Critique this leveraged portfolio

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Critique this leveraged portfolio

Post by danielc »

Recently there have been several threads about leveraged portfolios (e.g. HEDGEFUNDIE's excellent adventure, WisdomTree 90/60 Fund). According to MPT we should all be holding whichever portfolio has the highest Sharpe ratio, which in the US is around 30% stocks / 70% bonds, and then using leverage or cash to adjust the risk. Let me present a leveraged alternative to Vanguard's LifeStrategy Growth Fund, using 2x daily leveraged ETFs from ProShares:

Code: Select all

Leveraged portfolio:

18%  SSO   --  2 x daily S&P500               
 9%  EFO   --  2 x daily MSCI EAFE            
 3%  EET   --  2 x daily MSCI Emerging Markets
70%  UST   --  2 x daily 7-10 Year Treasury   
  • The stocks have the same 60% US / 40% Intl that Vanguard uses for its balanced funds.
  • The treasuries have a similar duration to Vanguard's total bond index fund.
  • The stock/treasury ratio is chosen to have a high Sharpe ratio, without a lot of fine tuning.
Using an admittedly short data set of 10 years, Portfolio Visualizer says that the leveraged portfolio has a higher CAGR, lower Stdev, smaller drops (both "worst year" and "max DD"), and better Sharpe and Sortino ratios than Vanguard's fund:

Code: Select all

            Vanguard VASGX     Leveraged Portfolio
CAGR          9.58%              11.26%
Stdev        10.32%               8.49%
Best Year    23.13%              25.02%
Worst Year   -6.90%              -6.34%
Max DD      -16.17%             -10.34%
Sharpe        0.88                1.24
Sortino       1.43                2.19
Recognizing that daily leveraged ETFs are not exactly what the creators of MPT had in mind... shouldn't we all be using leveraged ETFs as our core portfolio?
ukbogler
Posts: 195
Joined: Mon Sep 30, 2019 4:38 am

Re: Critique this leveraged portfolio

Post by ukbogler »

I'm no expert but I'd say you need a longer data series. 10 years = boomtime.
User avatar
Forester
Posts: 2400
Joined: Sat Jan 19, 2019 1:50 pm
Location: UK

Re: Critique this leveraged portfolio

Post by Forester »

Vulnerable to rising inflation, rising rates & a lost decade for US stocks.
Amateur Self-Taught Senior Macro Strategist
lazyday
Posts: 3849
Joined: Wed Mar 14, 2007 10:27 pm

Re: Critique this leveraged portfolio

Post by lazyday »

I wouldn’t depend on the future being like the past.

We’re living with trillions in negative nominal yield bonds, and massive QE. Nobody knows how that will change things.
Valuethinker
Posts: 49019
Joined: Fri May 11, 2007 11:07 am

Re: Critique this leveraged portfolio

Post by Valuethinker »

danielc wrote: Mon Jan 20, 2020 5:18 am Recently there have been several threads about leveraged portfolios (e.g. HEDGEFUNDIE's excellent adventure, WisdomTree 90/60 Fund). According to MPT we should all be holding whichever portfolio has the highest Sharpe ratio, which in the US is around 30% stocks / 70% bonds, and then using leverage or cash to adjust the risk. Let me present a leveraged alternative to Vanguard's LifeStrategy Growth Fund, using 2x daily leveraged ETFs from ProShares:

Code: Select all

Leveraged portfolio:

18%  SSO   --  2 x daily S&P500               
 9%  EFO   --  2 x daily MSCI EAFE            
 3%  EET   --  2 x daily MSCI Emerging Markets
70%  UST   --  2 x daily 7-10 Year Treasury   
  • The stocks have the same 60% US / 40% Intl that Vanguard uses for its balanced funds.
  • The treasuries have a similar duration to Vanguard's total bond index fund.
  • The stock/treasury ratio is chosen to have a high Sharpe ratio, without a lot of fine tuning.
Using an admittedly short data set of 10 years, Portfolio Visualizer says that the leveraged portfolio has a higher CAGR, lower Stdev, smaller drops (both "worst year" and "max DD"), and better Sharpe and Sortino ratios than Vanguard's fund:

Code: Select all

            Vanguard VASGX     Leveraged Portfolio
CAGR          9.58%              11.26%
Stdev        10.32%               8.49%
Best Year    23.13%              25.02%
Worst Year   -6.90%              -6.34%
Max DD      -16.17%             -10.34%
Sharpe        0.88                1.24
Sortino       1.43                2.19
Recognizing that daily leveraged ETFs are not exactly what the creators of MPT had in mind... shouldn't we all be using leveraged ETFs as our core portfolio?
What happens if you run it through 2008-09? 2000-03?

Were these instruments available in either time period? Because you do not know until it happens. Bonds you always want to run through 1994 (I think 1980 was even more awful but that is hopefully a non-repeatable outlier; ie short term interest rates going to 21%, the highest level ever achieved in American history outside (perhaps) of wartime?).

What's the position with dividends from the S&P 500 with these instruments.
redstar
Posts: 125
Joined: Thu Jul 13, 2017 11:15 pm

Re: Critique this leveraged portfolio

Post by redstar »

I wanted to try to recreate this with leverage via futures, both because it should be cheaper (fewer fees and maybe lower borrowing costs) and some people dislike the leverage reset (although it is likely safer).

The equity side is pretty easy to hold with just ETFs, and the treasuries can be leveraged (similar to DonIce's thread).

36% VTI (Total US Stock)
24% VXUS (Total Intl. Stock)
140% /ZN (10-year treasuries)
Total Leverage Ratio: 2.00

Actually implementing this can be a bit annoying due to the large size of the treasury futures. If we had a portfolio worth $100k, that would be:
$36k VTI
$24k VXUS
1 \ZN future (notional value: $128,937.50, collateral: $40k)
Meaning the actual leverage ratio is only: 1.89
User avatar
firebirdparts
Posts: 4411
Joined: Thu Jun 13, 2019 4:21 pm
Location: Southern Appalachia

Re: Critique this leveraged portfolio

Post by firebirdparts »

I may be getting beyond my ability here, but I am of the opinion that the 3X leveraged funds have proven to be okay to hold long term. 2X funds have not. I am not worried about investigating that any further, but you should be if you were going to buy and hold them.
This time is the same
User avatar
Uncorrelated
Posts: 1237
Joined: Sun Oct 13, 2019 3:16 pm

Re: Critique this leveraged portfolio

Post by Uncorrelated »

You are way overweight on leveraged bonds. Running a mean variance optimizer on a portfolio of stocks and bonds with realistic settings will generally result in a long term (20 year duration) bond exposure that is between 10 and 30% of the equity exposure. You have 50%. Yes I know nisiprius shows a max sharpe ratio at 30/70, the problem is that there is no reason to optimize for max sharpe ratio in the first place. The proper approach is to run a mean-variance optimization to find the portfolio with the highest utility, which unfortunately is somewhat complicated to do.

