HEDGEFUNDIE's excellent adventure Part II: The next journey

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

klaus14 wrote: Fri Sep 03, 2021 1:40 pm
skierincolorado wrote: Fri Sep 03, 2021 1:33 pm
klaus14 wrote: Fri Sep 03, 2021 12:47 pm
skierincolorado wrote: Thu Sep 02, 2021 4:17 pm
klaus14 wrote: Thu Sep 02, 2021 3:56 pm after reading this carry discussion i replaced my 1 /UB holding with 2 /ZB. /ZB seems to have more carry but also move similar to /UB. Like in the last 6 months yield curve flattened a bit and both /UB and /ZB interest rates declined similar amounts vs /ZN didn't even move. /ZB seems to be sweet spot to have some carry, equity risk reduction and BAB.
ZB (and UB) have done well recently (last 4 or 5 months). There is no denying that. However, over any extended period shorter durations have trounced ZB. The risk-adjusted retrurns of ZB are still quite poor - not as bad as UB, but still quite poor. I'd also point out the correlation of ZB and UB to stocks the last 5 months has been the opposite of what you want - it's been positive. ZB and UB have been going up the same time as the stock market.

ZB is still well within the range of experiencing negative effects from the BAB phenomenon. The higher yield and beta of a 17 year duration bond is appealing to yield chasing investors that are leverage constrained. Hence the popularity of TMF in this very thread! This yield chasing leads to overpricing and grossly inferior risk adjusted returns.

Personally I wouldn't own anything longer than ZN. I'd swap your 2 ZBs for 4 or 5 ZNs, or 6 or 7 ZFs, or some combo of the two. If the yield curve doesn't change at all, you'd nearly double your return from roll & carry alone. Doubling your roll & carry is a nice starting point, since in the long run that's where 100% of the return from bonds comes from. Betting on long-term rates falling forever is a losing proposition.

As the paper I posted a few posts back shows, the equity risk reduction is greatest for 4-7 year durations. You just have to own more.
I moved to ZN. It seems to have highest carry.
I used the CTD durations and futures yield curve from Treasury Analytics and seems like ZN has the highest yield per duration.

Method:
- Subtract financing cost from yield to find yield pickup. For ZN: 1.171%-0.1%=1.071%
- Scale its duration to some other contract, for example to ZF: 1.071*4.4412/6.3810= 0.74542%
- Compare this to ZF yield pickup: 0.7621%-0.1%=0.6621
===> ZN > ZF
Similarly ZN > UB, ZN > ZB and ZN > TN
Thanks! I need to pick up a contract in the next few days so this helps. One thing though, I get a duration for ZN of ~6.95 years with the CTD w/ a 8/15/2028 maturity. I also only see 1.115% as the yield. This gives me:

For ZN 1.115-.1 = 1.015 * 4.44/6.95 = .648% scaled to the 5 year
For ZF I get the same .6621%

So I get ZF ahead by a hair.

Second where did you get .1% for the financing cost? If we look at the Implied Repo Rate it is .03% for ZF and .11% for ZN. I think the financing cost is either equal to the IRR, or it's equal to IRR + current repo rate. I think it's the latter. So we need to add the current repo rate (.06) to these amounts. For ZF that would give .09 and for ZN it would be .17. If I'm doing this correctly, it would add to ZFs lead. The carry per unit duration would be nearly .1% higher for ZF at that point:

For ZN 1.115-.17 = .945 * 4.44/6.95 = .603%
For ZF .7621-.09 = .6721%


Of course one shouldn't just try to maximize carry, the market is efficient and can look at this too. The market may be giving higher carry to ZF because it expects the curve to flatten further. The higher carry is the reward for missing out on some of the flattening one might experience with ZN. I think the main thing though is that historically, nothing longer than ZN has performed well.
ZN duration is not 6.95 right? It is 6.381. Maybe you have a typo?
I remember reading somewhere financing cost is 3 month libor. Which is 0.12 according to this. I used 0.1 because it makes easy to do subtractions :)
implied repo there seems to fluctuate wildly so i just used a rule of thumb.

And disclaimer: i don't know if this method is a good approximation or not It made sense to me so i posted here to get feedback.
On the Analytics page the CTD for ZN has maturity of 8/15/2028. That's ~6.95 years isn't it? ZF is 2/28/2026, or ~4.4 years.

The IRR does seem to move around a lot, but I'm still fairly confident that IRR+repo rate represents the financing cost based on the definition of IRR.

https://www.cmegroup.com/education/cour ... s-ctd.html
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by klaus14 »

skierincolorado wrote: Fri Sep 03, 2021 1:47 pm
klaus14 wrote: Fri Sep 03, 2021 1:40 pm
skierincolorado wrote: Fri Sep 03, 2021 1:33 pm
klaus14 wrote: Fri Sep 03, 2021 12:47 pm
skierincolorado wrote: Thu Sep 02, 2021 4:17 pm

ZB (and UB) have done well recently (last 4 or 5 months). There is no denying that. However, over any extended period shorter durations have trounced ZB. The risk-adjusted retrurns of ZB are still quite poor - not as bad as UB, but still quite poor. I'd also point out the correlation of ZB and UB to stocks the last 5 months has been the opposite of what you want - it's been positive. ZB and UB have been going up the same time as the stock market.

ZB is still well within the range of experiencing negative effects from the BAB phenomenon. The higher yield and beta of a 17 year duration bond is appealing to yield chasing investors that are leverage constrained. Hence the popularity of TMF in this very thread! This yield chasing leads to overpricing and grossly inferior risk adjusted returns.

Personally I wouldn't own anything longer than ZN. I'd swap your 2 ZBs for 4 or 5 ZNs, or 6 or 7 ZFs, or some combo of the two. If the yield curve doesn't change at all, you'd nearly double your return from roll & carry alone. Doubling your roll & carry is a nice starting point, since in the long run that's where 100% of the return from bonds comes from. Betting on long-term rates falling forever is a losing proposition.

As the paper I posted a few posts back shows, the equity risk reduction is greatest for 4-7 year durations. You just have to own more.
I moved to ZN. It seems to have highest carry.
I used the CTD durations and futures yield curve from Treasury Analytics and seems like ZN has the highest yield per duration.

Method:
- Subtract financing cost from yield to find yield pickup. For ZN: 1.171%-0.1%=1.071%
- Scale its duration to some other contract, for example to ZF: 1.071*4.4412/6.3810= 0.74542%
- Compare this to ZF yield pickup: 0.7621%-0.1%=0.6621
===> ZN > ZF
Similarly ZN > UB, ZN > ZB and ZN > TN
Thanks! I need to pick up a contract in the next few days so this helps. One thing though, I get a duration for ZN of ~6.95 years with the CTD w/ a 8/15/2028 maturity. I also only see 1.115% as the yield. This gives me:

For ZN 1.115-.1 = 1.015 * 4.44/6.95 = .648% scaled to the 5 year
For ZF I get the same .6621%

So I get ZF ahead by a hair.

Second where did you get .1% for the financing cost? If we look at the Implied Repo Rate it is .03% for ZF and .11% for ZN. I think the financing cost is either equal to the IRR, or it's equal to IRR + current repo rate. I think it's the latter. So we need to add the current repo rate (.06) to these amounts. For ZF that would give .09 and for ZN it would be .17. If I'm doing this correctly, it would add to ZFs lead. The carry per unit duration would be nearly .1% higher for ZF at that point:

For ZN 1.115-.17 = .945 * 4.44/6.95 = .603%
For ZF .7621-.09 = .6721%


Of course one shouldn't just try to maximize carry, the market is efficient and can look at this too. The market may be giving higher carry to ZF because it expects the curve to flatten further. The higher carry is the reward for missing out on some of the flattening one might experience with ZN. I think the main thing though is that historically, nothing longer than ZN has performed well.
ZN duration is not 6.95 right? It is 6.381. Maybe you have a typo?
I remember reading somewhere financing cost is 3 month libor. Which is 0.12 according to this. I used 0.1 because it makes easy to do subtractions :)
implied repo there seems to fluctuate wildly so i just used a rule of thumb.

And disclaimer: i don't know if this method is a good approximation or not It made sense to me so i posted here to get feedback.
On the Analytics page the CTD for ZN has maturity of 8/15/2028. That's ~6.95 years isn't it? ZF is 2/28/2026, or ~4.4 years.

