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

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
CletusCaddy
Posts: 1577
Joined: Sun Sep 12, 2021 4:23 am

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

Post by CletusCaddy »

DMoogle wrote: Tue Jan 17, 2023 8:35 am I know it's a little speculative, but a US default would be absolutely devastating to this - to the extent it could completely wipe out who's pursuing this strategy.

With the debt ceiling deadline approaching, we're in a period where the risk of a US default is non-zero. What do people think about liquidating their treasury futures until it's resolved?
Wouldn’t a Treasury default mean sharply higher interest rates and therefore a huge tailwind to all Treasuries?
DMoogle
Posts: 502
Joined: Sat Oct 31, 2020 10:24 pm

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

Post by DMoogle »

CletusCaddy wrote: Tue Jan 17, 2023 10:22 pm
DMoogle wrote: Tue Jan 17, 2023 8:35 am I know it's a little speculative, but a US default would be absolutely devastating to this - to the extent it could completely wipe out who's pursuing this strategy.

With the debt ceiling deadline approaching, we're in a period where the risk of a US default is non-zero. What do people think about liquidating their treasury futures until it's resolved?
Wouldn’t a Treasury default mean sharply higher interest rates and therefore a huge tailwind to all Treasuries?
Er... the opposite. Rising interest rates are terrible for people holding bonds locked into lower rates, as can be seen in the past year. More than that though - if the lendee doesn't pay what they owe (defaults), then the bonds become worthless, right?

If the bonds associated with my futures become worthless, then I'll be ruined...
parval
Posts: 141
Joined: Tue Oct 22, 2019 9:23 pm

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

Post by parval »

DMoogle wrote: Wed Jan 18, 2023 3:17 pm
CletusCaddy wrote: Tue Jan 17, 2023 10:22 pm
DMoogle wrote: Tue Jan 17, 2023 8:35 am I know it's a little speculative, but a US default would be absolutely devastating to this - to the extent it could completely wipe out who's pursuing this strategy.

With the debt ceiling deadline approaching, we're in a period where the risk of a US default is non-zero. What do people think about liquidating their treasury futures until it's resolved?
Wouldn’t a Treasury default mean sharply higher interest rates and therefore a huge tailwind to all Treasuries?
Er... the opposite. Rising interest rates are terrible for people holding bonds locked into lower rates, as can be seen in the past year. More than that though - if the lendee doesn't pay what they owe (defaults), then the bonds become worthless, right?

If the bonds associated with my futures become worthless, then I'll be ruined...
Apparently CDS's are at an all time high, but you can always grab some to hedge

http://www.worldgovernmentbonds.com/cds ... s/5-years/
AlohaBill
Posts: 395
Joined: Thu Jan 24, 2008 12:20 pm
Location: California

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

Post by AlohaBill »

Just an added note. Don’t try this adventure if:
You have credit card debt
You have less than $4,000 to invest
You need money for your kids’ college education (or to buy a house)
You have a salary correlated with the market
You worry about losing money
You aren’t knowledgeable about margin debt or LEAPS

Schiller says this is only for retirement savings
Otherwise if you’re young and rich have at it.
impatientInv
Posts: 286
Joined: Fri Sep 03, 2021 1:26 pm

Re: SOFR futures

Post by impatientInv »

comeinvest wrote: Wed Nov 30, 2022 12:49 am
It is now official: I made the move to SOFR futures strips in lieu of the STT part of my interest rate / duration / term risk allocation. Here are examples of my futures portfolios in different accounts where I implement mHFEA - one 8-quarter strip, one 9-quarter strip, and a subsampled 4-position 2-year strip in my smaller IRA. My current target is 1/3 of treasuries exposure in STT, 1/3 in /ZF, and 1/3 in /ZN, measured in terms of duration exposure. First I thought I would hate the many line items, but now I find it so clean and easy to monitor, so many less transactions than with treasury futures, and hundreds of $ per year saved in transaction cost including bid/ask spreads. Each contract has a relatively manageable DV01 of exactly $25 at all times, not constantly changing duration exposures that you need to look up somewhere like with treasury futures. Want to adjust or rebalance your exposure? No problem, make the strip a little longer or a little shorter. Want to overweigh certain maturity ranges (e.g. STT)? No problem, add some SOFR futures in that area. Have a smaller account, or don't want as many line items? -> Sample the exposure in the maturity range that you desire - no problem, as the individual contracts that are "adjacent" to each other are highly correlated, because of the nature of the yield curve. (The P&L in the pics differ a little more, as some of the contracts I just bought today.) Nothing to adjust or to rebalance? -> Kick back, enjoy life and watch your SOFR futures roll down the yield curve for years to come. No worrying about anomalies like whether some quarterly roll is rich or cheap. Kiss goodbye the hassle with roll periods "honey I'm late for dinner I need to roll my futures, and b.t.w. ... can we postpone the Italy vacation for a week to after the roll period?" Instead, feed another contract into the sequence occasionally, on your own schedule and at your convenience.

I hope the followers of this thread find it interesting. I'm thinking of migrating the ITT exposure too, combining all former treasury maturities into a suitably designed sequence of SOFR futures. I feel that SOFR futures and leveraged buy-and-hold investors are a match made in heaven.
Thanks for sharing the info. Very helpful. How does one ZF future compare with say June 2026/2027 SOFR future? How is the value related, How would you replace 1 ZF or 2 ZF future with?

Are you concerned with temporary losses during a possible debt ceiling? If the interest rates spike temporarily what are the consequences?

Another question about SOFR gains/losses in a taxable account - if I hold June 2027 SOFR for > 1 year, will the gains/losses be calculated end of the year or when I sell? Will the taxes be like other futures 60-40 LTCG/ST even if held.for >1 year?
No individual stocks.
User avatar
typical.investor
Posts: 4032
Joined: Mon Jun 11, 2018 3:17 am

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

Post by typical.investor »

comeinvest wrote: Wed Jan 11, 2023 10:43 am
Kbg wrote: Wed Jan 11, 2023 9:22 am
comeinvest wrote: Tue Jan 10, 2023 4:29 am
Kbg wrote: Sun Jan 08, 2023 8:29 pm
comeinvest wrote: Sat Jan 07, 2023 6:58 pm

I think most major foreign equity index futures are treated as 1256 contracts. But it doesn't really matter for most, because using equity index futures in taxable account is not a good idea because of the bad tax treatment.
To be more precise with my wording...futures offered on foreign exchanges purchased by US investors are not given 1256 tax treatment. They are treated at securities for US taxation purposes.
I think most major foreign equity index futures are actually treated as 1256 contracts.
You assume that at your tax audit peril. It’s complex and the 6781 form instructions flag the issue. The foreign exchange has to make an application to the relevant US regulators and be approved for the legal status that enables US citizens to claim 1256. And many do not. As a minimum, one needs to do the research on whatever foreign contracts they are trading.
I don't have time right now to look it up again, but I did exactly that research. The major index futures on Eurex and in Osaka applied for approval. I don't care because I don't use equity index futures in taxable accounts. Let us know when you find the list.
Why wouldn’t a US person use NYSE Liffe? With MSCI EAFE and MSCI EM futures, what more do you need?
parval
Posts: 141
Joined: Tue Oct 22, 2019 9:23 pm

Re: SOFR futures

Post by parval »

comeinvest wrote: Wed Nov 30, 2022 12:49 am
comeinvest wrote: Sat Nov 26, 2022 7:41 pm So I tried to explore the possibility of using SOFR futures strips instead of STT futures, as has been proposed a few times before in this thread and in the HFEA thread, and in the Newedge paper https://coexpartnersaig.files.wordpress ... -carry.pdf . (SOFR replaces LIBOR, although there are a few conceptual differences.) I think the idea is to have an exposure to the average of the instantaneous forward interest rates between now and 2 years from now, and to the term premia of the term structure of interest rate estimates, which in theory, because of non-arbitrage and expectations hypothesis, should be very close in risk and performance to buying an equivalent number (i.e. duration matched) of 2y (/ZT) treasury futures, but with less transactions and related cost (commissions and bid/ask spread every quarter), and possibly less exposure to short-term market anomalies and fluctuations of implied repo rates. Also refer to this post viewtopic.php?p=6880351#p6880351 and the cost comparison tool https://www.cmegroup.com/tools-informat ... lysis.html .