I have previously mentioned that it is better to use SSO than UPRO but I have since then changed my mind because the expense ratio on these funds is not proportional to leverage. It is slightly better to buy UPRO + non-leveraged total stock market fund than it is to buy SSO.

I don't like leveraged ITT because the expense ratio is extremely high compared to the expected earnings. With the current yields assuming interest rates do not change, 3x leveraged ITT's yield less than "risk free" t-bills. However, more research is needed to confirm the dynamics of ITT in a complete portfolio.

Is this your entire portfolio or just a part? It is not a good idea to buy leveraged funds if you are only using it as a part of your portfolio. If you have an emergency fund, you might want to eliminate that too.
User avatar
firebirdparts
Posts: 4411
Joined: Thu Jun 13, 2019 4:21 pm
Location: Southern Appalachia

Re: Critique this leveraged portfolio

Post by firebirdparts »

danielc wrote: Mon Jan 20, 2020 5:18 am shouldn't we all be using leveraged ETFs as our core portfolio?
Let me just comment on this also. If you are going to get the benefits of leverage, you'd have to get it from somebody else. The cost of finding enough "somebody else's" would be an issue. If the leverage was based on actual borrowed money, then it is more scalable, but these funds don't really operate that way.
This time is the same
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: Critique this leveraged portfolio

Post by klaus14 »

Do you monitor stock-bond correlations? Will you change your alloc as correlations change?

I think you also need a uncorrelated 3rd asset to overweight in case stock-bond correlations turn positive. That asset is gold.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
guyinlaw
Posts: 727
Joined: Wed Jul 03, 2019 9:54 am

Re: Critique this leveraged portfolio

Post by guyinlaw »

redstar wrote: Mon Jan 20, 2020 9:10 am I wanted to try to recreate this with leverage via futures, both because it should be cheaper (fewer fees and maybe lower borrowing costs) and some people dislike the leverage reset (although it is likely safer).

The equity side is pretty easy to hold with just ETFs, and the treasuries can be leveraged (similar to DonIce's thread).

36% VTI (Total US Stock)
24% VXUS (Total Intl. Stock)
140% /ZN (10-year treasuries)
Total Leverage Ratio: 2.00

Actually implementing this can be a bit annoying due to the large size of the treasury futures. If we had a portfolio worth $100k, that would be:
$36k VTI
$24k VXUS
1 \ZN future (notional value: $128,937.50, collateral: $40k)
Meaning the actual leverage ratio is only: 1.89
1 \ZN 10 year future only needs around $1,400 margin. remaining collateral can be invested in Ultra short term bonds (SHY, ICSH). IMO you need around $10k collateral, see the max drawdowns..

One concern that I see is of performance when interest rates rise and stock fall.. (similar issues with NTSX) These risks have been discounted by some.

========

I would not recommend leveraged treasury ETFs, when it is so inexpensive and easy to get leverage using Treasury futures..

With around $10k collateral ($1,400 margin) you can get exposure to $128,937.50 in 10y treasury future..
Time is your friend; impulse is your enemy. - John C. Bogle
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Re: Critique this leveraged portfolio

Post by danielc »

Valuethinker wrote: Mon Jan 20, 2020 7:02 am What happens if you run it through 2008-09? 2000-03?

Were these instruments available in either time period?
The leveraged S&P 500 ETF was available in 2006, but the others were not. Most critically, the leveraged bonds did not exist, so you can't even look at a US-only portfolio. If you simulate the leveraged bonds by inserting negative CASHX in portfolio visualizer and focus on a US-only stocks, you get a leveraged portfolio that is better than the unleveraged one by every measure. Unfortunately, you can't really simulate daily leveraged ETFs this way; they don't behave like margin. Some of the people posting in HEDGEFUNDIE's thread have a lot of experience modelling these ETFs, so maybe they'll jump in.
Valuethinker wrote: Mon Jan 20, 2020 7:02 am What's the position with dividends from the S&P 500 with these instruments.
They follow the total return index, not just the price index. So you effectively get dividends too.
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Re: Critique this leveraged portfolio

Post by danielc »

firebirdparts wrote: Mon Jan 20, 2020 9:26 am I may be getting beyond my ability here, but I am of the opinion that the 3X leveraged funds have proven to be okay to hold long term. 2X funds have not. I am not worried about investigating that any further, but you should be if you were going to buy and hold them.
How would be possible for 3x leveraged funds to be ok, and 1x (i.e. unleveraged) to be ok, but somehow 2x is not ok? Also, the 2x leveraged funds have been around longer. The 2x S&P 500 appeared in 2006 so you can even see how it behaved during the financial crisis.
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Re: Critique this leveraged portfolio

Post by danielc »

Uncorrelated wrote: Mon Jan 20, 2020 10:17 am You are way overweight on leveraged bonds. Running a mean variance optimizer on a portfolio of stocks and bonds with realistic settings will generally result in a long term (20 year duration) bond exposure that is between 10 and 30% of the equity exposure. You have 50%.
None of the bonds are long bonds. They are 7-10 year treasuries.
Uncorrelated wrote: Mon Jan 20, 2020 10:17 am Yes I know nisiprius shows a max sharpe ratio at 30/70, the problem is that there is no reason to optimize for max sharpe ratio in the first place. The proper approach is to run a mean-variance optimization to find the portfolio with the highest utility, which unfortunately is somewhat complicated to do.
As far as I understand, max Sharpe ratio gives the tangency portfolio. Given that I just want to beat a generic LifeStrategy fund, and we don't want to over-optimize a 10 year sample, isn't max Sharpe ratio "good enough"?

Uncorrelated wrote: Mon Jan 20, 2020 10:17 am I have previously mentioned that it is better to use SSO than UPRO but I have since then changed my mind because the expense ratio on these funds is not proportional to leverage. It is slightly better to buy UPRO + non-leveraged total stock market fund than it is to buy SSO.
That's an interesting idea. So if I buy $1 of UPRO and $1 of some S&P 500 index fund, I would effectively have 2x leverage for cheaper than if I bought SSO? Wouldn't you need to rebalance daily if you really wanted to simulate SSO minus the expense ratio?