The IRR does seem to move around a lot, but I'm still fairly confident that IRR+repo rate represents the financing cost based on the definition of IRR.

https://www.cmegroup.com/education/cour ... s-ctd.html
I think it is more appropriate to use "average duration" instead of time to maturity in this calculation. It is also available in the same page as "Duration"
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

klaus14 wrote: Fri Sep 03, 2021 1:49 pm
skierincolorado wrote: Fri Sep 03, 2021 1:47 pm
klaus14 wrote: Fri Sep 03, 2021 1:40 pm
skierincolorado wrote: Fri Sep 03, 2021 1:33 pm
klaus14 wrote: Fri Sep 03, 2021 12:47 pm

I moved to ZN. It seems to have highest carry.
I used the CTD durations and futures yield curve from Treasury Analytics and seems like ZN has the highest yield per duration.

Method:
- Subtract financing cost from yield to find yield pickup. For ZN: 1.171%-0.1%=1.071%
- Scale its duration to some other contract, for example to ZF: 1.071*4.4412/6.3810= 0.74542%
- Compare this to ZF yield pickup: 0.7621%-0.1%=0.6621
===> ZN > ZF
Similarly ZN > UB, ZN > ZB and ZN > TN
Thanks! I need to pick up a contract in the next few days so this helps. One thing though, I get a duration for ZN of ~6.95 years with the CTD w/ a 8/15/2028 maturity. I also only see 1.115% as the yield. This gives me:

For ZN 1.115-.1 = 1.015 * 4.44/6.95 = .648% scaled to the 5 year
For ZF I get the same .6621%

So I get ZF ahead by a hair.

Second where did you get .1% for the financing cost? If we look at the Implied Repo Rate it is .03% for ZF and .11% for ZN. I think the financing cost is either equal to the IRR, or it's equal to IRR + current repo rate. I think it's the latter. So we need to add the current repo rate (.06) to these amounts. For ZF that would give .09 and for ZN it would be .17. If I'm doing this correctly, it would add to ZFs lead. The carry per unit duration would be nearly .1% higher for ZF at that point:

For ZN 1.115-.17 = .945 * 4.44/6.95 = .603%
For ZF .7621-.09 = .6721%


Of course one shouldn't just try to maximize carry, the market is efficient and can look at this too. The market may be giving higher carry to ZF because it expects the curve to flatten further. The higher carry is the reward for missing out on some of the flattening one might experience with ZN. I think the main thing though is that historically, nothing longer than ZN has performed well.
ZN duration is not 6.95 right? It is 6.381. Maybe you have a typo?
I remember reading somewhere financing cost is 3 month libor. Which is 0.12 according to this. I used 0.1 because it makes easy to do subtractions :)
implied repo there seems to fluctuate wildly so i just used a rule of thumb.

And disclaimer: i don't know if this method is a good approximation or not It made sense to me so i posted here to get feedback.
On the Analytics page the CTD for ZN has maturity of 8/15/2028. That's ~6.95 years isn't it? ZF is 2/28/2026, or ~4.4 years.

The IRR does seem to move around a lot, but I'm still fairly confident that IRR+repo rate represents the financing cost based on the definition of IRR.

https://www.cmegroup.com/education/cour ... s-ctd.html
I think it is more appropriate to use "average duration" instead of time to maturity in this calculation. It is also available in the same page as "Duration"
I don't think it would be, the contract is for the CTD security isn't it? If the contract is for a maturity with 6.95 years expiration that's what we should use. Do you have a link for this average duration sorry I can't find it, what is it an average of?
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by klaus14 »

skierincolorado wrote: Fri Sep 03, 2021 1:54 pm
klaus14 wrote: Fri Sep 03, 2021 1:49 pm
skierincolorado wrote: Fri Sep 03, 2021 1:47 pm
klaus14 wrote: Fri Sep 03, 2021 1:40 pm
skierincolorado wrote: Fri Sep 03, 2021 1:33 pm

Thanks! I need to pick up a contract in the next few days so this helps. One thing though, I get a duration for ZN of ~6.95 years with the CTD w/ a 8/15/2028 maturity. I also only see 1.115% as the yield. This gives me:

For ZN 1.115-.1 = 1.015 * 4.44/6.95 = .648% scaled to the 5 year
For ZF I get the same .6621%

So I get ZF ahead by a hair.

Second where did you get .1% for the financing cost? If we look at the Implied Repo Rate it is .03% for ZF and .11% for ZN. I think the financing cost is either equal to the IRR, or it's equal to IRR + current repo rate. I think it's the latter. So we need to add the current repo rate (.06) to these amounts. For ZF that would give .09 and for ZN it would be .17. If I'm doing this correctly, it would add to ZFs lead. The carry per unit duration would be nearly .1% higher for ZF at that point:

For ZN 1.115-.17 = .945 * 4.44/6.95 = .603%
For ZF .7621-.09 = .6721%


Of course one shouldn't just try to maximize carry, the market is efficient and can look at this too. The market may be giving higher carry to ZF because it expects the curve to flatten further. The higher carry is the reward for missing out on some of the flattening one might experience with ZN. I think the main thing though is that historically, nothing longer than ZN has performed well.
ZN duration is not 6.95 right? It is 6.381. Maybe you have a typo?
I remember reading somewhere financing cost is 3 month libor. Which is 0.12 according to this. I used 0.1 because it makes easy to do subtractions :)
implied repo there seems to fluctuate wildly so i just used a rule of thumb.

And disclaimer: i don't know if this method is a good approximation or not It made sense to me so i posted here to get feedback.
On the Analytics page the CTD for ZN has maturity of 8/15/2028. That's ~6.95 years isn't it? ZF is 2/28/2026, or ~4.4 years.

The IRR does seem to move around a lot, but I'm still fairly confident that IRR+repo rate represents the financing cost based on the definition of IRR.

https://www.cmegroup.com/education/cour ... s-ctd.html
I think it is more appropriate to use "average duration" instead of time to maturity in this calculation. It is also available in the same page as "Duration"
I don't think it would be, the contract is for the CTD security isn't it? If the contract is for a maturity with 6.95 years expiration that's what we should use. Do you have a link for this average duration sorry I can't find it, what is it an average of?
data is available in the same page at the top.
i meant the duration as defined in page 12 here. If you look at the table there you'll see it is different than maturity.
it is known as average duration in ETFs so i used it since it may be familiar. it is different than maturity like here.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
corpgator
Posts: 217
Joined: Wed Jul 22, 2015 12:00 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by corpgator »

Every time I come back to this thread, it gets more and more complicated. Yeesh.
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

klaus14 wrote: Fri Sep 03, 2021 1:58 pm
skierincolorado wrote: Fri Sep 03, 2021 1:54 pm
klaus14 wrote: Fri Sep 03, 2021 1:49 pm
skierincolorado wrote: Fri Sep 03, 2021 1:47 pm
klaus14 wrote: Fri Sep 03, 2021 1:40 pm

ZN duration is not 6.95 right? It is 6.381. Maybe you have a typo?
I remember reading somewhere financing cost is 3 month libor. Which is 0.12 according to this. I used 0.1 because it makes easy to do subtractions :)
implied repo there seems to fluctuate wildly so i just used a rule of thumb.

And disclaimer: i don't know if this method is a good approximation or not It made sense to me so i posted here to get feedback.
On the Analytics page the CTD for ZN has maturity of 8/15/2028. That's ~6.95 years isn't it? ZF is 2/28/2026, or ~4.4 years.

The IRR does seem to move around a lot, but I'm still fairly confident that IRR+repo rate represents the financing cost based on the definition of IRR.

https://www.cmegroup.com/education/cour ... s-ctd.html
I think it is more appropriate to use "average duration" instead of time to maturity in this calculation. It is also available in the same page as "Duration"
I don't think it would be, the contract is for the CTD security isn't it? If the contract is for a maturity with 6.95 years expiration that's what we should use. Do you have a link for this average duration sorry I can't find it, what is it an average of?
data is available in the same page at the top.
i meant the duration as defined in page 12 here. If you look at the table there you'll see it is different than maturity.
it is known as average duration in ETFs so i used it since it may be familiar. it is different than maturity like here.
Yep you're right, it looks like the average duration would be more appropriate. So now we have:

For ZN 1.117-.07 = 1.047 * 4.44/6.38 = .729%
For ZF .7621-.03 = .7321%


Using the definition of IRR from the paper you linked, I think IRR is the financing cost:

The IRR is calculated as the annualized rate of return
associated with the purchase of a security, sale of futures
and delivery of the same in satisfaction of the maturing
futures contract.
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by klaus14 »

skierincolorado wrote: Fri Sep 03, 2021 2:19 pm
klaus14 wrote: Fri Sep 03, 2021 1:58 pm
skierincolorado wrote: Fri Sep 03, 2021 1:54 pm
klaus14 wrote: Fri Sep 03, 2021 1:49 pm
skierincolorado wrote: Fri Sep 03, 2021 1:47 pm

On the Analytics page the CTD for ZN has maturity of 8/15/2028. That's ~6.95 years isn't it? ZF is 2/28/2026, or ~4.4 years.