...
It is now official: I made the move to SOFR futures strips in lieu of the STT part of my interest rate / duration / term risk allocation. Here are examples of my futures portfolios in different accounts where I implement mHFEA - one 8-quarter strip, one 9-quarter strip, and a subsampled 4-position 2-year strip in my smaller IRA. My current target is 1/3 of treasuries exposure in STT, 1/3 in /ZF, and 1/3 in /ZN, measured in terms of duration exposure. First I thought I would hate the many line items, but now I find it so clean and easy to monitor, so many less transactions than with treasury futures, and hundreds of $ per year saved in transaction cost including bid/ask spreads. Each contract has a relatively manageable DV01 of exactly $25 at all times, not constantly changing duration exposures that you need to look up somewhere like with treasury futures. Want to adjust or rebalance your exposure? No problem, make the strip a little longer or a little shorter. Want to overweigh certain maturity ranges (e.g. STT)? No problem, add some SOFR futures in that area. Have a smaller account, or don't want as many line items? -> Sample the exposure in the maturity range that you desire - no problem, as the individual contracts that are "adjacent" to each other are highly correlated, because of the nature of the yield curve. (The P&L in the pics differ a little more, as some of the contracts I just bought today.) Nothing to adjust or to rebalance? -> Kick back, enjoy life and watch your SOFR futures roll down the yield curve for years to come. No worrying about anomalies like whether some quarterly roll is rich or cheap. Kiss goodbye the hassle with roll periods "honey I'm late for dinner I need to roll my futures, and b.t.w. ... can we postpone the Italy vacation for a week to after the roll period?" Instead, feed another contract into the sequence occasionally, on your own schedule and at your convenience.

I hope the followers of this thread find it interesting. I'm thinking of migrating the ITT exposure too, combining all former treasury maturities into a suitably designed sequence of SOFR futures. I feel that SOFR futures and leveraged buy-and-hold investors are a match made in heaven.

Image

Image

Image
Any thoughts on migrating ZF/ZN? If I have 2 of each, can I roughly replace w/ 4 SQH30?
comeinvest
Posts: 1780
Joined: Mon Mar 12, 2012 6:57 pm

Re: SOFR futures

Post by comeinvest »

parval wrote: Sun Jan 22, 2023 9:27 pm Any thoughts on migrating ZF/ZN? If I have 2 of each, can I roughly replace w/ 4 SQH30?
I'm not sure what exactly you are asking; but if you are asking what one SOFR futures contract will do for you, then perhaps you first need to study a little more or read again my related posts.
Last edited by comeinvest on Sun Jan 22, 2023 11:18 pm, edited 2 times in total.
comeinvest
Posts: 1780
Joined: Mon Mar 12, 2012 6:57 pm

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

Post by comeinvest »

typical.investor wrote: Sun Jan 22, 2023 9:10 pm
comeinvest wrote: Wed Jan 11, 2023 10:43 am
Kbg wrote: Wed Jan 11, 2023 9:22 am
comeinvest wrote: Tue Jan 10, 2023 4:29 am
Kbg wrote: Sun Jan 08, 2023 8:29 pm

To be more precise with my wording...futures offered on foreign exchanges purchased by US investors are not given 1256 tax treatment. They are treated at securities for US taxation purposes.
I think most major foreign equity index futures are actually treated as 1256 contracts.
You assume that at your tax audit peril. It’s complex and the 6781 form instructions flag the issue. The foreign exchange has to make an application to the relevant US regulators and be approved for the legal status that enables US citizens to claim 1256. And many do not. As a minimum, one needs to do the research on whatever foreign contracts they are trading.
I don't have time right now to look it up again, but I did exactly that research. The major index futures on Eurex and in Osaka applied for approval. I don't care because I don't use equity index futures in taxable accounts. Let us know when you find the list.
Why wouldn’t a US person use NYSE Liffe? With MSCI EAFE and MSCI EM futures, what more do you need?
According to my research, the MSCI EAFE and MSCI EM futures have relatively high implied financing cost relative to the risk-free rate of the currency they are denominated in (USD), in comparison to the futures on EUREX and the Nikkei futures. My results are not yet very conclusive, but I posted some preliminary comparisons of the implied financing costs a few pages up in this thread viewtopic.php?p=6987262#p6987262 . I'll do some more backtests once I have time.
km91
Posts: 583
Joined: Wed Oct 13, 2021 12:32 pm

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

Post by km91 »

I asked this in one of the other HF threads but didn't get an answer. Re-posting here to see if I have better luck.

How bad is the tax/performance drag of using equity futures in a taxable account? I want to leverage a portfolio that has sizeable allocations to TIPS, SCV funds, and trend following funds, along with US TSM and Treasury exposures. I've seen it mentioned that the tax treatment is quite unfavorable vs buying the underlying outright on margin and I don't have any reason to doubt this. The tax rate on futures amounts to paying out ~20% of the yearly gains as taxes plus the opportunity cost of not being able to defer the taxes and let gains compound. I don't really like carrying such a large margin balance or using LEFTs but maybe these are my only options? Maybe some degree of tax loss harvesting amongst could offset the capital gains on the futures
comeinvest
Posts: 1780
Joined: Mon Mar 12, 2012 6:57 pm

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

Post by comeinvest »

km91 wrote: Mon Jan 23, 2023 6:37 pm I asked this in one of the other HF threads but didn't get an answer. Re-posting here to see if I have better luck.

How bad is the tax/performance drag of using equity futures in a taxable account? I want to leverage a portfolio that has sizeable allocations to TIPS, SCV funds, and trend following funds, along with US TSM and Treasury exposures. I've seen it mentioned that the tax treatment is quite unfavorable vs buying the underlying outright on margin and I don't have any reason to doubt this. The tax rate on futures amounts to paying out ~20% of the yearly gains as taxes plus the opportunity cost of not being able to defer the taxes and let gains compound. I don't really like carrying such a large margin balance or using LEFTs but maybe these are my only options? Maybe some degree of tax loss harvesting amongst could offset the capital gains on the futures
I think skier pointed out that with margin you pay ca. 1% above the rate that you can get with SPX box spreads or equity index futures, which might very well eat your entire excess return from leverage, not to mention it might not be tax deductible. Above $100k it's only ca. 0.5% above at IB. With equity index futures both your returns and your financing cost are taxable / tax deductible, as they are all embedded in the futures contracts and net each other. SPX box spreads can be better than futures, but in the long run you need to make sure that you have net gains (for example from treasury futures, or gains from some of your other strategies if they involve rebalancing and generate capital gains) to net the capital losses from the options against. The best approach depends on your specific tax situation and investment horizon; some people can itemize their paid investment interest on the tax return, while others can't. ETFs or stocks leveraged with box spreads, and treasury futures, can be a good way to leverage your portfolio in taxable accounts. This is how I do it. This is because treasury futures might have slightly lower financing cost than equity index futures, and on the other hand treasuries have more limited capital gains deferral possibilities in comparison with equities, which means the immediate taxable marked-to-market income from the treasury futures doesn't hurt that much. This was all discussed high up in this and the related HFEA threads. Unfortunately, with this method the annual income will probably fluctuate wildly and be rather unpredictable because of the fluctuating returns from the treasury futures, which can make estimated tax payments a bit difficult, and can catapult you into higher tax brackets in some years (footnote 1). On the other hand, if the term premium disappears and treasury futures never generate positive returns and recoup the losses from 2022, then the capital losses from the box spreads would not be deductible, but you can carry them forward until you have gains to net them against; this would be suboptimal, as you pay taxes on the dividends from the ETFs and stocks on an annual basis, but the deduction of your financing cost would be deferred, which reduces your capital efficiency.