Uncorrelated wrote: Mon Jan 20, 2020 10:17 am Is this your entire portfolio or just a part? It is not a good idea to buy leveraged funds if you are only using it as a part of your portfolio. If you have an emergency fund, you might want to eliminate that too.
Currently it is 0% of my portfolio, but I am currently thinking of putting maybe 10% of my portfolio in HEDGEFUNDIE's adventure portfolio (i.e. risky bet; might win big) and maybe 10-20% of my portfolio in something similar to what I showed here ---- a leveraged portfolio with lower volatility than then unleveraged portfolio that it replaces.
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Re: Critique this leveraged portfolio

Post by danielc »

guyinlaw wrote: Mon Jan 20, 2020 12:37 pm I would not recommend leveraged treasury ETFs, when it is so inexpensive and easy to get leverage using Treasury futures..

With around $10k collateral ($1,400 margin) you can get exposure to $128,937.50 in 10y treasury future..
Is it actually easy? I know very little about futures. How much would I need to learn before I was actually able to do what you suggested?
Last edited by danielc on Mon Jan 20, 2020 2:54 pm, edited 1 time in total.
langlands
Posts: 1093
Joined: Wed Apr 03, 2019 10:05 pm

Re: Critique this leveraged portfolio

Post by langlands »

Uncorrelated wrote: Mon Jan 20, 2020 10:17 am You are way overweight on leveraged bonds. Running a mean variance optimizer on a portfolio of stocks and bonds with realistic settings will generally result in a long term (20 year duration) bond exposure that is between 10 and 30% of the equity exposure. You have 50%. Yes I know nisiprius shows a max sharpe ratio at 30/70, the problem is that there is no reason to optimize for max sharpe ratio in the first place. The proper approach is to run a mean-variance optimization to find the portfolio with the highest utility, which unfortunately is somewhat complicated to do.
If we make the standard assumptions of lognormal returns, CRRA utility function etc. then optimizing for Sharpe ratio and utility function are related though, right? To maximize utility, you optimize for Sharpe ratio to get the correct relative weighting of assets, then solve the one dimensional Merton portfolio problem of determining how much leverage to take, which is (mu - r)/(sigma^2*gamma) where gamma is the CRRA.
User avatar
Uncorrelated
Posts: 1237
Joined: Sun Oct 13, 2019 3:16 pm

Re: Critique this leveraged portfolio

Post by Uncorrelated »

danielc wrote: Mon Jan 20, 2020 1:16 pm None of the bonds are long bonds. They are 7-10 year treasuries.
I'm aware of that. You currently have the equivalent exposure of 50% 20-year treasuries and 50% equities, but it would be better to pick 80% equities and 20% long-term treasuries because equities have higher return per unit of risk. (specifically, around 3 times as high).
As far as I understand, max Sharpe ratio gives the tangency portfolio. Given that I just want to beat a generic LifeStrategy fund, and we don't want to over-optimize a 10 year sample, isn't max Sharpe ratio "good enough"?
It is completely pointless to construct a tangency portfolio unless you have free leverage. There is no convincing reason to optimize for maximum sharpe ratio unless you have free leverage.
That's an interesting idea. So if I buy $1 of UPRO and $1 of some S&P 500 index fund, I would effectively have 2x leverage for cheaper than if I bought SSO? Wouldn't you need to rebalance daily if you really wanted to simulate SSO minus the expense ratio?
AFAIK the conclusion still holds if you rebalance yearly.
Uncorrelated wrote: Mon Jan 20, 2020 10:17 am Is this your entire portfolio or just a part? It is not a good idea to buy leveraged funds if you are only using it as a part of your portfolio. If you have an emergency fund, you might want to eliminate that too.
Currently it is 0% of my portfolio, but I am currently thinking of putting maybe 10% of my portfolio in HEDGEFUNDIE's adventure portfolio (i.e. risky bet; might win big) and maybe 10-20% of my portfolio in something similar to what I showed here ---- a leveraged portfolio with lower volatility than then unleveraged portfolio that it replaces.
That's a really bad idea. Don't think of your portfolio in buckets, think of your portfolio as a whole. It is likely that your goals are better served by replacing some bonds of your existing portfolio with equities, usage of factor funds, reverse mortgage or replacing one part of your portfolio (say, US stocks) with leveraged equities and keeping the rest the same.

Optimize your entire portfolio. Don't isolate a part of your portfolio and use that, that approach will very likely result in an inefficiently portfolio.
langlands wrote: Mon Jan 20, 2020 1:42 pm
Uncorrelated wrote: Mon Jan 20, 2020 10:17 am You are way overweight on leveraged bonds. Running a mean variance optimizer on a portfolio of stocks and bonds with realistic settings will generally result in a long term (20 year duration) bond exposure that is between 10 and 30% of the equity exposure. You have 50%. Yes I know nisiprius shows a max sharpe ratio at 30/70, the problem is that there is no reason to optimize for max sharpe ratio in the first place. The proper approach is to run a mean-variance optimization to find the portfolio with the highest utility, which unfortunately is somewhat complicated to do.
If we make the standard assumptions of lognormal returns, CRRA utility function etc. then optimizing for Sharpe ratio and utility function are related though, right? To maximize utility, you optimize for Sharpe ratio to get the correct relative weighting of assets, then solve the one dimensional Merton portfolio problem of determining how much leverage to take, which is (mu - r)/(sigma^2*gamma) where gamma is the CRRA.
That is correct if you have free leverage, but SSO costs approximately 2% per year.
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Critique this leveraged portfolio

Post by Steve Reading »

Uncorrelated wrote: Mon Jan 20, 2020 3:25 pm
That's a really bad idea. Don't think of your portfolio in buckets, think of your portfolio as a whole. It is likely that your goals are better served by replacing some bonds of your existing portfolio with equities, usage of factor funds, reverse mortgage or replacing one part of your portfolio (say, US stocks) with leveraged equities and keeping the rest the same.

Optimize your entire portfolio. Don't isolate a part of your portfolio and use that, that approach will very likely result in an inefficiently portfolio.
Whoah I couldn’t possibly agree more. It’s so amazing to me just how many BHs bucket off their portfolio, leading to inefficiencies like the one you mentioned. If you want more in stocks, don’t borrow from other. Sell your own bonds and borrow from yourself!

