Momentum and frequency of rebalancing
- martincmartin
- Posts: 900
- Joined: Wed Jul 02, 2014 3:04 pm
- Location: Boston, MA USA
Momentum and frequency of rebalancing
Just a reminder that the stock market isn't a complete random walk, there's a little bit of momentum.
This week was the 7th straight week of S&P 500 decline. Strings like this, or corresponding positive strings when markets are doing well, are more common than if each week's gains were chosen independently.
For the Boglehead, the take away is to not rebalance too often. Bernstein and McClung say once a year is about right.
Personally, I rebalance in the last week of December if I have losses to harvest, or first week of January for gains.
This week was the 7th straight week of S&P 500 decline. Strings like this, or corresponding positive strings when markets are doing well, are more common than if each week's gains were chosen independently.
For the Boglehead, the take away is to not rebalance too often. Bernstein and McClung say once a year is about right.
Personally, I rebalance in the last week of December if I have losses to harvest, or first week of January for gains.
Re: Momentum and frequency of rebalancing
Momentum is a thing, and so is Mean-reversion, that doesn't mean anyone can successfully time those to trade it at a profit.
Rebalancing isn't a trading scheme expected to have a positive return in all environments. If rebalancing is good for anything, it's to maintain ones "risk profile". To make sure you don't have too much exposure to any particular risk.
Rebalancing isn't a trading scheme expected to have a positive return in all environments. If rebalancing is good for anything, it's to maintain ones "risk profile". To make sure you don't have too much exposure to any particular risk.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
Re: Momentum and frequency of rebalancing
I use wide bands for re-balancing which sort of takes advantage of momentum (both directions). Historically, I had symmetric bands -- ie. +/- X% points. In 2018 I removed my band on the upside so I am now -5%/+55% (my equity "target" is nominally 45%). The only thing I have to resolve is whether I will ratchet. In other words, if my equity blows past 50% will my new "target" be 50% -5%/+50% or will I retain the 45% and wait to rebalance on the downside when equity gets to 40%? TBD.
Don't trust me, look it up. https://www.irs.gov/forms-instructions-and-publications
Re: Momentum and frequency of rebalancing
Several recent research papers on rebalancing state that the "optimal" algorithm is to check every two weeks and only rebalance if a category is off by more than 20%. As these studies were based on back-testing, some skepticism is warranted. Since I adopted this approach several years ago, I have rebalanced zero times.
YMMV.
Edit: These studies assume a particular style of sector investing. Broad index investing (in my case) might have different parameters for optimal rebalancing: likely a narrower band.
YMMV.
Edit: These studies assume a particular style of sector investing. Broad index investing (in my case) might have different parameters for optimal rebalancing: likely a narrower band.
Last edited by bh1 on Sat May 21, 2022 10:32 am, edited 1 time in total.
Re: Momentum and frequency of rebalancing
Momentum is a thing. However, you must read into the economic context for it to work, and you must set a few boundaries.
For example, suppose your timeframe is 5 years. The assets that returned the most during the last cycle were tech stocks.
Will tech stocks still provide good returns during the next 5 years? Or will inflation be so large that commodities will outperform?
Or should you "hide" into fixed income?
If that were easy, we'd all be rich.
My educated guess is that it will be very hard for the US to control current inflation, especially with so much debt. So, for the next 5 years, interest rates will either hike or stay high.
Of course, if a populist gets into office and chops down interest rates to zero again, that's another matter. What you can do is to try to optimize your risk according to your risk tolerance.
For example, suppose your timeframe is 5 years. The assets that returned the most during the last cycle were tech stocks.
Will tech stocks still provide good returns during the next 5 years? Or will inflation be so large that commodities will outperform?
Or should you "hide" into fixed income?
If that were easy, we'd all be rich.
My educated guess is that it will be very hard for the US to control current inflation, especially with so much debt. So, for the next 5 years, interest rates will either hike or stay high.
Of course, if a populist gets into office and chops down interest rates to zero again, that's another matter. What you can do is to try to optimize your risk according to your risk tolerance.