The IRR does seem to move around a lot, but I'm still fairly confident that IRR+repo rate represents the financing cost based on the definition of IRR.

https://www.cmegroup.com/education/cour ... s-ctd.html
I think it is more appropriate to use "average duration" instead of time to maturity in this calculation. It is also available in the same page as "Duration"
I don't think it would be, the contract is for the CTD security isn't it? If the contract is for a maturity with 6.95 years expiration that's what we should use. Do you have a link for this average duration sorry I can't find it, what is it an average of?
data is available in the same page at the top.
i meant the duration as defined in page 12 here. If you look at the table there you'll see it is different than maturity.
it is known as average duration in ETFs so i used it since it may be familiar. it is different than maturity like here.
Yep you're right, it looks like the average duration would be more appropriate. So now we have:

For ZN 1.117-.07 = 1.047 * 4.44/6.38 = .729%
For ZF .7621-.03 = .7321%


Using the definition of IRR from the paper you linked, I think IRR is the financing cost:

The IRR is calculated as the annualized rate of return
associated with the purchase of a security, sale of futures
and delivery of the same in satisfaction of the maturing
futures contract.
IRR i see are .15 and .08 respectively right now.

and interestingly IRR for TN is negative. so they are paying you to hold the future? It is probably anomaly.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

klaus14 wrote: Fri Sep 03, 2021 2:25 pm
skierincolorado wrote: Fri Sep 03, 2021 2:19 pm
klaus14 wrote: Fri Sep 03, 2021 1:58 pm
skierincolorado wrote: Fri Sep 03, 2021 1:54 pm
klaus14 wrote: Fri Sep 03, 2021 1:49 pm

I think it is more appropriate to use "average duration" instead of time to maturity in this calculation. It is also available in the same page as "Duration"
I don't think it would be, the contract is for the CTD security isn't it? If the contract is for a maturity with 6.95 years expiration that's what we should use. Do you have a link for this average duration sorry I can't find it, what is it an average of?
data is available in the same page at the top.
i meant the duration as defined in page 12 here. If you look at the table there you'll see it is different than maturity.
it is known as average duration in ETFs so i used it since it may be familiar. it is different than maturity like here.
Yep you're right, it looks like the average duration would be more appropriate. So now we have:

For ZN 1.117-.07 = 1.047 * 4.44/6.38 = .729%
For ZF .7621-.03 = .7321%


Using the definition of IRR from the paper you linked, I think IRR is the financing cost:

The IRR is calculated as the annualized rate of return
associated with the purchase of a security, sale of futures
and delivery of the same in satisfaction of the maturing
futures contract.
IRR i see are .15 and .08 respectively right now.

and interestingly IRR for TN is negative. so they are paying you to hold the future? It is probably anomaly.
That could be accurate though. The "optionality" discussed in the paper you linked explains why IRR is often lower than actual repo rates. But I assume this "optionality" that benefits the seller is just as detrimental to the buyer. Thus maybe you're right and it is best to use actual repo rates or LIBOR for the calculation.
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by klaus14 »

skierincolorado wrote: Fri Sep 03, 2021 2:33 pm
klaus14 wrote: Fri Sep 03, 2021 2:25 pm
skierincolorado wrote: Fri Sep 03, 2021 2:19 pm
klaus14 wrote: Fri Sep 03, 2021 1:58 pm
skierincolorado wrote: Fri Sep 03, 2021 1:54 pm

I don't think it would be, the contract is for the CTD security isn't it? If the contract is for a maturity with 6.95 years expiration that's what we should use. Do you have a link for this average duration sorry I can't find it, what is it an average of?
data is available in the same page at the top.
i meant the duration as defined in page 12 here. If you look at the table there you'll see it is different than maturity.
it is known as average duration in ETFs so i used it since it may be familiar. it is different than maturity like here.
Yep you're right, it looks like the average duration would be more appropriate. So now we have:

For ZN 1.117-.07 = 1.047 * 4.44/6.38 = .729%
For ZF .7621-.03 = .7321%


Using the definition of IRR from the paper you linked, I think IRR is the financing cost:

The IRR is calculated as the annualized rate of return
associated with the purchase of a security, sale of futures
and delivery of the same in satisfaction of the maturing
futures contract.
IRR i see are .15 and .08 respectively right now.

and interestingly IRR for TN is negative. so they are paying you to hold the future? It is probably anomaly.
That could be accurate though. The "optionality" discussed in the paper you linked explains why IRR is often lower than actual repo rates. But I assume this "optionality" that benefits the seller is just as detrimental to the buyer. Thus maybe you're right and it is best to use actual repo rates or LIBOR for the calculation.
yes.
sometimes a particular security may be in high demand for whatever reason. this can also cause negative IRR. but these effects should be temporary so shouldn't be relied upon for long term strategy.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

klaus14 wrote: Fri Sep 03, 2021 2:42 pm
skierincolorado wrote: Fri Sep 03, 2021 2:33 pm
klaus14 wrote: Fri Sep 03, 2021 2:25 pm
skierincolorado wrote: Fri Sep 03, 2021 2:19 pm
klaus14 wrote: Fri Sep 03, 2021 1:58 pm

data is available in the same page at the top.
i meant the duration as defined in page 12 here. If you look at the table there you'll see it is different than maturity.
it is known as average duration in ETFs so i used it since it may be familiar. it is different than maturity like here.
Yep you're right, it looks like the average duration would be more appropriate. So now we have:

For ZN 1.117-.07 = 1.047 * 4.44/6.38 = .729%
For ZF .7621-.03 = .7321%


Using the definition of IRR from the paper you linked, I think IRR is the financing cost:

The IRR is calculated as the annualized rate of return
associated with the purchase of a security, sale of futures
and delivery of the same in satisfaction of the maturing
futures contract.
IRR i see are .15 and .08 respectively right now.

and interestingly IRR for TN is negative. so they are paying you to hold the future? It is probably anomaly.
That could be accurate though. The "optionality" discussed in the paper you linked explains why IRR is often lower than actual repo rates. But I assume this "optionality" that benefits the seller is just as detrimental to the buyer. Thus maybe you're right and it is best to use actual repo rates or LIBOR for the calculation.
yes.
sometimes a particular security may be in high demand for whatever reason. this can also cause negative IRR. but these effects should be temporary so shouldn't be relied upon for long term strategy.
I think the "optionality" refers more to potential changes to the CTD that benefit the lender, hence they may be willing to lend at slightly lower rates than LIBOR
User avatar
LTCM
Posts: 412
Joined: Wed Sep 09, 2020 3:58 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by LTCM »

skierincolorado wrote: Fri Sep 03, 2021 2:45 pm I think the "optionality" refers more to potential changes to the CTD that benefit the lender, hence they may be willing to lend at slightly lower rates than LIBOR
Is this what CME mean when they talk about contracts rolling cheap?
55% VUG - 20% VEA - 20% EDV - 5% BNDX
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

LTCM wrote: Fri Sep 03, 2021 2:56 pm
skierincolorado wrote: Fri Sep 03, 2021 2:45 pm I think the "optionality" refers more to potential changes to the CTD that benefit the lender, hence they may be willing to lend at slightly lower rates than LIBOR
Is this what CME mean when they talk about contracts rolling cheap?
I'm not sure.. were getting into the nitty gritty not really necessary for some modest buy and hold (and roll) of futures. If you have a link I could take a guess.
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

klaus14 wrote: Fri Sep 03, 2021 2:25 pm
skierincolorado wrote: Fri Sep 03, 2021 2:19 pm
klaus14 wrote: Fri Sep 03, 2021 1:58 pm
skierincolorado wrote: Fri Sep 03, 2021 1:54 pm
klaus14 wrote: Fri Sep 03, 2021 1:49 pm