Uncorrelated wrote: Mon Jan 20, 2020 3:25 pm That is correct if you have free leverage, but SSO costs approximately 2% per year.
By free leverage you don’t actually mean borrowing for free right? You mean borrowing at the RFR yes?
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
User avatar
Uncorrelated
Posts: 1237
Joined: Sun Oct 13, 2019 3:16 pm

Re: Critique this leveraged portfolio

Post by Uncorrelated »

305pelusa wrote: Mon Jan 20, 2020 3:37 pm
Uncorrelated wrote: Mon Jan 20, 2020 3:25 pm That is correct if you have free leverage, but SSO costs approximately 2% per year.
By free leverage you don’t actually mean borrowing for free right? You mean borrowing at the RFR yes?

Correct.

For clarification, the leverage inside SSO costs approximately 1.4% per year (0.9% expense ratio, 0.5% borrow costs) on top of the risk free rate and UPRO costs around 2% (0.95% expense ratio 1% borrow cost). This makes UPRO a better deal on paper because you only need half as much to get the same equity exposure (if you combine it with a cheap unleveraged vanguard fund). I don't know if UPRO is a deal you should take, but it's clearly a better deal than SSO.
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Critique this leveraged portfolio

Post by Steve Reading »

danielc wrote: Mon Jan 20, 2020 5:18 am Recently there have been several threads about leveraged portfolios (e.g. HEDGEFUNDIE's excellent adventure, WisdomTree 90/60 Fund). According to MPT we should all be holding whichever portfolio has the highest Sharpe ratio, which in the US is around 30% stocks / 70% bonds, and then using leverage or cash to adjust the risk. Let me present a leveraged alternative to Vanguard's LifeStrategy Growth Fund, using 2x daily leveraged ETFs from ProShares:

Code: Select all

Leveraged portfolio:

18%  SSO   --  2 x daily S&P500               
 9%  EFO   --  2 x daily MSCI EAFE            
 3%  EET   --  2 x daily MSCI Emerging Markets
70%  UST   --  2 x daily 7-10 Year Treasury   
  • The stocks have the same 60% US / 40% Intl that Vanguard uses for its balanced funds.
  • The treasuries have a similar duration to Vanguard's total bond index fund.
  • The stock/treasury ratio is chosen to have a high Sharpe ratio, without a lot of fine tuning.
Using an admittedly short data set of 10 years, Portfolio Visualizer says that the leveraged portfolio has a higher CAGR, lower Stdev, smaller drops (both "worst year" and "max DD"), and better Sharpe and Sortino ratios than Vanguard's fund:

Code: Select all

            Vanguard VASGX     Leveraged Portfolio
CAGR          9.58%              11.26%
Stdev        10.32%               8.49%
Best Year    23.13%              25.02%
Worst Year   -6.90%              -6.34%
Max DD      -16.17%             -10.34%
Sharpe        0.88                1.24
Sortino       1.43                2.19
Recognizing that daily leveraged ETFs are not exactly what the creators of MPT had in mind... shouldn't we all be using leveraged ETFs as our core portfolio?
Perhaps I’m wrong in my thinking OP but:

I have to say it’s a little strange to agree with all of the (dubious?) assumptions CAPM makes in order to conclude you should leverage the max Sharpe, but then CONVENIENTLY forget that under those very assumptions, the max Sharpe is already the market portfolio, not some 30/70 data mined concoction.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Re: Critique this leveraged portfolio

Post by danielc »

Uncorrelated wrote: Mon Jan 20, 2020 3:25 pm
danielc wrote: Mon Jan 20, 2020 1:16 pm None of the bonds are long bonds. They are 7-10 year treasuries.
I'm aware of that. You currently have the equivalent exposure of 50% 20-year treasuries and 50% equities, but it would be better to pick 80% equities and 20% long-term treasuries because equities have higher return per unit of risk. (specifically, around 3 times as high).
Can you show me how you derived those numbers? I guess that you are trying to work out the portfolio exposure to equity risk vs term risk. I figure that the leveraged portfolio has roughly the same ratio of equity risk and term risk as the 50/50 equity/LTT portfolio you suggest, but long treasuries aren't exactly equivalent to 2x leveraged intermediate treasuries. The yield curve is not a straight line, and they don't have the same correlation with equities. So maybe it's best if you explain your reasoning instead of me trying to guess.

Uncorrelated wrote: Mon Jan 20, 2020 3:25 pm It is completely pointless to construct a tangency portfolio unless you have free leverage. There is no convincing reason to optimize for maximum sharpe ratio unless you have free leverage.
Ok. What would be your recommended leveraged portfolio that keeps the 60/40 US/Intl equity exposure of Vanguard balanced funds ?

Uncorrelated wrote: Mon Jan 20, 2020 3:25 pm
danielc wrote:
Uncorrelated wrote: Mon Jan 20, 2020 10:17 am Is this your entire portfolio or just a part? It is not a good idea to buy leveraged funds if you are only using it as a part of your portfolio. If you have an emergency fund, you might want to eliminate that too.
Currently it is 0% of my portfolio, but I am currently thinking of putting maybe 10% of my portfolio in HEDGEFUNDIE's adventure portfolio (i.e. risky bet; might win big) and maybe 10-20% of my portfolio in something similar to what I showed here ---- a leveraged portfolio with lower volatility than then unleveraged portfolio that it replaces.
That's a really bad idea. Don't think of your portfolio in buckets, think of your portfolio as a whole. It is likely that your goals are better served by replacing some bonds of your existing portfolio with equities, usage of factor funds, reverse mortgage or replacing one part of your portfolio (say, US stocks) with leveraged equities and keeping the rest the same.
Up to this point I have always treated my entire portfolio as a whole. I don't do buckets, and I haven't made changes yet. Hedgefundie's adventure portfolio is a bit of a gamble and I definitely want to separate it from my main investment; so maybe that's a bucket, or maybe it's "play money".

But I am completely open to looking at everything else as a whole. Now, I don't have a house, and I don't have a lot of bonds that I can practically move to stocks. Most of my bonds are my emergency fund. Even if I had a lot of bonds, I am not sure that replacing bonds with stocks is better than leverage. Bonds (other than very short T-Bills) are themselves risky assets with a relatively low correlation with equities. An optimal portfolio should be the result of finding the right mix of stocks and bonds and then either add T-Bills or leverage to adjust the risk and return.
Last edited by danielc on Mon Jan 20, 2020 4:45 pm, edited 2 times in total.
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Re: Critique this leveraged portfolio

Post by danielc »

305pelusa wrote: Mon Jan 20, 2020 3:37 pm Whoah I couldn’t possibly agree more. It’s so amazing to me just how many BHs bucket off their portfolio, leading to inefficiencies like the one you mentioned. If you want more in stocks, don’t borrow from other. Sell your own bonds and borrow from yourself!
1) You and @Uncorrelated are both assuming that I am bucketing my portfolio. I never said that I was or that I plan to. I merely answered @Uncorrelated's question which is whether this is my entire portfolio. I said that it is not part of my portfolio at all, and that I'm thinking of making it maybe 10-20% of my portfolio. This does not mean that I was planning to "bucket" (I was not).