(footnote 1) Assume you have a 1M taxable portfolio with a 250% ITT allocation (based on 5-year ITT duration), where the ITT are implemented with treasury or SOFR futures. Let's say prevailing ITT interest rates are 3.5%, then comes a macroeconomic event that induces a recession and a stock market crash, and subsequently ITT rates sink to 1.5%. If you have no prior deferred losses, then this would translate to an up to ca. $250k immediately taxable marked-to-market income from the treasury or interest rate futures, which is more than many of us typically earn in a year from our jobs. Ouch!

My financial life definitely became more complex since I started mHFEA.

Would you mind sharing your asset allocation? What trend following funds do you use?
sharukh
Posts: 571
Joined: Mon Jun 20, 2016 10:19 am

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

Post by sharukh »

comeinvest wrote: Sun Jan 22, 2023 11:01 pm
typical.investor wrote: Sun Jan 22, 2023 9:10 pm
comeinvest wrote: Wed Jan 11, 2023 10:43 am
Kbg wrote: Wed Jan 11, 2023 9:22 am
comeinvest wrote: Tue Jan 10, 2023 4:29 am

I think most major foreign equity index futures are actually treated as 1256 contracts.
You assume that at your tax audit peril. It’s complex and the 6781 form instructions flag the issue. The foreign exchange has to make an application to the relevant US regulators and be approved for the legal status that enables US citizens to claim 1256. And many do not. As a minimum, one needs to do the research on whatever foreign contracts they are trading.
I don't have time right now to look it up again, but I did exactly that research. The major index futures on Eurex and in Osaka applied for approval. I don't care because I don't use equity index futures in taxable accounts. Let us know when you find the list.
Why wouldn’t a US person use NYSE Liffe? With MSCI EAFE and MSCI EM futures, what more do you need?
According to my research, the MSCI EAFE and MSCI EM futures have relatively high implied financing cost relative to the risk-free rate of the currency they are denominated in (USD), in comparison to the futures on EUREX and the Nikkei futures. My results are not yet very conclusive, but I posted some preliminary comparisons of the implied financing costs a few pages up in this thread viewtopic.php?p=6987262#p6987262 . I'll do some more backtests once I have time.
For emini sp500 we can go long box spread on options for getting a yield on the margin ( I usually keep 60% of notional value).
Similar things don't exist for MSCI index futures at NYSE Liffe.

For eurex the yield is negative. So not sure if it helps.
km91
Posts: 583
Joined: Wed Oct 13, 2021 12:32 pm

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

Post by km91 »

comeinvest wrote: Mon Jan 23, 2023 7:21 pm My financial life definitely became more complex since I started mHFEA.

Would you mind sharing your asset allocation? What trend following funds do you use?
How does unwinding or rolling the box work? Is there ever a situation the box could get broken or collateral called, or liquidated by your broker?

50/50 DBMF/KMLM. AUM and volume are fine for the size of my position and the expense ratio is ok considering the asset class and there's good diversity in the markets they trade across. I don't know enough to be able to verify that they deliver the pure academic trend exposure that someone like AQR promises but AQMIX, DBMF, and KMLM are all in the same place over the last two years.

This is more of a model portfolio, I have some of the elements implemented in my own portfolio but not at this leverage

Code: Select all

| Exposure          | Within Asset Class | NAV %   | Gross % |
|-------------------|--------------------|---------|---------|
|                   |                    |         |         |
| Rates             |                    |         |         |
| ITT - Nominal     | 30.00%             | 43.500% | 13.81%  |
| LTT - Nominal     | 20.00%             | 29.000% | 9.21%   |
| ITT - TIPS        | 30.00%             | 43.500% | 13.81%  |
| LTT - TIPS        | 20.00%             | 29.000% | 9.21%   |
|                   | 100.00%            | 145.00% | 46.03%  |
|                   |                    |         |         |
| Equity            |                    |         |         |
| TSM - US          | 25.00%             | 32.50%  | 10.32%  |
| SCV   - US        | 15.00%             | 19.50%  | 6.19%   |
| TSM - Developed   | 25.00%             | 32.50%  | 10.32%  |
| SCV   - Developed | 15.00%             | 19.50%  | 6.19%   |
| TSM - EM          | 20.00%             | 26.00%  | 8.25%   |
|                   | 100.00%            | 130.00% | 41.27%  |
| Commodity         |                    |         |         |
| Gold              | 30.00%             | 12.00%  | 3.81%   |
| Trend             | 70.00%             | 28.00%  | 8.89%   |
|                   | 100.00%            | 40.00%  | 12.70%  |
|                   |                    |         |         |
| Total             |                    | 315.00% | 100.00% |
comeinvest
Posts: 1780
Joined: Mon Mar 12, 2012 6:57 pm

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

Post by comeinvest »

sharukh wrote: Mon Jan 23, 2023 8:16 pm
comeinvest wrote: Sun Jan 22, 2023 11:01 pm
typical.investor wrote: Sun Jan 22, 2023 9:10 pm
comeinvest wrote: Wed Jan 11, 2023 10:43 am
Kbg wrote: Wed Jan 11, 2023 9:22 am

You assume that at your tax audit peril. It’s complex and the 6781 form instructions flag the issue. The foreign exchange has to make an application to the relevant US regulators and be approved for the legal status that enables US citizens to claim 1256. And many do not. As a minimum, one needs to do the research on whatever foreign contracts they are trading.
I don't have time right now to look it up again, but I did exactly that research. The major index futures on Eurex and in Osaka applied for approval. I don't care because I don't use equity index futures in taxable accounts. Let us know when you find the list.
Why wouldn’t a US person use NYSE Liffe? With MSCI EAFE and MSCI EM futures, what more do you need?
According to my research, the MSCI EAFE and MSCI EM futures have relatively high implied financing cost relative to the risk-free rate of the currency they are denominated in (USD), in comparison to the futures on EUREX and the Nikkei futures. My results are not yet very conclusive, but I posted some preliminary comparisons of the implied financing costs a few pages up in this thread viewtopic.php?p=6987262#p6987262 . I'll do some more backtests once I have time.
For emini sp500 we can go long box spread on options for getting a yield on the margin ( I usually keep 60% of notional value).
Similar things don't exist for MSCI index futures at NYSE Liffe.

For eurex the yield is negative. So not sure if it helps.
You can do the same with the collateral for the MSCI futures that you do with the collateral for the /ES futures. What makes you believe it's different?

The EUR yields are no longer negative, unfortunately. Now you have to factor the missed interest on the EUR based maintenance margin requirement into the cost of leverage. I checked with Interactive Brokers, but was not able to get an answer if any EUR denominated government bonds can be used in lieu of cash collateral, as is possible with T-bills for USD denominated futures. So yes, that's a drawback. But Nikkei futures seem to have very low financing cost even relative to JPY risk-free rates, and on top you still lose next to nothing on interest on the collateral as prevailing interest rates are still near zero. So that's a very efficient way of leverage.
comeinvest
Posts: 1780
Joined: Mon Mar 12, 2012 6:57 pm

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

Post by comeinvest »

km91 wrote: Mon Jan 23, 2023 9:09 pm
comeinvest wrote: Mon Jan 23, 2023 7:21 pm My financial life definitely became more complex since I started mHFEA.