I think it is more appropriate to use "average duration" instead of time to maturity in this calculation. It is also available in the same page as "Duration"
I don't think it would be, the contract is for the CTD security isn't it? If the contract is for a maturity with 6.95 years expiration that's what we should use. Do you have a link for this average duration sorry I can't find it, what is it an average of?
data is available in the same page at the top.
i meant the duration as defined in page 12 here. If you look at the table there you'll see it is different than maturity.
it is known as average duration in ETFs so i used it since it may be familiar. it is different than maturity like here.
Yep you're right, it looks like the average duration would be more appropriate. So now we have:

For ZN 1.117-.07 = 1.047 * 4.44/6.38 = .729%
For ZF .7621-.03 = .7321%


Using the definition of IRR from the paper you linked, I think IRR is the financing cost:

The IRR is calculated as the annualized rate of return
associated with the purchase of a security, sale of futures
and delivery of the same in satisfaction of the maturing
futures contract.
IRR i see are .15 and .08 respectively right now.

and interestingly IRR for TN is negative. so they are paying you to hold the future? It is probably anomaly.
OK so final numbers!
For ZN 1.117 -.1 = 1.017 * 4.44/6.38 = .708
For ZF .7621 -.1 = .6621

This I think is the carry relative to the effective duration. So it's basically how much (net) interest you are getting relative to how much risk. For ZB I got ~.45.
User avatar
LTCM
Posts: 412
Joined: Wed Sep 09, 2020 3:58 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by LTCM »

skierincolorado wrote: Fri Sep 03, 2021 3:19 pm
LTCM wrote: Fri Sep 03, 2021 2:56 pm
skierincolorado wrote: Fri Sep 03, 2021 2:45 pm I think the "optionality" refers more to potential changes to the CTD that benefit the lender, hence they may be willing to lend at slightly lower rates than LIBOR
Is this what CME mean when they talk about contracts rolling cheap?
I'm not sure.. were getting into the nitty gritty not really necessary for some modest buy and hold (and roll) of futures. If you have a link I could take a guess.
The implied financing is pretty important with this strategy imo. Its one of the main reasons we're using futures. I do agree its pretty far away from HFEA. HFEA is simple but less efficient. It might be worth having a dedicated futures thread?

https://www.cmegroup.com/trading/equity ... -etfs.html
ctrl-f "Observations on the Futures Roll"

non CME paper here:
https://cdar.berkeley.edu/sites/default ... 022021.pdf
55% VUG - 20% VEA - 20% EDV - 5% BNDX
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

LTCM wrote: Fri Sep 03, 2021 3:46 pm
skierincolorado wrote: Fri Sep 03, 2021 3:19 pm
LTCM wrote: Fri Sep 03, 2021 2:56 pm
skierincolorado wrote: Fri Sep 03, 2021 2:45 pm I think the "optionality" refers more to potential changes to the CTD that benefit the lender, hence they may be willing to lend at slightly lower rates than LIBOR
Is this what CME mean when they talk about contracts rolling cheap?
I'm not sure.. were getting into the nitty gritty not really necessary for some modest buy and hold (and roll) of futures. If you have a link I could take a guess.
The implied financing is pretty important with this strategy imo. Its one of the main reasons we're using futures. I do agree its pretty far away from HFEA. HFEA is simple but less efficient. It might be worth having a dedicated futures thread?

https://www.cmegroup.com/trading/equity ... -etfs.html
ctrl-f "Observations on the Futures Roll"

non CME paper here:
https://cdar.berkeley.edu/sites/default ... 022021.pdf
I guess I just feel pretty confident in efficient markets making futures the cheaper option. You've cut out the middle man. My understanding is LETFs are just using futures and other derivates to obtain leverage, but then charging you a 0.9% fee in addition to passing on the borrowing costs. It's good to be able to check the borrowing cost calculations ourselves, but not totally mandatory IMO. Futures should have similar borrowing costs to LETFs, have much lower fees, and allow for greater leverage of ITT. They also cut out the daily rebalancing, which I don't like because it can produce losses in volatile sideways markets (although it can also enhance returns in vertical markets).

I agree a thread on futures and ITT variations of HFEA would be a good idea, I'll start one.

Oh and the "cheap" vs "rich" they are referring to is relative to LIBOR. They don't really suggest why roll could be cheaper than LIBOR. They do suggest that regulatory tightness on banks was believed to contribute to rich/expensive rolls. This is on S&P futures. The other paper that was posted was talking about cheap rolls on Treasury Futures being related to "optionality" that the lender has about which contract to deliver (or something like that).
Last edited by skierincolorado on Fri Sep 03, 2021 4:13 pm, edited 1 time in total.
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by klaus14 »

Paper about max-carry approach.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by klaus14 »

skierincolorado wrote: Fri Sep 03, 2021 4:04 pm
LTCM wrote: Fri Sep 03, 2021 3:46 pm
skierincolorado wrote: Fri Sep 03, 2021 3:19 pm
LTCM wrote: Fri Sep 03, 2021 2:56 pm
skierincolorado wrote: Fri Sep 03, 2021 2:45 pm I think the "optionality" refers more to potential changes to the CTD that benefit the lender, hence they may be willing to lend at slightly lower rates than LIBOR
Is this what CME mean when they talk about contracts rolling cheap?
I'm not sure.. were getting into the nitty gritty not really necessary for some modest buy and hold (and roll) of futures. If you have a link I could take a guess.
The implied financing is pretty important with this strategy imo. Its one of the main reasons we're using futures. I do agree its pretty far away from HFEA. HFEA is simple but less efficient. It might be worth having a dedicated futures thread?

https://www.cmegroup.com/trading/equity ... -etfs.html
ctrl-f "Observations on the Futures Roll"

non CME paper here:
https://cdar.berkeley.edu/sites/default ... 022021.pdf
I guess I just feel pretty confident in efficient markets making futures the cheaper option. You've cut out the middle man. My understanding is LETFs are just using futures and other derivates to obtain leverage, but then charging you a 0.9% fee in addition to passing on the borrowing costs. It's good to be able to check the borrowing cost calculations ourselves, but not totally mandatory IMO. Futures should have similar borrowing costs to LETFs, have much lower fees, and allow for greater leverage of ITT. They also cut out the daily rebalancing, which I don't like because it can produce losses in volatile sideways markets (although it can also enhance returns in vertical markets).

I agree a thread on futures and ITT variations of HFEA would be a good idea, I'll start one.
Agreed. It is clear futures are superior to ETF (if you don't mind size and can avoid tax issues)
However it doesn't end there. There maybe a reason market assigns more carry to ZN than ZB or UB. If it is mostly a BAB story then it is almost a free lunch.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
User avatar
LTCM
Posts: 412
Joined: Wed Sep 09, 2020 3:58 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by LTCM »

skierincolorado wrote: Fri Sep 03, 2021 4:04 pm I guess I just feel pretty confident in efficient markets making futures the cheaper option. You've cut out the middle man. My understanding is LETFs are just using futures and other derivates to obtain leverage, but then charging you a 0.9% fee in addition to passing on the borrowing costs. It's good to be able to check the borrowing cost calculations ourselves, but not totally mandatory IMO. Futures should have similar borrowing costs to LETFs, have much lower fees, and allow for greater leverage of ITT. They also cut out the daily rebalancing, which I don't like because it can produce losses in volatile sideways markets (although it can also enhance returns in vertical markets).