Re: Momentum and frequency of rebalancing
so, if you have, say a target of 60% equity and 40% fixed income you rebalance if equity drops below 40% or goes above 80%?bh1 wrote: ↑Sat May 21, 2022 10:22 am Several recent research papers on rebalancing state that the "optimal" algorithm is to check every two weeks and only rebalance if a category is off by more than 20%. As these studies were based on back-testing, some skepticism is warranted. Since I adopted this approach several years ago, I have rebalanced zero times.
YMMV.
Don't trust me, look it up. https://www.irs.gov/forms-instructions-and-publications
Re: Momentum and frequency of rebalancing
FED officials are not elected.
Don't trust me, look it up. https://www.irs.gov/forms-instructions-and-publications
Re: Momentum and frequency of rebalancing
My reading of one of the papers is that you calculate the dollars that should be in a sector, based on your AA. If the actual dollars is 20% high or low, then adjust that category. My SP500 completion index should be about 6% of my portfolio. It has been 15% low twice in the last few years, including now. 20% low would be (4/5)6% = 4.8%.
The second paper I found I didn't read in detail, because rebalancing doesn't make that much difference. I was just looking for a rule of thumb that seemed rationally derived and stopped when I found it.
Re: Momentum and frequency of rebalancing
I've been considering the same thing.jebmke wrote: ↑Sat May 21, 2022 8:31 am I use wide bands for re-balancing which sort of takes advantage of momentum (both directions). Historically, I had symmetric bands -- ie. +/- X% points. In 2018 I removed my band on the upside so I am now -5%/+55% (my equity "target" is nominally 45%). The only thing I have to resolve is whether I will ratchet. In other words, if my equity blows past 50% will my new "target" be 50% -5%/+50% or will I retain the 45% and wait to rebalance on the downside when equity gets to 40%? TBD.
I don't think that I have ever sold equities to rebalance anyway since all income and contributions go into fixed income. I might as well acknowledge it.
As to the second point about changing the lower equity action point I will probably increase it as time goes by. We have enough income to fund our needs and our portfolio is likely all going to our grandchildren and the (UGH) government.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.
- martincmartin
- Posts: 900
- Joined: Wed Jul 02, 2014 3:04 pm
- Location: Boston, MA USA
Re: Momentum and frequency of rebalancing
Interesting. Do you have a reference?bh1 wrote: ↑Sat May 21, 2022 10:22 am Several recent research papers on rebalancing state that the "optimal" algorithm is to check every two weeks and only rebalance if a category is off by more than 20%. As these studies were based on back-testing, some skepticism is warranted. Since I adopted this approach several years ago, I have rebalanced zero times.
YMMV.
Edit: These studies assume a particular style of sector investing. Broad index investing (in my case) might have different parameters for optimal rebalancing: likely a narrower band.
Re: Momentum and frequency of rebalancing
I searched for a few minutes on my computer but didn't find anything I saved. I'll leave the google hunt to the reader.
Re: Momentum and frequency of rebalancing
That's probably this old reference:
http://resource.fpanet.org/resource/09B ... yanani.pdf
Linked from bogleheads.org/wiki/rebalancing
Many of these rebalancing studies use month-end data and miss some intra-month opportunities where markets fall 20% before the end of the month, but rise again before the end of the month.
Re: Momentum and frequency of rebalancing
martincmartin wrote: ↑Sat May 21, 2022 10:49 amInteresting. Do you have a reference?bh1 wrote: ↑Sat May 21, 2022 10:22 am Several recent research papers on rebalancing state that the "optimal" algorithm is to check every two weeks and only rebalance if a category is off by more than 20%. As these studies were based on back-testing, some skepticism is warranted. Since I adopted this approach several years ago, I have rebalanced zero times.
YMMV.
Edit: These studies assume a particular style of sector investing. Broad index investing (in my case) might have different parameters for optimal rebalancing: likely a narrower band.
This recent paper might not be exactly on point but is certainly relevant: https://faculty.fuqua.duke.edu/~charvey ... ancing.pdf
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Re: Momentum and frequency of rebalancing
And this article uses monthly data.vineviz wrote: ↑Sat May 21, 2022 12:33 pmThis recent paper might not be exactly on point but is certainly relevant: https://faculty.fuqua.duke.edu/~charvey ... ancing.pdf
In contrast, siamond has a 3-part series on rebalancing
viewtopic.php?t=322344 in the Bogleheads blog that does address some of this.