Would you mind sharing your asset allocation? What trend following funds do you use?
How does unwinding or rolling the box work? Is there ever a situation the box could get broken or collateral called, or liquidated by your broker?

50/50 DBMF/KMLM. AUM and volume are fine for the size of my position and the expense ratio is ok considering the asset class and there's good diversity in the markets they trade across. I don't know enough to be able to verify that they deliver the pure academic trend exposure that someone like AQR promises but AQMIX, DBMF, and KMLM are all in the same place over the last two years.

This is more of a model portfolio, I have some of the elements implemented in my own portfolio but not at this leverage

Code: Select all

| Exposure          | Within Asset Class | NAV %   | Gross % |
|-------------------|--------------------|---------|---------|
|                   |                    |         |         |
| Rates             |                    |         |         |
| ITT - Nominal     | 30.00%             | 43.500% | 13.81%  |
| LTT - Nominal     | 20.00%             | 29.000% | 9.21%   |
| ITT - TIPS        | 30.00%             | 43.500% | 13.81%  |
| LTT - TIPS        | 20.00%             | 29.000% | 9.21%   |
|                   | 100.00%            | 145.00% | 46.03%  |
|                   |                    |         |         |
| Equity            |                    |         |         |
| TSM - US          | 25.00%             | 32.50%  | 10.32%  |
| SCV   - US        | 15.00%             | 19.50%  | 6.19%   |
| TSM - Developed   | 25.00%             | 32.50%  | 10.32%  |
| SCV   - Developed | 15.00%             | 19.50%  | 6.19%   |
| TSM - EM          | 20.00%             | 26.00%  | 8.25%   |
|                   | 100.00%            | 130.00% | 41.27%  |
| Commodity         |                    |         |         |
| Gold              | 30.00%             | 12.00%  | 3.81%   |
| Trend             | 70.00%             | 28.00%  | 8.89%   |
|                   | 100.00%            | 40.00%  | 12.70%  |
|                   |                    |         |         |
| Total             |                    | 315.00% | 100.00% |
Nice, thanks. You create a new box on the date of expiration of the old one. Theoretical risk is debatable. Head over to the SPX box spread thread, plenty of information.

TIPS have higher cost of leverage than treasury futures, as they have no futures; in light of this, are they worth it in a leveraged portfolio? Did they do much good in 2022? Real rates rose, TIPS tanked like everything else, didn't they? With >= 100% of NAV in equities and with net debt, plenty of inflation protection already, or not?
km91
Posts: 583
Joined: Wed Oct 13, 2021 12:32 pm

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

Post by km91 »

comeinvest wrote: Tue Jan 24, 2023 1:21 am Nice, thanks. You create a new box on the date of expiration of the old one. Theoretical risk is debatable. Head over to the SPX box spread thread, plenty of information.

TIPS have higher cost of leverage than treasury futures, as they have no futures; in light of this, are they worth it in a leveraged portfolio? Did they do much good in 2022? Real rates rose, TIPS tanked like everything else, didn't they? With >= 100% of NAV in equities and with net debt, plenty of inflation protection already, or not?
Thanks

For this type of portfolio I don't necessarily view TIPS as inflation protection per se, more like a diversifier of the interest rate risk across real and nominal rates
comeinvest
Posts: 1780
Joined: Mon Mar 12, 2012 6:57 pm

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

Post by comeinvest »

km91 wrote: Tue Jan 24, 2023 11:22 am For this type of portfolio I don't necessarily view TIPS as inflation protection per se, more like a diversifier of the interest rate risk across real and nominal rates
Isn't it a rising expected inflation scenario by definition, when nominal rates were to rise more than real rates? In which case equites will rise in relatively short order, I would think. Perhaps with some delay.
sharukh
Posts: 571
Joined: Mon Jun 20, 2016 10:19 am

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

Post by sharukh »

comeinvest wrote: Tue Jan 24, 2023 12:03 am
sharukh wrote: Mon Jan 23, 2023 8:16 pm
comeinvest wrote: Sun Jan 22, 2023 11:01 pm
typical.investor wrote: Sun Jan 22, 2023 9:10 pm
comeinvest wrote: Wed Jan 11, 2023 10:43 am

I don't have time right now to look it up again, but I did exactly that research. The major index futures on Eurex and in Osaka applied for approval. I don't care because I don't use equity index futures in taxable accounts. Let us know when you find the list.
Why wouldn’t a US person use NYSE Liffe? With MSCI EAFE and MSCI EM futures, what more do you need?
According to my research, the MSCI EAFE and MSCI EM futures have relatively high implied financing cost relative to the risk-free rate of the currency they are denominated in (USD), in comparison to the futures on EUREX and the Nikkei futures. My results are not yet very conclusive, but I posted some preliminary comparisons of the implied financing costs a few pages up in this thread viewtopic.php?p=6987262#p6987262 . I'll do some more backtests once I have time.
For emini sp500 we can go long box spread on options for getting a yield on the margin ( I usually keep 60% of notional value).
Similar things don't exist for MSCI index futures at NYSE Liffe.

For eurex the yield is negative. So not sure if it helps.
You can do the same with the collateral for the MSCI futures that you do with the collateral for the /ES futures. What makes you believe it's different?

The EUR yields are no longer negative, unfortunately. Now you have to factor the missed interest on the EUR based maintenance margin requirement into the cost of leverage. I checked with Interactive Brokers, but was not able to get an answer if any EUR denominated government bonds can be used in lieu of cash collateral, as is possible with T-bills for USD denominated futures. So yes, that's a drawback. But Nikkei futures seem to have very low financing cost even relative to JPY risk-free rates, and on top you still lose next to nothing on interest on the collateral as prevailing interest rates are still near zero. So that's a very efficient way of leverage.
For some reason I believed that mxea futures are traded at ICE/NYSE Liffe. But it's options are traded on CME, a different exchange.
So a long options box at CME won't count as collateral for NYSE Liffe.

Also adding a mxea box options in IBKR watch list didn't clearly say what exchange is it traded. It clearly says for ES box options in watchlist.
km91
Posts: 583
Joined: Wed Oct 13, 2021 12:32 pm

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

Post by km91 »

comeinvest wrote: Tue Jan 24, 2023 3:15 pm Isn't it a rising expected inflation scenario by definition, when nominal rates were to rise more than real rates? In which case equites will rise in relatively short order, I would think. Perhaps with some delay.
The honest answer is I don't know. There's expected relationships between real rates, nominal rates, growth, inflation, and equity prices, but I won't pretend that I fully understand the long run implications on the returns of a portfolio. Given that, I feel the simplest solution, and the one that relies on the least amount of assumptions about the future is to hold nominal and real bonds. This is all in theory though, in the real world there's other practical considerations and trade offs to be made to actually execute such a portfolio, and nominal bonds might be good enough
comeinvest
Posts: 1780
Joined: Mon Mar 12, 2012 6:57 pm

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

Post by comeinvest »

sharukh wrote: Tue Jan 24, 2023 4:20 pm
comeinvest wrote: Tue Jan 24, 2023 12:03 am
sharukh wrote: Mon Jan 23, 2023 8:16 pm
comeinvest wrote: Sun Jan 22, 2023 11:01 pm
typical.investor wrote: Sun Jan 22, 2023 9:10 pm

Why wouldn’t a US person use NYSE Liffe? With MSCI EAFE and MSCI EM futures, what more do you need?
According to my research, the MSCI EAFE and MSCI EM futures have relatively high implied financing cost relative to the risk-free rate of the currency they are denominated in (USD), in comparison to the futures on EUREX and the Nikkei futures. My results are not yet very conclusive, but I posted some preliminary comparisons of the implied financing costs a few pages up in this thread viewtopic.php?p=6987262#p6987262 . I'll do some more backtests once I have time.
For emini sp500 we can go long box spread on options for getting a yield on the margin ( I usually keep 60% of notional value).
Similar things don't exist for MSCI index futures at NYSE Liffe.

