of the 2y bonds vs treasury futures performance in 2017-2018 in the second chart: about 1.5% divergence within about 2 years, eyeballing the chart.
But also notice how the fed funds rate didn't drop below 3% in the 1990ies, and stayed higher than that for decades before (not shown in this chart). So don't count on anything happening.

https://www.advisorperspectives.com/art ... sury-bonds
https://www.advisorperspectives.com/art ... res-market

Image

Image

Image
Post Reply