Re: Momentum and frequency of rebalancing
Sure, but there's no solid evidence that rebalancing more frequently than monthly offers any benefits much less enough of them to offset the costs.livesoft wrote: ↑Sat May 21, 2022 12:46 pmAnd this article uses monthly data.vineviz wrote: ↑Sat May 21, 2022 12:33 pmThis recent paper might not be exactly on point but is certainly relevant: https://faculty.fuqua.duke.edu/~charvey ... ancing.pdf
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Re: Momentum and frequency of rebalancing
OK, so what?
It should be intuitively obvious that the best time to rebalance into equities is when the market has reached a significant low. That will not happen every month nor is it likely to happen on the first trading day of the month nor on the last trading day of the month.
In my opinion, any rebalancing study that uses periodic rebalancing based on the calendar is flawed. You don't have to agree with my opinion.
Re: Momentum and frequency of rebalancing
It might be intuitive, but it's not necessarily true. And the evidence we do have suggests it's NOT true.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
-
- Posts: 1488
- Joined: Wed Aug 08, 2018 6:34 am
Re: Momentum and frequency of rebalancing
If optimizing a 50-50 portfolio for 4% SWR, mathematically optimal frequency is 6 years. (since 1926)
If young and accumulating, never rebalance....stay 100% stocks...(unless of course, valuations reach nosebleed levels)
If young and accumulating, never rebalance....stay 100% stocks...(unless of course, valuations reach nosebleed levels)
Re: Momentum and frequency of rebalancing
These are precisely the types of questions that are amplified when working with LETFs. I recently posted a long entry here that documents some outcomes of my grappling with these issues. My goal has been to use 3x LETFs in a way that gives a much smoother portfolio ride than the underlying 3x LETFs while being largely agnostic to market conditions (sort of a supercharged permanent portfolio, in a way).
IMO, rebalancing accomplishes two things: (i) reset to the desired risk profile and (ii) take advantage of a rebalancing bonus from volatility. These operate on different time scales. When one is setting the risk profile to represent long time frames (e.g., >5 years), one ends up with a "fixed" asset allocation and you can get away with much larger intervals between rebalances. When one is setting the risk profile to represent shorter time frames (less than a year), momentum can be used to some extent but you need to rebalance frequently enough to take advantage (even weekly or biweekly with 3x LETFs).
The rebalancing bonus from volatility kicks in most powerfully with daily resets (and seems to be negligible with weekly resets), but one has to deal with slippage costs that may not make it worthwhile. My rough estimate is that the rebalancing bonus from daily rebalancing might increase CAGR by 1 percent (e.g., go from 7 to 8 percent) for unleveraged funds (not accounting for slippage), but may increase CAGR by 4 or 5 percent for 3x funds. Maybe not worth it for 1x funds, maybe worth it for 3x funds.
So rebalancing frequency should very much be tied to how volatile the portfolio assets are, how active you are willing to be, and how willing you are to accept timing luck for returns.
Just my 2 cents (1.84 cents after inflation).
IMO, rebalancing accomplishes two things: (i) reset to the desired risk profile and (ii) take advantage of a rebalancing bonus from volatility. These operate on different time scales. When one is setting the risk profile to represent long time frames (e.g., >5 years), one ends up with a "fixed" asset allocation and you can get away with much larger intervals between rebalances. When one is setting the risk profile to represent shorter time frames (less than a year), momentum can be used to some extent but you need to rebalance frequently enough to take advantage (even weekly or biweekly with 3x LETFs).
The rebalancing bonus from volatility kicks in most powerfully with daily resets (and seems to be negligible with weekly resets), but one has to deal with slippage costs that may not make it worthwhile. My rough estimate is that the rebalancing bonus from daily rebalancing might increase CAGR by 1 percent (e.g., go from 7 to 8 percent) for unleveraged funds (not accounting for slippage), but may increase CAGR by 4 or 5 percent for 3x funds. Maybe not worth it for 1x funds, maybe worth it for 3x funds.
So rebalancing frequency should very much be tied to how volatile the portfolio assets are, how active you are willing to be, and how willing you are to accept timing luck for returns.