For eurex the yield is negative. So not sure if it helps.
You can do the same with the collateral for the MSCI futures that you do with the collateral for the /ES futures. What makes you believe it's different?

The EUR yields are no longer negative, unfortunately. Now you have to factor the missed interest on the EUR based maintenance margin requirement into the cost of leverage. I checked with Interactive Brokers, but was not able to get an answer if any EUR denominated government bonds can be used in lieu of cash collateral, as is possible with T-bills for USD denominated futures. So yes, that's a drawback. But Nikkei futures seem to have very low financing cost even relative to JPY risk-free rates, and on top you still lose next to nothing on interest on the collateral as prevailing interest rates are still near zero. So that's a very efficient way of leverage.
For some reason I believed that mxea futures are traded at ICE/NYSE Liffe. But it's options are traded on CME, a different exchange.
So a long options box at CME won't count as collateral for NYSE Liffe.

Also adding a mxea box options in IBKR watch list didn't clearly say what exchange is it traded. It clearly says for ES box options in watchlist.
MXEA index options are traded on CBOE, same as SPX. So you can just as well use SPX options, they are probably more liquid. MXEA futures have no futures options.
sharukh
Posts: 571
Joined: Mon Jun 20, 2016 10:19 am

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

Post by sharukh »

comeinvest wrote: Tue Jan 24, 2023 5:01 pm
sharukh wrote: Tue Jan 24, 2023 4:20 pm
comeinvest wrote: Tue Jan 24, 2023 12:03 am
sharukh wrote: Mon Jan 23, 2023 8:16 pm
comeinvest wrote: Sun Jan 22, 2023 11:01 pm

According to my research, the MSCI EAFE and MSCI EM futures have relatively high implied financing cost relative to the risk-free rate of the currency they are denominated in (USD), in comparison to the futures on EUREX and the Nikkei futures. My results are not yet very conclusive, but I posted some preliminary comparisons of the implied financing costs a few pages up in this thread viewtopic.php?p=6987262#p6987262 . I'll do some more backtests once I have time.
For emini sp500 we can go long box spread on options for getting a yield on the margin ( I usually keep 60% of notional value).
Similar things don't exist for MSCI index futures at NYSE Liffe.

For eurex the yield is negative. So not sure if it helps.
You can do the same with the collateral for the MSCI futures that you do with the collateral for the /ES futures. What makes you believe it's different?

The EUR yields are no longer negative, unfortunately. Now you have to factor the missed interest on the EUR based maintenance margin requirement into the cost of leverage. I checked with Interactive Brokers, but was not able to get an answer if any EUR denominated government bonds can be used in lieu of cash collateral, as is possible with T-bills for USD denominated futures. So yes, that's a drawback. But Nikkei futures seem to have very low financing cost even relative to JPY risk-free rates, and on top you still lose next to nothing on interest on the collateral as prevailing interest rates are still near zero. So that's a very efficient way of leverage.
For some reason I believed that mxea futures are traded at ICE/NYSE Liffe. But it's options are traded on CME, a different exchange.
So a long options box at CME won't count as collateral for NYSE Liffe.

Also adding a mxea box options in IBKR watch list didn't clearly say what exchange is it traded. It clearly says for ES box options in watchlist.
MXEA index options are traded on CBOE, same as SPX. So you can just as well use SPX options, they are probably more liquid. MXEA futures have no futures options.
Very helpful thanks so much.

If you don't mind, can you please tell what is the relationship between NYSE Liffe and cboe w.r.t. mxea futures.
I thought NYSE Liffe is an exchange and mxea is traded there. Cboe is a different exchange and spx is traded there.
User avatar
typical.investor
Posts: 4032
Joined: Mon Jun 11, 2018 3:17 am

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

Post by typical.investor »

sharukh wrote: Tue Jan 24, 2023 6:07 pm
comeinvest wrote: Tue Jan 24, 2023 5:01 pm
sharukh wrote: Tue Jan 24, 2023 4:20 pm
comeinvest wrote: Tue Jan 24, 2023 12:03 am
sharukh wrote: Mon Jan 23, 2023 8:16 pm

For emini sp500 we can go long box spread on options for getting a yield on the margin ( I usually keep 60% of notional value).
Similar things don't exist for MSCI index futures at NYSE Liffe.

For eurex the yield is negative. So not sure if it helps.
You can do the same with the collateral for the MSCI futures that you do with the collateral for the /ES futures. What makes you believe it's different?

The EUR yields are no longer negative, unfortunately. Now you have to factor the missed interest on the EUR based maintenance margin requirement into the cost of leverage. I checked with Interactive Brokers, but was not able to get an answer if any EUR denominated government bonds can be used in lieu of cash collateral, as is possible with T-bills for USD denominated futures. So yes, that's a drawback. But Nikkei futures seem to have very low financing cost even relative to JPY risk-free rates, and on top you still lose next to nothing on interest on the collateral as prevailing interest rates are still near zero. So that's a very efficient way of leverage.
For some reason I believed that mxea futures are traded at ICE/NYSE Liffe. But it's options are traded on CME, a different exchange.
So a long options box at CME won't count as collateral for NYSE Liffe.

Also adding a mxea box options in IBKR watch list didn't clearly say what exchange is it traded. It clearly says for ES box options in watchlist.
MXEA index options are traded on CBOE, same as SPX. So you can just as well use SPX options, they are probably more liquid. MXEA futures have no futures options.
Very helpful thanks so much.

If you don't mind, can you please tell what is the relationship between NYSE Liffe and cboe w.r.t. mxea futures.
I thought NYSE Liffe is an exchange and mxea is traded there. Cboe is a different exchange and spx is traded there.
I don't know myself but at IBKR, if I put MXEA NYSELIFFE in my watchlist, I can (and have) trade it.
If I try to put MXEA CBOE in my watchlist ... making sure to select futures and not the index ... then selecting a contract, it gets added as MXEA NYSELIFFE.

Same if try to open a trade and search for it there it's always MXEA NYSELIFFE. So I don't see a way to trade it on CBOE.
comeinvest
Posts: 1780
Joined: Mon Mar 12, 2012 6:57 pm

Re: SOFR futures

Post by comeinvest »

comeinvest wrote: Fri Jan 06, 2023 6:27 pm
I found another paper on STIR (Eurodollar) futures in investment portfolios: https://papers.ssrn.com/sol3/papers.cfm ... id=1992153

They say: "Using various methods of analysis, we find that an equity portfolio is better diversified and provides higher returns by overlaying the short rate futures term premium strategy versus Treasuries."

I think mHFEA people who studied the performance of STT, ITT, and LTT, and who opted for STT and ITT vs. LTT based on historical risk-adjusted performance, should read this paper, which studies the effect of concentrating the duration risk and harvesting the term premium in the quarter that has the highest risk-adjusted return, which per the paper is the 4th quarter from present - a potential refinement of the mHFEA strategy.