2) You overestimate how many bonds I have available to sell.

3) I don't agree at all that selling my own bonds is equivalent to "borrowing from myself". Bonds aren't cash, and they are not the risk-free asset. My money market account is similar to the risk free asset. Bonds are a risky asset with a different type of risk (i.e. term risk) than stocks. "It's so amazing how many BHs treat ``bonds`` or ``treasuries`` as the risk-free asset."
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Re: Critique this leveraged portfolio

Post by danielc »

305pelusa wrote: Mon Jan 20, 2020 4:09 pm Perhaps I’m wrong in my thinking OP but:

I have to say it’s a little strange to agree with all of the (dubious?) assumptions CAPM makes in order to conclude you should leverage the max Sharpe, but then CONVENIENTLY forget that under those very assumptions, the max Sharpe is already the market portfolio, not some 30/70 data mined concoction.
1) Your comment is needlessly antagonistic. Please be civil.

2) I could as easily say that you are "conveniently" forgetting that the market portfolio includes bonds too, and the bond market is larger than the stock market. The market portfolio is heavy on bonds.
MotoTrojan
Posts: 11259
Joined: Wed Feb 01, 2017 7:39 pm

Re: Critique this leveraged portfolio

Post by MotoTrojan »

Tracking error is bad on the non-US 3x funds. DZK was around 3% I believe. That’s a significant drag.

I’d also use EDV before TYD.
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Critique this leveraged portfolio

Post by Steve Reading »

danielc wrote: Mon Jan 20, 2020 4:42 pm
305pelusa wrote: Mon Jan 20, 2020 3:37 pm Whoah I couldn’t possibly agree more. It’s so amazing to me just how many BHs bucket off their portfolio, leading to inefficiencies like the one you mentioned. If you want more in stocks, don’t borrow from other. Sell your own bonds and borrow from yourself!
1) You and @Uncorrelated are both assuming that I am bucketing my portfolio. I never said that I was or that I plan to. I merely answered @Uncorrelated's question which is whether this is my entire portfolio. I said that it is not part of my portfolio at all, and that I'm thinking of making it maybe 10-20% of my portfolio. This does not mean that I was planning to "bucket" (I was not).

2) You overestimate how many bonds I have available to sell.

3) I don't agree at all that selling my own bonds is equivalent to "borrowing from myself". Bonds aren't cash, and they are not the risk-free asset. My money market account is similar to the risk free asset. Bonds are a risky asset with a different type of risk (i.e. term risk) than stocks. "It's so amazing how many BHs treat ``bonds`` or ``treasuries`` as the risk-free asset."
1) Making this just 10% of your portfolio is what I call a “bucket” (or mental accounting). So its just nomenclature here. This can lead to inefficiency.
2) Maybe.
3) A bond is literally you loaning money out to someone. Are you then also going to borrow? This might make sense in some circumstances sure. But I’m suggesting that instead of you lending money and then borrowing money, you cut out the two people and just “borrow” whatever you were planning on “lending”.
danielc wrote: Mon Jan 20, 2020 4:54 pm
305pelusa wrote: Mon Jan 20, 2020 4:09 pm Perhaps I’m wrong in my thinking OP but:

I have to say it’s a little strange to agree with all of the (dubious?) assumptions CAPM makes in order to conclude you should leverage the max Sharpe, but then CONVENIENTLY forget that under those very assumptions, the max Sharpe is already the market portfolio, not some 30/70 data mined concoction.
1) Your comment is needlessly antagonistic. Please be civil.

2) I could as easily say that you are "conveniently" forgetting that the market portfolio includes bonds too, and the bond market is larger than the stock market. The market portfolio is heavy on bonds.
1) Whoah you completely mistook my tone. I even added a “correct me if I’m wrong” because I’m not sure I’m correct in my thinking. I’m going to assume this is just a silly web misunderstanding.
2) First of all, I’m not forgetting that at all. The bond market is larger, yes but it’s not 30/70. More like 45/55 if I recall correctly.

Secondly, and most importantly, government debt (unlike corporate debt) is considered a zero-supply asset. Which means it can be printed at will with no claim to any asset and hence does not actually belong in the market portfolio. Corporate debt does have claims on real assets.

So the true market portfolio that would have max Sharpe based on the assumptions you are making, is really more like 70+% stocks.

Again, not saying any of this to antagonize you. Just trying to give you perspective.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Re: Critique this leveraged portfolio

Post by danielc »

305pelusa wrote: Mon Jan 20, 2020 5:29 pm 1) Making this just 10% of your portfolio is what I call a “bucket” (or mental accounting). So its just nomenclature here. This can lead to inefficiency.
Nonsense. What if instead of 2x leverage I want lower leverage? Then it's entirely reasonable to have x% of my portfolio in 2x leverage ETFs and (1-x%) in unleveraged funds. What matters is that when I make a pie chart of my portfolio (or any other kind of analysis) all the components of the portfolio are included together. Finally, for those of us who have the majority of our portfolio in a 401k or 403b, leveraging those portions is simply not an option, even if I could find ETFs that offered 1.1x leverage.

305pelusa wrote: Mon Jan 20, 2020 5:29 pm 3) A bond is literally you loaning money out to someone. Are you then also going to borrow? This might make sense in some circumstances sure. But I’m suggesting that instead of you lending money and then borrowing money, you cut out the two people and just “borrow” whatever you were planning on “lending”.
Borrowing at the risk-free rate and using the money to buy a 10-year treasury exposes you to the term premium in a way that holding cash does not. Depending on the situation, that could make sense for some people. Based on my example, and Hedgefundie's thread, and the thread about the WT 90/60 fund, it appears that a balanced fund with leverage can often give a better risk-adjusted return than what you get by increasing the stock allocation.

305pelusa wrote: Mon Jan 20, 2020 5:29 pm 1) Whoah you completely mistook my tone. I even added a “correct me if I’m wrong” because I’m not sure I’m correct in my thinking. I’m going to assume this is just a silly web misunderstanding.
You're probably right.
305pelusa wrote: Mon Jan 20, 2020 5:29 pm 2) First of all, I’m not forgetting that at all. The bond market is larger, yes but it’s not 30/70. More like 45/55 if I recall correctly.
For the US, I think it is 43/57 (based on a random article I found online) and globally it seems to be 33/67 (also based on a random article). So, sure, it's not quite 30/70. The main point of my point of my post was not 30/70 is perfect. I didn't put a lot of effort into tuning it; I just wanted to show an example of a leveraged portfolio that seems to be superior than another example of an unleveraged portfolio by every measure reported by PV.