Just my 2 cents (1.84 cents after inflation).
- martincmartin
- Posts: 900
- Joined: Wed Jul 02, 2014 3:04 pm
- Location: Boston, MA USA
Re: Momentum and frequency of rebalancing
Why would you have bonds and leverage at the same time? Leverage is borrowing money, bonds are loaning money. Why are you borrowing money to buy stocks, then separately loaning money to governments/companies? If you're using 3x leverage, it seems your desired risk profile is > 100% stocks, unless I'm missing something.Hydromod wrote: ↑Sun May 22, 2022 11:36 am These are precisely the types of questions that are amplified when working with LETFs. I recently posted a long entry here that documents some outcomes of my grappling with these issues. My goal has been to use 3x LETFs in a way that gives a much smoother portfolio ride than the underlying 3x LETFs while being largely agnostic to market conditions (sort of a supercharged permanent portfolio, in a way).
IMO, rebalancing accomplishes two things: (i) reset to the desired risk profile and (ii) take advantage of a rebalancing bonus from volatility. These operate on different time scales. When one is setting the risk profile to represent long time frames (e.g., >5 years), one ends up with a "fixed" asset allocation and you can get away with much larger intervals between rebalances. When one is setting the risk profile to represent shorter time frames (less than a year), momentum can be used to some extent but you need to rebalance frequently enough to take advantage (even weekly or biweekly with 3x LETFs).
The rebalancing bonus from volatility kicks in most powerfully with daily resets (and seems to be negligible with weekly resets), but one has to deal with slippage costs that may not make it worthwhile. My rough estimate is that the rebalancing bonus from daily rebalancing might increase CAGR by 1 percent (e.g., go from 7 to 8 percent) for unleveraged funds (not accounting for slippage), but may increase CAGR by 4 or 5 percent for 3x funds. Maybe not worth it for 1x funds, maybe worth it for 3x funds.
So rebalancing frequency should very much be tied to how volatile the portfolio assets are, how active you are willing to be, and how willing you are to accept timing luck for returns.
Just my 2 cents (1.84 cents after inflation).
- nisiprius
- Advisory Board
- Posts: 52215
- Joined: Thu Jul 26, 2007 9:33 am
- Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry
Re: Momentum and frequency of rebalancing
Yep, just like Cowles and Jones found in... 1937.
Values above 1.0 are "inertia" (now called "momentum"), values below are "reversals" (now called "mean reversion.")
Values above 1.0 are "inertia" (now called "momentum"), values below are "reversals" (now called "mean reversion.")
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
Re: Momentum and frequency of rebalancing
The stuff I talk about is an outgrowth of the Hedgefundie Excellent Adventure (HFEA) threads. The basic idea is that you are trying to create a portfolio that has a favorable Sharpe ratio, somewhere near the tangency on the efficient frontier, then leveraging that portfolio.martincmartin wrote: ↑Sun May 22, 2022 11:48 am Why would you have bonds and leverage at the same time? Leverage is borrowing money, bonds are loaning money. Why are you borrowing money to buy stocks, then separately loaning money to governments/companies? If you're using 3x leverage, it seems your desired risk profile is > 100% stocks, unless I'm missing something.
The levered tangency portfolio should have better risk-adjusted returns than stocks by themselves.
The portfolio has both stocks and bonds, so both need to have some leverage beyond a certain overall portfolio leverage.
The HFEA approach was developed with 3x LETFs (or 2x LETFs, for the more risk adverse) to get the leverage. In this case, the borrowing is under the hood of the LETF.
Others like to use methods that leverage 1x funds.
- firebirdparts
- Posts: 4412
- Joined: Thu Jun 13, 2019 4:21 pm
- Location: Southern Appalachia
Re: Momentum and frequency of rebalancing
I think quarterly is about as fast as I would contemplate but I certainly wasn’t feeling very good about it April 1. It felt really good in 2020 though. You win some and lose some.
If I was going to read the current market, I would say let’s believe the Fed. Let’s believe they will do what they expect to do. No hurry.
If I was going to read the current market, I would say let’s believe the Fed. Let’s believe they will do what they expect to do. No hurry.
This time is the same