The results seem to be remarkably consistent across various international markets.

Unfortunately the paper doesn't compare the performance of Eurodollar futures to that of treasury futures of equal duration (only treasuries); but we know that STIR futures must theoretically mimic treasury futures and treasuries.

The periods of positive correlation between equities and treasuries are unfortunately also missing, as STIR futures didn't exist before the late 1980ies.

Per the first chart below, for example the average return of a position in the 4th quarter is about 50% higher than that of a position the 12th quarter, with almost equal risk, if my interpretation is right. Unfortunately the diagram doesn't tell the drawdown risks of each quarter, but the paper compares equities portfolios diversified with both concentrated STIR futures or treasuries.

TP in the third chart is term premium 4 quarters out, TSY is 10-year treasury.

Image

Image

Image
Last edited by comeinvest on Thu Feb 02, 2023 9:11 pm, edited 12 times in total.
comeinvest
Posts: 1780
Joined: Mon Mar 12, 2012 6:57 pm

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

Post by comeinvest »

comeinvest wrote: Sun Oct 23, 2022 3:35 am Interest rates and term premia are the most important ingredients affecting return and risk of HFEA style strategies, in comparison to equities-only or unleveraged strategies. I did some reading on the weekend on expected interest rates, and found some relevant papers. There are many more, so this sample of papers is rather arbitrary.

The bond market term premium: what is it, and how can we measure it: https://www.bis.org/publ/qtrpdf/r_qt0706e.pdf
The term premium: https://www.nbim.no/en/publications/dis ... rm-premium
Predicting Risk Premia for Treasury Bonds: https://risk.edhec.edu/sites/risk/files ... _bonds.pdf
The Historical Decline in Real Interest Rates and Its Implications for CBO’s Projections: https://www.cbo.gov/system/files/2020-1 ... -rates.pdf
slides: https://www.cbo.gov/system/files/2021-0 ... -Rates.pdf

The last paper studies mostly expected *real* interest rates and their determinants, with a literature review. For me it was sobering to see how little understood interest rates and term premia are, and how difficult it is to forecast them. And that is even for real rates alone, not to mention expected inflation. There seems to be no consistent theory, but only (rather desperate) attempts at explaining, or shall we say describing, past interest rate levels and term premia, with a wide dispersion of expected or perceived equilibrium rates.

I only had time to glance over the paper, but based on the charts, my personal conclusion is that it is anybody's guess whether or not rates will keep increasing, or ever come down again from their current levels in our lifetimes. Please let me know if you come to a different conclusion.

I want to present only one out of many interesting charts from the paper:
Please note that the chart shows *real* interest rates. There seem to be decades-long prevailing rates levels, with no easily discernible long term trend, pattern, or mean reversion. I think current real ITT forward rates (forward rates ca. 4%, 2.5% expected equilibrium inflation -> 1.5% real forward rates) are approximately at the level of the purple line between 1972 and 1983 - not very high in historical context. I think this has vast implications for the expected risk and return of HFEA style strategies with relatively high bond allocations.

Non-leveraged bond investors are in a much better position. The higher (and positive) real rates will sooner or later catch up and compensate the buy-and-hold investor for the capital losses suffered in 2022. For leveraged investors it is not so clear. The term premium is much smaller than real rates in the first place, and it is not even clear if it increased from Dec 2021 to now, or if it will be positive in the future.
Let's do some math, very simplified (for example ignoring the effect of deleveraging, and no contributions) and with what I hope are reasonable and comparable expectations, just to illustrate risks and returns in comparison. With a 0.5% p.a. combined future term premium and rebalancing benefit of a leveraged fixed income overlay of 5-year duration ITT to an equities portfolio, and with a (for example, assuming) 3% ITT interest rate hike risk (as coincidentally happened in 2022, but it could stretch over longer periods with similarly negative consequences), the annual return / risk of the overlay would be 0.5 / 15 = ca. 0.03. (Which implies, among other things, that the annual excess returns might not be able to catch up with the generated portfolio loss within your investment horizon of 30 years, if the level of rates resets to this higher level and won't mean-revert.)
For the stock market, assuming 4% p.a. future equity risk premium and 50% drawdown risk (perhaps roughly the stock market equivalent to the likelihood of a 3% ITT interest rate hike in the bond markets), the annual return / risk of an unfunded overlay would be 4 / 50 = 0.08, by magnitudes higher than the ratio for ITT. The stock market performance will typically catch up within an investing career (probably much sooner) with any P/E revaluation (even if the revaluation is permanent), due to the nature of the exponential performance of its intrinsic value with an exponential base of 1.04 (assuming 4% real growth): 1.04^30 = ca. 3.2. Looking at the vastly different benefit/risk ratios makes me wonder why (m)HFEA with the percentage allocations most often proposed (e.g. 140% / 250%) has about equal risk allocations to the two assets. Shouldn't the allocations also reflect expected returns above the financing rate, and/or the risk of negative performance of the respective unfunded overlay (i.e. above the risk-free rate) over the investment horizon?
I think the key difference of my analysis to previous ones in this thread is the risk and return consideration over a typical investor lifetime or investment horizon of (m)HFEA (for example 30 years) instead of shorter term risk and return measures, and considering the spreads to the risk-free rate of each asset class, as both are leveraged above 100%. Please critique my analysis.

EDIT: I ignored the financing cost of treasury futures. There was some discussion recently whether it's closer the 0% or closer the 0.25%. If the financing cost is positive, it would tilt the above comparison even significantly more in favor of equities. Financing cost of equities have much less impact on relative returns.
EDIT: The rebalancing bonus might actually be very small or zero. Historically equities and treasuries had about zero correlation on average. Teh addition of treasuries will result in a small rebalancing bonus when the assets are less than perfectly correlated; but they will also add to the overall portfolio volatility decay during times when the assets have positive correlation like in 2022.

Image
Following up on my previous post on historical term premia and interest rates, I wanted to cross-link a discussion on historical term premia from another thread, along with interesting charts:
viewtopic.php?p=6129783#p6129783
viewtopic.php?p=6144396#p6144396

With so much inconsistent data regarding both the existence of a term premium and the negative correlation of bonds and stocks, please convince me that (m)HFEA still makes sense! (Should history repeat, I'm not too keen on inadvertently leveraging the 1871 to 1930 portion of the yield spread chart 3x or even 2x, if I can avoid it.)

Image

Image
DMoogle
Posts: 502
Joined: Sat Oct 31, 2020 10:24 pm

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

Post by DMoogle »

So deploying this strategy may require a significant amount of cash just sitting in an account. I currently have my IRA in Etrade, which appears to only offer 0.01%-0.15% (https://us.etrade.com/l/options-uninvested-cash). Pitiful. Now that savings rates are much higher, this is actually relevant. I know IB offers benchmark minus 0.5% or so. Are there any other services that offer more, or should I just move my IRA to IB?
OrangeMan234
Posts: 4
Joined: Sun Dec 05, 2021 1:12 pm

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

Post by OrangeMan234 »

I'm interested in implementing mHFEA in a taxable account using LETFs rather than futures with a relatively small share of my portfolio. (I have a pretty basic understanding of futures, I’m not that interested in becoming proficient, and I worry that the added complexity will lead to mistakes over time.)

My target allocation is 135 SPY / 315 ITT — it's not quite the 165 / 400 skierincolorado landed on, but it still yields a historical annual return north of 20% with a max drawdown just over 50% and a lower leverage requirement. Like a few others up-thread, I plan to use TYA and UPRO to get most of the way there, but I'll need to add about 35% portfolio margin to reach my intended allocation. I plan to increase or decrease margin when I rebalance quarterly to maintain a constant leverage rate.