I agree a thread on futures and ITT variations of HFEA would be a good idea, I'll start one.
I agree futures are going to beat LETF whichever way you slice it. I'm talking about being able to calculate the implied financing on different treasury duration futures to better pick between /ZN /ZF each quarter. Or at least understand why the implied financing might be different for each.
55% VUG - 20% VEA - 20% EDV - 5% BNDX
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

klaus14 wrote: Fri Sep 03, 2021 4:11 pm
skierincolorado wrote: Fri Sep 03, 2021 4:04 pm
LTCM wrote: Fri Sep 03, 2021 3:46 pm
skierincolorado wrote: Fri Sep 03, 2021 3:19 pm
LTCM wrote: Fri Sep 03, 2021 2:56 pm

Is this what CME mean when they talk about contracts rolling cheap?
I'm not sure.. were getting into the nitty gritty not really necessary for some modest buy and hold (and roll) of futures. If you have a link I could take a guess.
The implied financing is pretty important with this strategy imo. Its one of the main reasons we're using futures. I do agree its pretty far away from HFEA. HFEA is simple but less efficient. It might be worth having a dedicated futures thread?

https://www.cmegroup.com/trading/equity ... -etfs.html
ctrl-f "Observations on the Futures Roll"

non CME paper here:
https://cdar.berkeley.edu/sites/default ... 022021.pdf
I guess I just feel pretty confident in efficient markets making futures the cheaper option. You've cut out the middle man. My understanding is LETFs are just using futures and other derivates to obtain leverage, but then charging you a 0.9% fee in addition to passing on the borrowing costs. It's good to be able to check the borrowing cost calculations ourselves, but not totally mandatory IMO. Futures should have similar borrowing costs to LETFs, have much lower fees, and allow for greater leverage of ITT. They also cut out the daily rebalancing, which I don't like because it can produce losses in volatile sideways markets (although it can also enhance returns in vertical markets).

I agree a thread on futures and ITT variations of HFEA would be a good idea, I'll start one.
Agreed. It is clear futures are superior to ETF (if you don't mind size and can avoid tax issues)
However it doesn't end there. There maybe a reason market assigns more carry to ZN than ZB or UB. If it is mostly a BAB story then it is almost a free lunch.
I think it's a free lunch, to quote from the BAB paper.

Since (leverage) constrained investors bid up high-beta assets, high beta is associated with low alpha, as we
find empirically for U.S. equities, 20 international equity markets, Treasury bonds, corporate
bonds, and futures;

Our finding is that the compensation per unit of risk is in fact larger for the
1-year bond than for the 10-year bond. Hence, a portfolio that has a leveraged long
position in 1-year (and other short-term) bonds and a short position in long-term
bonds produces positive returns.


I know the max-carry paper says max-carry beats BAB, but that doesn't explain why some bonds (usually STT or ITT) have higher carry. I think it's the free lunch of BAB.
Last edited by skierincolorado on Fri Sep 03, 2021 4:52 pm, edited 1 time in total.
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

LTCM wrote: Fri Sep 03, 2021 4:15 pm
skierincolorado wrote: Fri Sep 03, 2021 4:04 pm I guess I just feel pretty confident in efficient markets making futures the cheaper option. You've cut out the middle man. My understanding is LETFs are just using futures and other derivates to obtain leverage, but then charging you a 0.9% fee in addition to passing on the borrowing costs. It's good to be able to check the borrowing cost calculations ourselves, but not totally mandatory IMO. Futures should have similar borrowing costs to LETFs, have much lower fees, and allow for greater leverage of ITT. They also cut out the daily rebalancing, which I don't like because it can produce losses in volatile sideways markets (although it can also enhance returns in vertical markets).

I agree a thread on futures and ITT variations of HFEA would be a good idea, I'll start one.
I agree futures are going to beat LETF whichever way you slice it. I'm talking about being able to calculate the implied financing on different treasury duration futures to better pick between /ZN /ZF each quarter. Or at least understand why the implied financing might be different for each.
Oh right, I think Klaus has convinced me the borrowing costs should be the same. The carry is lower for the long-bonds UB and ZB. The differences in IRR on the Treasury Analytics page could be due to differences in the "optionality", but that would lead to differences that affect the buyer as well, so it's a wash. Theoretically, I can't think of a reason the financing would be different on different terms. An efficient market should have all of this priced in. If you are getting lower financing on one contract, it's because you're being hit in another way, like that duration being less desirable for some reason, or optionality on the CTD. And treasury futures are the most liquid market in the world I think. Maybe a strategy like max-carry can help us pick winning durations over time, but holding a mix of ZT/ZF/ZN should be nearly or just as good. I enjoy the complexity to an extent but there's a limit haha.. I don't see myself calculating implied financing costs every time I roll. Maybe I'd calculate the carry since that is simpler.
Last edited by skierincolorado on Fri Sep 03, 2021 5:08 pm, edited 2 times in total.
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

klaus14 wrote: Fri Sep 03, 2021 4:06 pm Paper about max-carry approach.
The very strong results they find for curve carry diversified internationally has me wondering how hard this would be to implement. I also don't understand if they hedged their currency risk or if the returns are in USD.
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by klaus14 »

skierincolorado wrote: Fri Sep 03, 2021 3:34 pm
klaus14 wrote: Fri Sep 03, 2021 2:25 pm
skierincolorado wrote: Fri Sep 03, 2021 2:19 pm
klaus14 wrote: Fri Sep 03, 2021 1:58 pm
skierincolorado wrote: Fri Sep 03, 2021 1:54 pm

I don't think it would be, the contract is for the CTD security isn't it? If the contract is for a maturity with 6.95 years expiration that's what we should use. Do you have a link for this average duration sorry I can't find it, what is it an average of?
data is available in the same page at the top.
i meant the duration as defined in page 12 here. If you look at the table there you'll see it is different than maturity.
it is known as average duration in ETFs so i used it since it may be familiar. it is different than maturity like here.
Yep you're right, it looks like the average duration would be more appropriate. So now we have:

For ZN 1.117-.07 = 1.047 * 4.44/6.38 = .729%
For ZF .7621-.03 = .7321%


Using the definition of IRR from the paper you linked, I think IRR is the financing cost:

The IRR is calculated as the annualized rate of return
associated with the purchase of a security, sale of futures
and delivery of the same in satisfaction of the maturing
futures contract.
IRR i see are .15 and .08 respectively right now.

and interestingly IRR for TN is negative. so they are paying you to hold the future? It is probably anomaly.
OK so final numbers!
For ZN 1.117 -.1 = 1.017 * 4.44/6.38 = .708
For ZF .7621 -.1 = .6621

This I think is the carry relative to the effective duration. So it's basically how much (net) interest you are getting relative to how much risk. For ZB I got ~.45.
i am not sure if financing cost is already factored in futures yield or not. If factored in then simply:
For ZN 1.117 * 4.44/6.38 = .777
vs ZF .7621
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

klaus14 wrote: Fri Sep 03, 2021 5:27 pm
skierincolorado wrote: Fri Sep 03, 2021 3:34 pm
klaus14 wrote: Fri Sep 03, 2021 2:25 pm
skierincolorado wrote: Fri Sep 03, 2021 2:19 pm
klaus14 wrote: Fri Sep 03, 2021 1:58 pm

data is available in the same page at the top.
i meant the duration as defined in page 12 here. If you look at the table there you'll see it is different than maturity.
it is known as average duration in ETFs so i used it since it may be familiar. it is different than maturity like here.
Yep you're right, it looks like the average duration would be more appropriate. So now we have:

For ZN 1.117-.07 = 1.047 * 4.44/6.38 = .729%
For ZF .7621-.03 = .7321%


Using the definition of IRR from the paper you linked, I think IRR is the financing cost:

The IRR is calculated as the annualized rate of return
associated with the purchase of a security, sale of futures
and delivery of the same in satisfaction of the maturing
futures contract.
IRR i see are .15 and .08 respectively right now.

and interestingly IRR for TN is negative. so they are paying you to hold the future? It is probably anomaly.
OK so final numbers!
For ZN 1.117 -.1 = 1.017 * 4.44/6.38 = .708
For ZF .7621 -.1 = .6621

This I think is the carry relative to the effective duration. So it's basically how much (net) interest you are getting relative to how much risk. For ZB I got ~.45.
i am not sure if financing cost is already factored in futures yield or not. If factored in then simply:
For ZN 1.117 * 4.44/6.38 = .777
vs ZF .7621
Pretty sure it doesn't, the screenshot on this page shows a futures yield of 2.2% for the 10y ultra. It looks to be from Sept or Oct 2017. The 3 month rate was near 1% then and the 10 year rate was indeed around 2.2% (not 3.2%). There's no way (assuming this is from Sep/Oct 2017) that the yield after financing was 2.2%. Before financing makes more sense.
https://www.cmegroup.com/tools-informat ... guide.html

Another interesting question is does their calculation of futures yield include roll. I think it does. ZF has a futures yield of .76% right now with a duration and time to maturity both around 4.4 years. Looking at the yield curve, I would expect the yield of a 4.4y bond to be a lot less than .76%... more like .65% or .70%. So maybe it is including the roll? The interest rate on a 4.4 y bond would only be ~.65% but the yield in the first year would be more than that because it would become a 3.4y bond.
investor.was.here
Posts: 88
Joined: Thu Oct 08, 2020 2:52 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by investor.was.here »

Playing around with some stuff... first, add monthly contributions to HFEA. Second, add 15% of margin. I ignored borrow costs here. And... here's what you'd actually have gotten since these ETFs were released.