305pelusa wrote: Mon Jan 20, 2020 5:29 pm Secondly, and most importantly, government debt (unlike corporate debt) is considered a zero-supply asset. Which means it can be printed at will with no claim to any asset and hence does not actually belong in the market portfolio. Corporate debt does have claims on real assets.
I have raised a similar point in other posts. You might want to debate @longinvest about that. William Sharpe treats government bonds as part of the market. My take is that this is an example of why "the market portfolio" should be taken with a big grain of sand when it comes to government debt, and why I think you shouldn't be arguing with me about whether 30/70 is the market portfolio or not.
User avatar
gordoni2
Posts: 213
Joined: Wed Aug 15, 2007 12:20 am
Contact:

Re: Critique this leveraged portfolio

Post by gordoni2 »

danielc wrote: Mon Jan 20, 2020 5:18 am According to MPT we should all be holding whichever portfolio has the highest Sharpe ratio, which in the US is around 30% stocks / 70% bonds, and then using leverage or cash to adjust the risk.
MPT assumes assets returns are normally distributed. This may be very approximately true for stocks, but isn't for bonds. Bonds returns do not follow a random walk but presumably exhibit a high degree of mean reversion and auto-correlation. I am thus not sure it is meaningful to construct a mean-variance plot using long run bond return statistics.

The graph showing 30/70 as optimal was for average returns over 1926-2017. Bond rates and the risk free rate were around 5% and 3% over this time period. They are much lower at present. Using values of 2% and 1% for bond returns and the risk free rate would drastically change the optimal point.

MPT assumes you can borrow at the risk free rate. You can't. Instead the capital allocation line is I think made up of two tangency lines and a connecting portion of the efficient frontier curve. Thus you don't necessarily want to hold the portfolio with the highest Sharpe ratio and simply apply leverage.

MPT is a single time period model. If you plan to be invested over a period of time you should be using a multi time period model that can handle the optimal spreading of risk over time. Merton's portfolio problem and stochastic dynamic programming are two examples of such models.
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Critique this leveraged portfolio

Post by Steve Reading »

danielc wrote: Mon Jan 20, 2020 6:35 pm
305pelusa wrote: Mon Jan 20, 2020 5:29 pm 1) Making this just 10% of your portfolio is what I call a “bucket” (or mental accounting). So its just nomenclature here. This can lead to inefficiency.
Nonsense. What if instead of 2x leverage I want lower leverage? Then it's entirely reasonable to have x% of my portfolio in 2x leverage ETFs and (1-x%) in unleveraged funds. What matters is that when I make a pie chart of my portfolio (or any other kind of analysis) all the components of the portfolio are included together. Finally, for those of us who have the majority of our portfolio in a 401k or 403b, leveraging those portions is simply not an option, even if I could find ETFs that offered 1.1x leverage.
Then that’s one circumstance when bucketing happened to work out correctly because your buckets are consistent with your desired total allocation. It’s correct by luck.

Many others do things like “10% dedicated to the Excellent Adventure, 25% to the permanent portfolio, etc”. Such a portfolio might be made more efficiently, for instance, by deleveraging the stocks from the Excellent Adventure and moving the cash from PP into stocks. Identical overall exposures but cheaper.

You get the deal.
danielc wrote: Mon Jan 20, 2020 6:35 pm
Borrowing at the risk-free rate and using the money to buy a 10-year treasury exposes you to the term premium in a way that holding cash does not. Depending on the situation, that could make sense for some people. Based on my example, and Hedgefundie's thread, and the thread about the WT 90/60 fund, it appears that a balanced fund with leverage can often give a better risk-adjusted return than what you get by increasing the stock allocation.
No I think you’re misunderstanding me. My point is ^ that particular example is that if you’re going to borrow money at the daily rate and lend it out at the 10 Y mark just to gain exposure to term, you might be able to do the same thing by just not borrowing and lending at a longer term.

IMO, trying to get into nitty gritty like “leverage Intermediate Term treasuries instead of just buying LTTs” starts to get kinda cute. I don’t buy there’s really that much to be gained from such specific, data mined strategies. And that’s all they are btw, data mined. I don’t think any one really could say why getting term exposure via leverage ITTs would be better than just buying longer term treasuries. It’s just that the data shows it and that’s good enough for some.
danielc wrote: Mon Jan 20, 2020 6:35 pm
For the US, I think it is 43/57 (based on a random article I found online) and globally it seems to be 33/67 (also based on a random article). So, sure, it's not quite 30/70. The main point of my point of my post was not 30/70 is perfect. I didn't put a lot of effort into tuning it; I just wanted to show an example of a leveraged portfolio that seems to be superior than another example of an unleveraged portfolio by every measure reported by PV.
Whoah I didn’t know the global scene was really that lopsided. Either way government bonds don’t really count so it’s much more equity heavy.

As to the portfolio, based on the fact that your OP has very specific %s and the thread is titled “Critique this portfolio”, I really thought you wanted feedback on that one. If all you’re presenting is the idea of leveraging a portfolio with high Sharpe (and aren’t too concerned about exact %s) then I think that’s reasonable.
danielc wrote: Mon Jan 20, 2020 6:35 pm
I have raised a similar point in other posts. You might want to debate @longinvest about that. William Sharpe treats government bonds as part of the market. My take is that this is an example of why "the market portfolio" should be taken with a big grain of sand when it comes to government debt, and why I think you shouldn't be arguing with me about whether 30/70 is the market portfolio or not.
If you’ve brought this point before, then you’re correct.There’s no grain of salt here. Government debt is not an asset to be taken into account into the Market Portfolio. If Sharpe truly said that, either he is just wrong, or maybe he was oversimplifying for viewers (like in that Andrew Lo interview).

In fact, in that interview, he even says something like “you should use the market portfolio. I even started doing this, I put some of my money last summer this way...”