A couple of questions for the brain trust. First off, I notice that TYA seems to generate 3.5x the returns of VFITX (backtest here: https://www.portfoliovisualizer.com/bac ... on3_1=-250), not the 2.5/2.7/3x I’ve seen elsewhere. Any idea why that might be? Obviously TYA hasn't been around very long… Would we expect that relationship to change?

Anyway, assuming that relationship holds up, I would reproduce the performance of 135 VFINX/315 VFITX using 33% UPRO / 67% TYA and adding 35% margin for 45% UPRO / 90% TYA. This backtest (https://www.portfoliovisualizer.com/bac ... on10_2=-35) matches the performance of 135 VFINX /315 VFITX fairly closely. Have I calculated the combined leverage from the ETFs and portfolio margin correctly?

And lastly, holding well-known inflation risks aside, I'm open to thoughts on whether this is ill-advised, particularly relative to ETF-only solutions that avoid portfolio margin (like pure HFEA or lower leverage mHFEA). I gather rebalancing will lead to tax drag of 1-2% p.a.; ETF fees and tracking error may lop off another percent relative to the back tests; and portfolio margin will take another 1.5-2% when averaged across the entire strategy in the near term. On the flip side, I can probably avoid taxes for a few years via deferred tax loss harvesting and by rebalancing with new contributions; margin interest is obviously deductible; and rates are expected to come down over the long term.
Topic Author
skierincolorado
Posts: 2063
Joined: Sat Mar 21, 2020 10:56 am

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

Post by skierincolorado »

OrangeMan234 wrote: Mon Feb 06, 2023 3:16 pm I'm interested in implementing mHFEA in a taxable account using LETFs rather than futures with a relatively small share of my portfolio. (I have a pretty basic understanding of futures, I’m not that interested in becoming proficient, and I worry that the added complexity will lead to mistakes over time.)

My target allocation is 135 SPY / 315 ITT — it's not quite the 165 / 400 skierincolorado landed on, but it still yields a historical annual return north of 20% with a max drawdown just over 50% and a lower leverage requirement. Like a few others up-thread, I plan to use TYA and UPRO to get most of the way there, but I'll need to add about 35% portfolio margin to reach my intended allocation. I plan to increase or decrease margin when I rebalance quarterly to maintain a constant leverage rate.

A couple of questions for the brain trust. First off, I notice that TYA seems to generate 3.5x the returns of VFITX (backtest here: https://www.portfoliovisualizer.com/bac ... on3_1=-250), not the 2.5/2.7/3x I’ve seen elsewhere. Any idea why that might be? Obviously TYA hasn't been around very long… Would we expect that relationship to change?

Anyway, assuming that relationship holds up, I would reproduce the performance of 135 VFINX/315 VFITX using 33% UPRO / 67% TYA and adding 35% margin for 45% UPRO / 90% TYA. This backtest (https://www.portfoliovisualizer.com/bac ... on10_2=-35) matches the performance of 135 VFINX /315 VFITX fairly closely. Have I calculated the combined leverage from the ETFs and portfolio margin correctly?

And lastly, holding well-known inflation risks aside, I'm open to thoughts on whether this is ill-advised, particularly relative to ETF-only solutions that avoid portfolio margin (like pure HFEA or lower leverage mHFEA). I gather rebalancing will lead to tax drag of 1-2% p.a.; ETF fees and tracking error may lop off another percent relative to the back tests; and portfolio margin will take another 1.5-2% when averaged across the entire strategy in the near term. On the flip side, I can probably avoid taxes for a few years via deferred tax loss harvesting and by rebalancing with new contributions; margin interest is obviously deductible; and rates are expected to come down over the long term.
Don't get sucked in by pretty backtests... the backtest in the OP was purely for comparison to HFEA and not something I would implement.

1) don't break your portfolio up, it introduces inefficiencies, behavioral errors and irrational path dependency. Look at your portfolio as a whole and decide how much leverage is appropriate for the whole portfolio and then implement that. To implement you can leverage just part but it should be viewed as a whole when rebalancing.

2) leverage mostly only makes sense for people with significant contributions remaining in their life, ie lifecycle investing

3) leveraging stocks is more beneficial than leveraging bonds in lifecycle investing. I'd pick something more stock heavy.
comeinvest
Posts: 1780
Joined: Mon Mar 12, 2012 6:57 pm

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

Post by comeinvest »

OrangeMan234 wrote: Mon Feb 06, 2023 3:16 pm My target allocation is 135 SPY / 315 ITT — it's not quite the 165 / 400 skierincolorado landed on, but it still yields a historical annual return north of 20% with a max drawdown just over 50% and a lower leverage requirement. Like a few others up-thread, I plan to use TYA and UPRO to get most of the way there, but I'll need to add about 35% portfolio margin to reach my intended allocation. I plan to increase or decrease margin when I rebalance quarterly to maintain a constant leverage rate.
135/315 had 71% max drawdown in 2022 per portfoliovisualizer; probably more than that intra-month. I think most of skier's recommendations were less bonds than that - depending on age and investment horizon of course.

You are mixing several implementations. When you already use margin, then it's probably more efficient to just leverage low-cost ETFs. However, margin is so expensive that it may negate the excess returns of the strategy.

I don't mean to change your mind, but I think whoever is capable of studying dozens of pages of (m)HFEA material, run and interpret backtests and risks, and plan out (m)HFEA for a time horizon appropriate for these strategies (probably your life), is also capable (at a fraction of the time it takes to understand the risk and return of those strategies) of learning how to implement it with futures or with SPX options boxes, and pocket the additional risk-free 1% p.a. (UPRO/TMF) or whatever the ER of your ETFs is, for the rest of your life. Just my personal $0.02. Think of it, despite all backtests, the excess return from (m)HFEA is unknown in the end, and may or may not materialize; the expenses are certain, and compound over your life at a rate of 1.01 to a multiplier of 1.01^(#years). P.S.: I was an options and futures novice just a short time ago.
OrangeMan234
Posts: 4
Joined: Sun Dec 05, 2021 1:12 pm

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

Post by OrangeMan234 »

skierincolorado wrote: Mon Feb 06, 2023 4:09 pm don't break your portfolio up, it introduces inefficiencies, behavioral errors and irrational path dependency. Look at your portfolio as a whole and decide how much leverage is appropriate for the whole portfolio and then implement that. To implement you can leverage just part but it should be viewed as a whole when rebalancing.
I was implicitly following the approach of bucketing this strategy off in it's own corner of the portfolio (a la Hedgefundie's original post) because of the high risk rather than following the lifecycle investing model. I'll be roughly 20% leveraged across my portfolio, though, and will reduce leverage in this bucket as I approach retirement, so the net impact likely won't be all that different.
skierincolorado wrote: Mon Feb 06, 2023 4:09 pm leverage mostly only makes sense for people with significant contributions remaining in their life, ie lifecycle investing
I've probably contributed about 35% of my expected inflation-adjusted lifetime contributions (started at 30 and my income has risen significantly since hitting 40).
comeinvest wrote: Mon Feb 06, 2023 4:16 pm 135/315 had 71% max drawdown in 2022 per portfoliovisualizer; probably more than that intra-month. I think most of skier's recommendations were less bonds than that - depending on age and investment horizon of course.
I was incorrectly referencing annual drawdowns when I cited a "max" drawdown in the low 50s earlier for 135/315. Looks like the actual max drawdown is ~58%, I think? https://www.portfoliovisualizer.com/bac ... on3_1=-350.
comeinvest wrote: Mon Feb 06, 2023 4:16 pm You are mixing several implementations. When you already use margin, then it's probably more efficient to just leverage low-cost ETFs. However, margin is so expensive that it may negate the excess returns of the strategy.
I share the concerns on (a) mixing implementations, and (b) using portfolio margin. My addition of portfolio margin may have made more sense when portfolio margin rates were only 3.5% a year ago. I think I'm now learning towards allocating a slightly larger slice of taxable to this strategy, but without the addition of portfolio margin.
Topic Author
skierincolorado
Posts: 2063
Joined: Sat Mar 21, 2020 10:56 am