That's $10k + $1k/mo inflation-adjusted to $300mil in a decade.
telandra
Posts: 5
Joined: Sun May 16, 2021 4:00 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by telandra »

investor.was.here wrote: Fri Sep 03, 2021 6:24 pm Playing around with some stuff... first, add monthly contributions to HFEA. Second, add 15% of margin. I ignored borrow costs here. And... here's what you'd actually have gotten since these ETFs were released.

That's $10k + $1k/mo inflation-adjusted to $300mil in a decade.
HFEA would be 55/45. You are tripling that to 165/135 and then adding even more margin.
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

telandra wrote: Fri Sep 03, 2021 6:37 pm
investor.was.here wrote: Fri Sep 03, 2021 6:24 pm Playing around with some stuff... first, add monthly contributions to HFEA. Second, add 15% of margin. I ignored borrow costs here. And... here's what you'd actually have gotten since these ETFs were released.

That's $10k + $1k/mo inflation-adjusted to $300mil in a decade.
HFEA would be 55/45. You are tripling that to 165/135 and then adding even more margin.
Yep Port 1 is 3x HFEA and Port 2 is like 4x HFEA.. crazy amounts of leverage.. zeros out immediately in the first market dip in 1987:

https://www.portfoliovisualizer.com/bac ... bol7=VFITX


Just for fun, the blue here is the max-leverage you could get since 1985 without zero-ing out... takes you to 6 Bil... red is 'regular' HFEA... this is just for laughs of course, if the market dropped a few percent more in '87 or '08 it would zero you out.

https://www.portfoliovisualizer.com/bac ... bol7=VFITX
investor.was.here
Posts: 88
Joined: Thu Oct 08, 2020 2:52 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by investor.was.here »

Oh, I forgot to change the ticker symbols lol. Woops. So... nay on the 9x leverage?
constructor
Posts: 17
Joined: Fri Sep 03, 2021 6:18 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by constructor »

Lovely to see the renewed futures discussion here.
skierincolorado wrote: Fri Sep 03, 2021 1:33 pm
klaus14 wrote: Fri Sep 03, 2021 12:47 pm
skierincolorado wrote: Thu Sep 02, 2021 4:17 pm
klaus14 wrote: Thu Sep 02, 2021 3:56 pm after reading this carry discussion i replaced my 1 /UB holding with 2 /ZB. /ZB seems to have more carry but also move similar to /UB. Like in the last 6 months yield curve flattened a bit and both /UB and /ZB interest rates declined similar amounts vs /ZN didn't even move. /ZB seems to be sweet spot to have some carry, equity risk reduction and BAB.
ZB (and UB) have done well recently (last 4 or 5 months). There is no denying that. However, over any extended period shorter durations have trounced ZB. The risk-adjusted retrurns of ZB are still quite poor - not as bad as UB, but still quite poor. I'd also point out the correlation of ZB and UB to stocks the last 5 months has been the opposite of what you want - it's been positive. ZB and UB have been going up the same time as the stock market.

ZB is still well within the range of experiencing negative effects from the BAB phenomenon. The higher yield and beta of a 17 year duration bond is appealing to yield chasing investors that are leverage constrained. Hence the popularity of TMF in this very thread! This yield chasing leads to overpricing and grossly inferior risk adjusted returns.

Personally I wouldn't own anything longer than ZN. I'd swap your 2 ZBs for 4 or 5 ZNs, or 6 or 7 ZFs, or some combo of the two. If the yield curve doesn't change at all, you'd nearly double your return from roll & carry alone. Doubling your roll & carry is a nice starting point, since in the long run that's where 100% of the return from bonds comes from. Betting on long-term rates falling forever is a losing proposition.

As the paper I posted a few posts back shows, the equity risk reduction is greatest for 4-7 year durations. You just have to own more.
I moved to ZN. It seems to have highest carry.
I used the CTD durations and futures yield curve from Treasury Analytics and seems like ZN has the highest yield per duration.

Method:
- Subtract financing cost from yield to find yield pickup. For ZN: 1.171%-0.1%=1.071%
- Scale its duration to some other contract, for example to ZF: 1.071*4.4412/6.3810= 0.74542%
- Compare this to ZF yield pickup: 0.7621%-0.1%=0.6621
===> ZN > ZF
Similarly ZN > UB, ZN > ZB and ZN > TN
Thanks! I need to pick up a contract in the next few days so this helps. One thing though, I get a duration for ZN of ~6.95 years with the CTD w/ a 8/15/2028 maturity. I also only see 1.115% as the yield. This gives me:

For ZN 1.115-.1 = 1.015 * 4.44/6.95 = .648% scaled to the 5 year
For ZF I get the same .6621%

So I get ZF ahead by a hair.

Second where did you get .1% for the financing cost? If we look at the Implied Repo Rate it is .03% for ZF and .11% for ZN. I think the financing cost is either equal to the IRR, or it's equal to IRR + current repo rate. I think it's the latter. So we need to add the current repo rate (.06) to these amounts. For ZF that would give .09 and for ZN it would be .17. If I'm doing this correctly, it would add to ZFs lead. The carry per unit duration would be nearly .1% higher for ZF at that point:

For ZN 1.115-.17 = .945 * 4.44/6.95 = .603%
For ZF .7621-.09 = .6721%

I'm not sure if this is correct though because I've never been able to confirm that the financing cost is equal to IRR+rep rate.

Of course one shouldn't just try to maximize carry, the market is efficient and can look at this too. The market may be giving higher carry to ZF because it expects the curve to flatten further. The higher carry is the reward for missing out on some of the flattening one might experience with ZN. I think the main thing though is that historically, nothing longer than ZN has performed well.
A while ago someone mentioned using Eurodollar futures here: https://bogleheads.org/forum/viewtopic. ... 9#p4760749

Any thoughts? Since it's not settled with real bonds does it mean it misses out on the "carry"?
anaparafunds
Posts: 51
Joined: Sun Jul 18, 2021 5:50 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by anaparafunds »

What does 55%/45% mean?

55 shares of UPRO and 45 shares of TMF or 55% dollar amount of UPRO and 45% dollar amount TMF?
User avatar
drumboy256
Posts: 673
Joined: Sat Jun 06, 2020 2:21 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by drumboy256 »

anaparafunds wrote: Sat Sep 04, 2021 6:28 pm What does 55%/45% mean?

55 shares of UPRO and 45 shares of TMF or 55% dollar amount of UPRO and 45% dollar amount TMF?
Percentages, not shares.
Promise is one thing. Fulfilling that promise is quite another. - Sir Alex Ferguson | 20% IVV / 40% IBIT / 20% IXUS / 20% VGLT + chill
anaparafunds
Posts: 51
Joined: Sun Jul 18, 2021 5:50 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by anaparafunds »

drumboy256 wrote: Sat Sep 04, 2021 6:35 pm
anaparafunds wrote: Sat Sep 04, 2021 6:28 pm What does 55%/45% mean?

55 shares of UPRO and 45 shares of TMF or 55% dollar amount of UPRO and 45% dollar amount TMF?
Percentages, not shares.
Thanks.
BayStater
Posts: 97
Joined: Sun Mar 29, 2020 11:57 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by BayStater »

tomphilly wrote: Wed Sep 01, 2021 1:57 pm This week I was recommended an alternative volatility hedge that I had never heard of, and I decided to backtest it - VIRT. It only goes back to 2015 but it holds up in backtests.
VIRT as in the company Virtu Financial? That's what shows on the PV link you sent.