I’m sorry what? Your whole thesis is the market portfolio William. I would think you’d have done this for decades, with your entire portfolio. Not recently, with some of your money. Oh well.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
UberGrub
Posts: 160
Joined: Wed Aug 21, 2019 3:47 pm

Re: Critique this leveraged portfolio

Post by UberGrub »

305pelusa wrote: Mon Jan 20, 2020 8:32 pm. Either way government bonds don’t really count so it’s much more equity heavy.
Wait so why exactly don’t they count?
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Critique this leveraged portfolio

Post by Steve Reading »

UberGrub wrote: Mon Jan 20, 2020 8:41 pm
305pelusa wrote: Mon Jan 20, 2020 8:32 pm. Either way government bonds don’t really count so it’s much more equity heavy.
Wait so why exactly don’t they count?
A zero net supply is just a transaction between two parties. There’s no actual asset in there. Think of derivatives. If you add up all the long and short positions, you get zero. The market portfolio shouldn’t contain derivatives. Just because more people decide to establish futures positions amongst each other (which are essentially bets) doesn’t mean the market is allocating capital there. It isn’t.

Commodities, stocks and corporate bonds are net positive supply. If you add up all long and short, you end up with a positive number of longs perfectly balanced by the assets it lays claim too. If a corporation released more bonds with no more infrastructure/capital/etc behind it, the value of the other bonds would drop commensurately so the capitalization of the bonds was the same before and after.

Government bonds are not like this. As they have no capitalized asset behind it, they’re just net zero supply. They only have a “capitalization” because you add up all the long’s but then conveniently leave out the debt (that the government owes). But if you capitalize stocks, that is literally what the companies themselves are worth. Hope that makes sense. Read about it!
Last edited by Steve Reading on Mon Jan 20, 2020 8:48 pm, edited 1 time in total.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
MotoTrojan
Posts: 11259
Joined: Wed Feb 01, 2017 7:39 pm

Re: Critique this leveraged portfolio

Post by MotoTrojan »

305pelusa wrote: Mon Jan 20, 2020 8:32 pm
IMO, trying to get into nitty gritty like “leverage Intermediate Term treasuries instead of just buying LTTs” starts to get kinda cute. I don’t buy there’s really that much to be gained from such specific, data mined strategies. And that’s all they are btw, data mined. I don’t think any one really could say why getting term exposure via leverage ITTs would be better than just buying longer term treasuries. It’s just that the data shows it and that’s good enough for some.
+1. I prefer to just pick a target volatility/duration and most efficiently achieve that with taxes, expenses, and volatility decay in mind. For example I would use EDV before TYD, but if I needed more volatility than EDV could provide I’d start to add TMF.
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Re: Critique this leveraged portfolio

Post by danielc »

305pelusa wrote: Mon Jan 20, 2020 8:32 pm
danielc wrote: Mon Jan 20, 2020 6:35 pm
305pelusa wrote: Mon Jan 20, 2020 5:29 pm 1) Making this just 10% of your portfolio is what I call a “bucket” (or mental accounting). So its just nomenclature here. This can lead to inefficiency.
Nonsense. What if instead of 2x leverage I want lower leverage? Then it's entirely reasonable to have x% of my portfolio in 2x leverage ETFs and (1-x%) in unleveraged funds. What matters is that when I make a pie chart of my portfolio (or any other kind of analysis) all the components of the portfolio are included together. Finally, for those of us who have the majority of our portfolio in a 401k or 403b, leveraging those portions is simply not an option, even if I could find ETFs that offered 1.1x leverage.
Then that’s one circumstance when bucketing happened to work out correctly because your buckets are consistent with your desired total allocation. It’s correct by luck.
It's not luck. I am informing you that my plan is to look at my overall portfolio. The limitations of my 401k force me to slice and dice my portfolio in ways that I wouldn't otherwise. That's not me making buckets; that's the govt and my employer and my 401k provider forcing my hand. But I can still choose to put everything in the same spreadsheet and make decisions about the whole, to the best of my ability.

Having said that, if I do implement hedgefundie's adventure portfolio, I do plan to bucket it because I think that rebalancing between it and the rest of the portfolio is too risky.

305pelusa wrote: Mon Jan 20, 2020 8:32 pm No I think you’re misunderstanding me. My point is ^ that particular example is that if you’re going to borrow money at the daily rate and lend it out at the 10 Y mark just to gain exposure to term, you might be able to do the same thing by just not borrowing and lending at a longer term.
Now I see your point. Maybe it is better to replace leveraged ITT with unleveraged LTT. I recently discovered EDV which has a very long duration. Has a mean duration of 25 years :shock: . Maybe I'll grab that. One advantage of your suggestion is a lower expense ratio, on top of the point you already made that borrowing costs money.

305pelusa wrote: Mon Jan 20, 2020 8:32 pm If you’ve brought this point before, then you’re correct.There’s no grain of salt here. Government debt is not an asset to be taken into account into the Market Portfolio. If Sharpe truly said that, either he is just wrong, or maybe he was oversimplifying for viewers (like in that Andrew Lo interview).
Yes, the comment I'm thinking of did come from the Andrew Lo interview. But in that comment, Sharpe was describing his own portfolio too. The question was along the lines of "so... you're the father of MPT... tell me, whats the ideal portfolio?".
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Re: Critique this leveraged portfolio

Post by danielc »

305pelusa wrote: Mon Jan 20, 2020 8:46 pm
UberGrub wrote: Mon Jan 20, 2020 8:41 pm Wait so why exactly don’t they count?
A zero net supply is just a transaction between two parties. There’s no actual asset in there. Think of derivatives. If you add up all the long and short positions, you get zero. The market portfolio shouldn’t contain derivatives. Just because more people decide to establish futures positions amongst each other (which are essentially bets) doesn’t mean the market is allocating capital there. It isn’t.

Commodities, stocks and corporate bonds are net positive supply. If you add up all long and short, you end up with a positive number of longs perfectly balanced by the assets it lays claim too. If a corporation released more bonds with no more infrastructure/capital/etc behind it, the value of the other bonds would drop commensurately so the capitalization of the bonds was the same before and after.

Government bonds are not like this. As they have no capitalized asset behind it, they’re just net zero supply. They only have a “capitalization” because you add up all the long’s but then conveniently leave out the debt (that the government owes). But if you capitalize stocks, that is literally what the companies themselves are worth. Hope that makes sense. Read about it!
That is very interesting. So a government bond is a bit like saying that my left pocket owes 10 bucks to my right pocket.
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Critique this leveraged portfolio

Post by Steve Reading »

danielc wrote: Mon Jan 20, 2020 9:17 pm
305pelusa wrote: Mon Jan 20, 2020 8:46 pm
UberGrub wrote: Mon Jan 20, 2020 8:41 pm Wait so why exactly don’t they count?
A zero net supply is just a transaction between two parties. There’s no actual asset in there. Think of derivatives. If you add up all the long and short positions, you get zero. The market portfolio shouldn’t contain derivatives. Just because more people decide to establish futures positions amongst each other (which are essentially bets) doesn’t mean the market is allocating capital there. It isn’t.