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

Post by skierincolorado »

OrangeMan234 wrote: Mon Feb 06, 2023 7:18 pm
skierincolorado wrote: Mon Feb 06, 2023 4:09 pm don't break your portfolio up, it introduces inefficiencies, behavioral errors and irrational path dependency. Look at your portfolio as a whole and decide how much leverage is appropriate for the whole portfolio and then implement that. To implement you can leverage just part but it should be viewed as a whole when rebalancing.
I was implicitly following the approach of bucketing this strategy off in it's own corner of the portfolio (a la Hedgefundie's original post) because of the high risk rather than following the lifecycle investing model. I'll be roughly 20% leveraged across my portfolio, though, and will reduce leverage in this bucket as I approach retirement, so the net impact likely won't be all that different.
skierincolorado wrote: Mon Feb 06, 2023 4:09 pm leverage mostly only makes sense for people with significant contributions remaining in their life, ie lifecycle investing
I've probably contributed about 35% of my expected inflation-adjusted lifetime contributions (started at 30 and my income has risen significantly since hitting 40).
comeinvest wrote: Mon Feb 06, 2023 4:16 pm 135/315 had 71% max drawdown in 2022 per portfoliovisualizer; probably more than that intra-month. I think most of skier's recommendations were less bonds than that - depending on age and investment horizon of course.
I was incorrectly referencing annual drawdowns when I cited a "max" drawdown in the low 50s earlier for 135/315. Looks like the actual max drawdown is ~58%, I think? https://www.portfoliovisualizer.com/bac ... on3_1=-350.
comeinvest wrote: Mon Feb 06, 2023 4:16 pm You are mixing several implementations. When you already use margin, then it's probably more efficient to just leverage low-cost ETFs. However, margin is so expensive that it may negate the excess returns of the strategy.
I share the concerns on (a) mixing implementations, and (b) using portfolio margin. My addition of portfolio margin may have made more sense when portfolio margin rates were only 3.5% a year ago. I think I'm now learning towards allocating a slightly larger slice of taxable to this strategy, but without the addition of portfolio margin.
HFEAs bucketing is a mistake and a psychological crutch. You've addressed one half of the mistake which is that you say you will reduce the amount in the bucket as the bucket gets bigger and/or you age. But you have to address the other half of the mistake which means increasing the size of the bucket if the market goes down. It doesn't make sense to take 20% leverage when the market is at X but only 10% leverage if the market drops 20%. If anything you should be more leveraged if the market drops, or at least the same leverage as initial.


Why would you sell shares of SPY as the market drops? That's what UPRO does.

Thr problem of inefficiency and irrational leverage glidepath is made worse the more leveraged the bucket is. Taken to the extreme you could have a tiny bucket that is leveraged 20x and say that overall you are 1.3x leveraged. But the extreme volatility decay of a 20x portfolio is guaranteed to underform an alternate version of 1.3x using a larger 3x bucket (due to volatility decay 20x leverage will drop to zero while a bigger 3x bucket will not). Take that a step further and you will realize a 3x bucket is guaranteed to underperform just leveraging the whole portfolio 1.3x.

So yes I greatly prefer a larger bucket that is less leveraged without using margin. And only if you have a plan for increasing the size of the bucket if it shrinks and decreasing the size of the bucket if it grows. Otherwise you are just giving up too much to volatility.



Those of us who did not segregated buckets have faired better this year because we maintained our leverage when the market dropped and have benefited from the rebound in bonds and stocks. If I had a segregated bucket I would be much poorer today. I do use LETF (UPRO) but I rebalanced into it heavily to maintain my overall 1.5x leverage. Otherwise my leverage would have dropped. Not only did I maintain my leverage, I increased it. And I leveraged my new contributions.


There's a lot of volatility decay in LETFs that's hidden. If you were using margin you wouldn't sell shares as the market dropped would you? If you had a lot of leverage your plan might involve selling some to control risk... but it would not sell as much as Upro.
Last edited by skierincolorado on Mon Feb 06, 2023 10:21 pm, edited 1 time in total.
lawyeredCLO
Posts: 42
Joined: Wed Feb 09, 2022 10:52 am

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

Post by lawyeredCLO »

[/quote]
HFEAs bucketing is a mistake and a psychological crutch. You've addressed one half of the mistake which is that you say you will reduce the amount in the bucket as the bucket gets bigger and/or you age. But you have to address the other half of the mistake which means increasing the size of the bucket if the market goes down. It doesn't make sense to take 20% leverage when the market is at X but only 10% leverage if the market drops 20%. If anything you should be more leveraged if the market drops, or at least the same leverage as initial.


Why would you sell shares of SPY as the market drops? That's what UPRO does.

Thr problem of inefficiency and irrational leverage glidepath is made worse the more leveraged the bucket is. Taken to the extreme you could have a tiny bucket that is leveraged 20x and say that overall you are 1.3x leveraged. But the extreme volatility decay of a 20x portfolio is guaranteed to underform an alternate version of 1.3x using a larger 3x bucket (due to volatility decay 20x leverage will drop to zero). Which is guaranteed to underperform just leveraging the whole portfolio 1.3x.

So yes I greatly prefer a larger bucket that is less leveraged without using margin. And only if you have a plan for increasing the size of the bucket if it shrinks and decreasing the size of the bucket if it grows. Otherwise you are just giving up too much to volatility.



Those of us who did not segregated buckets have faired better this year because we maintained our leverage when the market dropped and have benefited from the rebound in bonds and stocks. If I had a segregated bucket I would be much poorer today. I do use LETF (UPRO) but I rebalanced into it heavily to maintain my overall 1.5x leverage. Otherwise my leverage would have dropped. Not only did I maintain my leverage, I increased it. And I leveraged my new contributions.


There's a lot of volatility decay in LETFs that's hidden. If you were using margin you wouldn't sell shares as the market dropped would you? If you had a lot of leverage your plan might involve selling some to control risk... but it would not sell as much as Upro.
[/quote]

I originally liked the idea of bucketing (and yes it was a crutch), but the more you think about it, it makes more sense to try and run some sort of a 1.25x-2x leverage across your entire assets for all the reasons you stated.
Topic Author
skierincolorado
Posts: 2063
Joined: Sat Mar 21, 2020 10:56 am

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

Post by skierincolorado »

lawyeredCLO wrote: Mon Feb 06, 2023 8:16 pm

I originally liked the idea of bucketing (and yes it was a crutch), but the more you think about it, it makes more sense to try and run some sort of a 1.25x-2x leverage across your entire assets for all the reasons you stated.
Glad to hear it! Most people aren't that receptive to those points. Bucketing LETFs is popular for the simplicity and psychological ease. But it glosses over a lot of the complexities and volatility decay that happens internally. A strong backtest in a strong market hides all that even more. But bucketing LETFs is certainly not the AA or Investment Plan that we would design if we were implementing from scratch with margin. We might sell shares during market drops to control risk, but certainly we would sell as many shares as UPRO sells. LETFs are still useful because the internal borrowing rate is much less than most broker's margin rates, but we have to be careful about how we use them.
Post Reply