There was an earlier discussion on here about using Direxion's FAS (3x Financial Services). But what's the thesis behind Virtu Financial being a volatility hedge (at least more than any other market maker)? Why would you put 20% of your portfolio in a single company?
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by klaus14 »

skierincolorado wrote: Fri Sep 03, 2021 6:19 pm
klaus14 wrote: Fri Sep 03, 2021 5:27 pm
skierincolorado wrote: Fri Sep 03, 2021 3:34 pm
klaus14 wrote: Fri Sep 03, 2021 2:25 pm
skierincolorado wrote: Fri Sep 03, 2021 2:19 pm

Yep you're right, it looks like the average duration would be more appropriate. So now we have:

For ZN 1.117-.07 = 1.047 * 4.44/6.38 = .729%
For ZF .7621-.03 = .7321%


Using the definition of IRR from the paper you linked, I think IRR is the financing cost:

The IRR is calculated as the annualized rate of return
associated with the purchase of a security, sale of futures
and delivery of the same in satisfaction of the maturing
futures contract.
IRR i see are .15 and .08 respectively right now.

and interestingly IRR for TN is negative. so they are paying you to hold the future? It is probably anomaly.
OK so final numbers!
For ZN 1.117 -.1 = 1.017 * 4.44/6.38 = .708
For ZF .7621 -.1 = .6621

This I think is the carry relative to the effective duration. So it's basically how much (net) interest you are getting relative to how much risk. For ZB I got ~.45.
i am not sure if financing cost is already factored in futures yield or not. If factored in then simply:
For ZN 1.117 * 4.44/6.38 = .777
vs ZF .7621
Pretty sure it doesn't, the screenshot on this page shows a futures yield of 2.2% for the 10y ultra. It looks to be from Sept or Oct 2017. The 3 month rate was near 1% then and the 10 year rate was indeed around 2.2% (not 3.2%). There's no way (assuming this is from Sep/Oct 2017) that the yield after financing was 2.2%. Before financing makes more sense.
https://www.cmegroup.com/tools-informat ... guide.html

Another interesting question is does their calculation of futures yield include roll. I think it does. ZF has a futures yield of .76% right now with a duration and time to maturity both around 4.4 years. Looking at the yield curve, I would expect the yield of a 4.4y bond to be a lot less than .76%... more like .65% or .70%. So maybe it is including the roll? The interest rate on a 4.4 y bond would only be ~.65% but the yield in the first year would be more than that because it would become a 3.4y bond.
"Futures Yield" seems to be different than "yield to maturity"
As of now ZF CTD Futures Yield is 0.7621% but yield to maturity is 0.7085%

So spot cash bond is more expensive than future bond. My totally uninformed speculation is that it is because of this:
In addition to paying the (negotiated) price of the couponbearing security, the buyer also typically compensates
the seller for any interest accrued between the last semiannual coupon payment date and the settlement date of
the security.

E.g., it is October 10, 2017. You purchase $1 million face
value of the 2-1/4% Treasury security maturing in August 2027
(a ten-year note) for a price of 99-01 ($990,312.50) to yield
2.36%, for settlement on the next day, October 11, 2017.

In addition to the price of the security, you must further
compensate the seller for interest of $3,485.05 accrued
during the 57 days between the original issue date of August
15, 2017 and the settlement date of October 11, 2017
(Source)

--> Futures price should reflect this, and should be cheaper than otherwise because it excludes this payment.

This implies "yield to maturity" is the real yield you are getting. And that is before financing cost.
Also see this answer
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
Mickelous
Posts: 198
Joined: Mon Apr 20, 2020 11:24 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Mickelous »

Aren't futures similar to options and that they will perform terribly in a crash versus daily rebalancing?
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

klaus14 wrote: Mon Sep 06, 2021 5:18 am
skierincolorado wrote: Fri Sep 03, 2021 6:19 pm
klaus14 wrote: Fri Sep 03, 2021 5:27 pm
skierincolorado wrote: Fri Sep 03, 2021 3:34 pm
klaus14 wrote: Fri Sep 03, 2021 2:25 pm

IRR i see are .15 and .08 respectively right now.

and interestingly IRR for TN is negative. so they are paying you to hold the future? It is probably anomaly.
OK so final numbers!
For ZN 1.117 -.1 = 1.017 * 4.44/6.38 = .708
For ZF .7621 -.1 = .6621

This I think is the carry relative to the effective duration. So it's basically how much (net) interest you are getting relative to how much risk. For ZB I got ~.45.
i am not sure if financing cost is already factored in futures yield or not. If factored in then simply:
For ZN 1.117 * 4.44/6.38 = .777
vs ZF .7621
Pretty sure it doesn't, the screenshot on this page shows a futures yield of 2.2% for the 10y ultra. It looks to be from Sept or Oct 2017. The 3 month rate was near 1% then and the 10 year rate was indeed around 2.2% (not 3.2%). There's no way (assuming this is from Sep/Oct 2017) that the yield after financing was 2.2%. Before financing makes more sense.
https://www.cmegroup.com/tools-informat ... guide.html

Another interesting question is does their calculation of futures yield include roll. I think it does. ZF has a futures yield of .76% right now with a duration and time to maturity both around 4.4 years. Looking at the yield curve, I would expect the yield of a 4.4y bond to be a lot less than .76%... more like .65% or .70%. So maybe it is including the roll? The interest rate on a 4.4 y bond would only be ~.65% but the yield in the first year would be more than that because it would become a 3.4y bond.
"Futures Yield" seems to be different than "yield to maturity"
As of now ZF CTD Futures Yield is 0.7621% but yield to maturity is 0.7085%

So spot cash bond is more expensive than future bond. My totally uninformed speculation is that it is because of this:
In addition to paying the (negotiated) price of the couponbearing security, the buyer also typically compensates
the seller for any interest accrued between the last semiannual coupon payment date and the settlement date of
the security.

E.g., it is October 10, 2017. You purchase $1 million face
value of the 2-1/4% Treasury security maturing in August 2027
(a ten-year note) for a price of 99-01 ($990,312.50) to yield
2.36%, for settlement on the next day, October 11, 2017.

In addition to the price of the security, you must further
compensate the seller for interest of $3,485.05 accrued
during the 57 days between the original issue date of August
15, 2017 and the settlement date of October 11, 2017
(Source)

--> Futures price should reflect this, and should be cheaper than otherwise because it excludes this payment.

This implies "yield to maturity" is the real yield you are getting. And that is before financing cost.
Also see this answer
That could be part of it, but I don't think yield to maturity is right either. If the YTM is 0.70% that would mean if you held to maturity the yield would be 0.70%, but with a futures contract we are holding for 3 months at most. The yield curve is steeper at 5 years than at 1 year, so we should get more yield than 0.70%. Likewise, holding an actual bond with 5 years to maturity and YTM of 0.70%, we would get more than 0.70% in the first couple years of holding, and less than 0.70% in the last couple years (assuming no change in the yield curve). I believe this phenomenon is referred to as "roll."

It's also possible that the *price* of the contract includes the interest payment you reference, but the 'futures yield' is calculated taking this into account. But because of the "roll" the futures yield is higher than YTM. In fact,it wouldn't make much sense to calculate futures yield without taking into account the known interest payment.
Hfearless
Posts: 135
Joined: Sat Aug 14, 2021 9:53 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Hfearless »

Klaus and Skier:

Could you please consider being a bit more careful with quoting posts? It’s hard to read short replies to very long quotes.

In phpBB, a convenient way of quoting people is to first click Post Reply without quoting anything, then to use the Topic Review panel below the message composition box to select the relevant text and then to press the Quote button within the post containing that text. This only quotes the selected text, and in addition to being concise, also makes it possible to quote different people in a single reply.
ChinchillaWhiplash
Posts: 1574
Joined: Sat Jan 20, 2018 4:40 pm
Location: Land of Hypoxia

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by ChinchillaWhiplash »

Anyone ever figure out any good TLH partners for UPRO and TMF? Tried to search through the threads but it is too much. :shock:
Hfearless
Posts: 135
Joined: Sat Aug 14, 2021 9:53 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Hfearless »

Tax loss harvesting? TBH, this was mentioned multiple times. The short answer is UPRO↔SPXL, and margin loan/box spread leverage for the Treasurys—but given that in days of prosperity you expect UPRO/SPXL to grow and the Treasurys to lose money, you might as well use futures for the Treasurys and incur your losses that way.
Z33
Posts: 12
Joined: Fri Jun 11, 2021 11:21 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Z33 »

skierincolorado wrote: Fri Sep 03, 2021 10:17 am
DMoogle wrote: Fri Sep 03, 2021 10:01 am
skierincolorado wrote: Thu Sep 02, 2021 3:29 pm
DMoogle wrote: Thu Sep 02, 2021 3:13 pm I think IB does actually count stock holdings as collateral to some extent. Reason being I purchased some futures a few months ago while having a negative balance (margin) in my account, and I didn't run into any issues.

I do have a portfolio margin account, not sure if that matters.
Yeah that's in a taxable account with margin. The cash collateral requirement for a futures contract is just deducted from the cash balance on the account. The cash balance can go substantially negative depending on what else is held in the account.
Oh wow that's way simpler than I was expecting. I assume interest is charged on that difference? So say someone has $10k cash balance, but needs $15k in collateral - the remaining $5k would have interest charged, even though it's not truly a negative cash balance.

skier you've been writing some of the best, most informative posts on here recently, thanks a lot for that. I appreciate the breakdown + reference on implied financing and futures portfolio construction in particular.

You've mentioned before using box spreads in taxable - could you give an overview of what that would look like? I majored in finance in college 10 years ago, but never actually touched options personally. Right now I'm just using margin + NTSX + VXUS in my taxable (plus some individual stocks from before I knew any better).
I believe that's how it works I'll check my account.

Box spreads are awesome. There's some horror stories of people who took absurd leverage or implemented them wrong. One guy used SPY instead of SPX and one leg of the box was called early and blew up his position. European options on SPX can't be called early. Box spreads give us access to commercial lending markets at market short-term interest rates near LIBOR without a middle man, which I find to just be truly awesome. Writing a box is a bit tricky to learn, but other than that it's almost simpler than futures because it's exactly the same as a margin loan from your broker.

Here's a thread on them: viewtopic.php?t=344667

Basically if you write a call for 4000, buy a call for 4500, sell a put for 4500, and buy a put for 4000, you have yourself a 50,000 risk free loan with 0.6% financing (currently). The four options need to be written at the same time as part of the same trade. Some other caveats in the links. You basically refinance your broker margin at this lower rate. So if your IBKR balance is -30,000 you would write a 30,000 box spread, and the cash balance would go back to $0.

This guy explains it partially. He used his very large box spread to borrow at near-LIBOR rates and withdraw the cash and put the money in a CD. Pretty smart, but I would rather borrow less and invest in stocks.

https://www.lesswrong.com/posts/8NSKMMD ... tly-faster

This has some pointers too

https://www.reddit.com/r/investing/comm ... cheap_085/
Amazing sharing! Thank you. Been reading about box spreads but not sure how to execute it. Will study some more.

I read the separate IBKR thread on this but wasn't conclusive on the instrument in terms of eating up buying power of the portfolio?
Cheers
anaparafunds
Posts: 51
Joined: Sun Jul 18, 2021 5:50 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by anaparafunds »

Hfearless wrote: Mon Sep 06, 2021 2:39 pm Tax loss harvesting? TBH, this was mentioned multiple times. The short answer is UPRO↔SPXL, and margin loan/box spread leverage for the Treasurys—but given that in days of prosperity you expect UPRO/SPXL to grow and the Treasurys to lose money, you might as well use futures for the Treasurys and incur your losses that way.
They look very similar. Where does IRS draw the line?

Thanks.
anaparafunds
Posts: 51
Joined: Sun Jul 18, 2021 5:50 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by anaparafunds »

When does Portfolio Visualizer Rebalance when it's set to "Rebalance Quarterly"?

What month and date?
User avatar
firebirdparts
Posts: 4411
Joined: Thu Jun 13, 2019 4:21 pm
Location: Southern Appalachia

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by firebirdparts »

investor.was.here wrote: Fri Sep 03, 2021 6:24 pm That's $10k + $1k/mo inflation-adjusted to $300mil in a decade.
if you had a time machine, 100X leverage would be okay. You just want to know when to do it, but we don't.
Last edited by firebirdparts on Tue Sep 07, 2021 9:22 am, edited 1 time in total.
This time is the same
User avatar
firebirdparts
Posts: 4411
Joined: Thu Jun 13, 2019 4:21 pm
Location: Southern Appalachia

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by firebirdparts »

BayStater wrote: Sun Sep 05, 2021 9:36 pm VIRT as in the company Virtu Financial? That's what shows on the PV link you sent.
That's a bold strategy, cotton, let's see how it works out for him.

I think it's interesting. As I started to even understand it, I saw that there are some articles already about how their earnings might be trending down because volatility is low. So it's interesting.
This time is the same
TurtleVA
Posts: 1
Joined: Tue Sep 07, 2021 11:34 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by TurtleVA »

OP states the biggest risk to this strategy is positive correlation between treasuries and s&p500, but how about an extended drawdown forcing closure of one or both of UPRO and TMF? Is that a realistic scenario?
Last edited by TurtleVA on Tue Sep 07, 2021 11:37 am, edited 1 time in total.
User avatar
firebirdparts
Posts: 4411
Joined: Thu Jun 13, 2019 4:21 pm
Location: Southern Appalachia

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by firebirdparts »

It's been asked before, of course. The secret to TMF and UPRO is that there is somebody "out there" willing to be the counterparty. So I think there's a greater risk of closure if the fund does well. They are vacuuming up all that money from "other people" and at some point the other people may decide they don't want to do that anymore. As long as they find counterparties every day, it doesn't really go negative. If they have trouble doing this, what I see is they would quickly close the fund. that's just one ignorant person's opinion.

If the fund legitimately loses a lot of money, they'd have to pay the counterparties. If it closed due to tiny AUM, I just assume investors would be entitled to that. I mean it could go to zero because that's what you deserve.
Last edited by firebirdparts on Tue Sep 07, 2021 12:39 pm, edited 1 time in total.
This time is the same
telandra
Posts: 5
Joined: Sun May 16, 2021 4:00 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by telandra »

TurtleVA wrote: Tue Sep 07, 2021 11:36 am OP states the biggest risk to this strategy is positive correlation between treasuries and s&p500, but how about an extended drawdown forcing closure of one or both of UPRO and TMF? Is that a realistic scenario?
Long as the correlation is not positive, the strategy should work fine. Even if say UPRO closed, you'd have had 45% invested in TMF, which should have gone up in value.

If both closed down, that would mean they were positively correlated.
Ramjet
Posts: 1464
Joined: Thu Feb 06, 2020 10:45 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Ramjet »

TurtleVA wrote: Tue Sep 07, 2021 11:36 am OP states the biggest risk to this strategy is positive correlation between treasuries and s&p500, but how about an extended drawdown forcing closure of one or both of UPRO and TMF? Is that a realistic scenario?
UPRO/TMF rebalanced quarterly would have survived The Great Financial Crisis, one of only two 50% drawdowns in U.S. history
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by skierincolorado »

firebirdparts wrote: Tue Sep 07, 2021 11:56 am It's been asked before, of course. The secret to TMF and UPRO is that there is somebody "out there" willing to be the counterparty. So I think there's a greater risk of closure if the fund does well. They are vacuuming up all that money from "other people" and at some point the other people may decide they don't want to do that anymore. As long as they find counterparties every day, it doesn't really go negative. If they have trouble doing this, what I see is they would quickly close the fund. that's just one ignorant person's opinion.

If the fund legitimately loses a lot of money, they'd have to pay the counterparties. If it closed due to tiny AUM, I just assume investors would be entitled to that. I mean it could go to zero because that's what you deserve.
There may be other ways they obtain leverage, but they also can obtain leverage just by borrowing money. In that case, the counter party isn’t betting stocks or bonds go down. The counter party is just loaning money. They make money either way.
Ramjet
Posts: 1464
Joined: Thu Feb 06, 2020 10:45 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Ramjet »

firebirdparts wrote: Tue Sep 07, 2021 11:56 am It's been asked before, of course. The secret to TMF and UPRO is that there is somebody "out there" willing to be the counterparty. So I think there's a greater risk of closure if the fund does well. They are vacuuming up all that money from "other people" and at some point the other people may decide they don't want to do that anymore. As long as they find counterparties every day, it doesn't really go negative. If they have trouble doing this, what I see is they would quickly close the fund. that's just one ignorant person's opinion.

If the fund legitimately loses a lot of money, they'd have to pay the counterparties. If it closed due to tiny AUM, I just assume investors would be entitled to that. I mean it could go to zero because that's what you deserve.
Not UPRO/TMF but PSLDX was around for the great financial crisis and survived. Same counterparty concerns during one of the worst economic events in history
Z33
Posts: 12
Joined: Fri Jun 11, 2021 11:21 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Z33 »

Barrons starting to talk about stagflation....#1 nemesis of this strategy.

Any thoughts gents?

Cheers
Post Reply