Commodities, stocks and corporate bonds are net positive supply. If you add up all long and short, you end up with a positive number of longs perfectly balanced by the assets it lays claim too. If a corporation released more bonds with no more infrastructure/capital/etc behind it, the value of the other bonds would drop commensurately so the capitalization of the bonds was the same before and after.

Government bonds are not like this. As they have no capitalized asset behind it, they’re just net zero supply. They only have a “capitalization” because you add up all the long’s but then conveniently leave out the debt (that the government owes). But if you capitalize stocks, that is literally what the companies themselves are worth. Hope that makes sense. Read about it!
That is very interesting. So a government bond is a bit like saying that my left pocket owes 10 bucks to my right pocket.
Yeah the only way to get a capitalization is if you only look at the left pocket in that example. You have to purposefully ignore the right. Same with derivatives, personal bets, student loans, etc. These aren't assets and aren't part of a capitalization-weighted market portfolio.

This is different than real estate, corporate bonds, stocks, etc. If you added up all the longs and shorts you actually do get a net positive long. And that net long IS the capitalization of the asset. So holding it at capitalization weight represents how the market is allocating capital.

Again, I know Sharpe has said it. I can only convince myself that he's just trying to simplify things a bit with basic Vanguard mutual funds (could you imagine getting into such technicalities in that interview?).
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
User avatar
Uncorrelated
Posts: 1237
Joined: Sun Oct 13, 2019 3:16 pm

Re: Critique this leveraged portfolio

Post by Uncorrelated »

danielc wrote: Mon Jan 20, 2020 9:10 pm It's not luck. I am informing you that my plan is to look at my overall portfolio. The limitations of my 401k force me to slice and dice my portfolio in ways that I wouldn't otherwise. That's not me making buckets; that's the govt and my employer and my 401k provider forcing my hand. But I can still choose to put everything in the same spreadsheet and make decisions about the whole, to the best of my ability.

Having said that, if I do implement hedgefundie's adventure portfolio, I do plan to bucket it because I think that rebalancing between it and the rest of the portfolio is too risky.
That is a mistake. The theory we are applying here is that of Merton's portfolio model, which is a continuous time variant of MPT (MPT optimizes over only a single horizon). According to Mertons portfolio model the optimal portfolio over time is a constant (if we assume constant relative risk aversion). In practice this means that if you allocate 10% to of your portfolio to this adventure and you don't rebalance (it becomes 5%, or 20%), then your portfolio is a violation of Merton's portfolio model.

There can be a good reason to stray from Merton's portfolio model, but in general (99% of the time) I find that it indicates a portfolio construction error.
Now I see your point. Maybe it is better to replace leveraged ITT with unleveraged LTT. I recently discovered EDV which has a very long duration. Has a mean duration of 25 years :shock: . Maybe I'll grab that. One advantage of your suggestion is a lower expense ratio, on top of the point you already made that borrowing costs money.
It is not a good idea to use EDV due to a bet-against-beta anomaly in bonds. The longer the duration goes, the less you get compensated for your risk. According to simba's bactesting spreadsheet, ITT7-index has a return of 1.87% per year (+ risk free) and standard deviation of 6.36%. EDV has a return of 3.23% (+ risk free rate) and standard deviation of 24.40%. That is twice the return for 4 times as much risk. According to the same spreadsheet 3x leveraged ITT has higher returns and lower risk than EDV.

I'm not a fan of the target duration approach. Not all durations are created equal. A portfolio with total bond market has a substantially higher sharpe ratio than a portfolio of t-bills and long-term bonds calibrated to the same duration exposure.
danielc wrote: Mon Jan 20, 2020 4:42 pm 1) You and @Uncorrelated are both assuming that I am bucketing my portfolio. I never said that I was or that I plan to. I merely answered @Uncorrelated's question which is whether this is my entire portfolio. I said that it is not part of my portfolio at all, and that I'm thinking of making it maybe 10-20% of my portfolio. This does not mean that I was planning to "bucket" (I was not).
That is what I'd call bucketing your portfolio. But what we call it is not important, the point is that it is impossible to judge your portfolio because you only gave a part of it.
danielc wrote: Mon Jan 20, 2020 4:30 pm
Uncorrelated wrote: Mon Jan 20, 2020 3:25 pm
danielc wrote: Mon Jan 20, 2020 1:16 pm None of the bonds are long bonds. They are 7-10 year treasuries.
I'm aware of that. You currently have the equivalent exposure of 50% 20-year treasuries and 50% equities, but it would be better to pick 80% equities and 20% long-term treasuries because equities have higher return per unit of risk. (specifically, around 3 times as high).
Can you show me how you derived those numbers? I guess that you are trying to work out the portfolio exposure to equity risk vs term risk. I figure that the leveraged portfolio has roughly the same ratio of equity risk and term risk as the 50/50 equity/LTT portfolio you suggest, but long treasuries aren't exactly equivalent to 2x leveraged intermediate treasuries. The yield curve is not a straight line, and they don't have the same correlation with equities. So maybe it's best if you explain your reasoning instead of me trying to guess.
I used a custom mean-variance optimizer to draw these conclusions, it uses data since 1934 or 1955. This optimizer consistently allocates the majority of it's funds to equities instead of bonds (total bond market or TMF).

The problem with leveraged ITT instead of leveraged LTT is that is it less space efficient. If you use ITT to get the same duration exposure as with LTT you must take on more leveraged funds on the equities side to keep the appropriate ratio of equities and bonds. However, leverage on equities is significantly more expensive than on bonds. Although I have not implemented leveraged ITT in my optimizer I suspect that the optimal allocation would be 0% everywhere along the efficient frontier. I have implemented EDV and unleveraged long term bonds, but my optimizer never chooses those.

If we are talking about non-leveraged funds, total bond market is the best choice in the vast majority of cases.
Uncorrelated wrote: Mon Jan 20, 2020 3:25 pm It is completely pointless to construct a tangency portfolio unless you have free leverage. There is no convincing reason to optimize for maximum sharpe ratio unless you have free leverage.
Ok. What would be your recommended leveraged portfolio that keeps the 60/40 US/Intl equity exposure of Vanguard balanced funds ?
I don't recommend any particular exposure. A 60/40 ratio is optimal for individuals with high risk aversion, but if you don't have high risk aversion then the optimal ratio is different from 60/40. If you add in leverage or factor funds, the the optimal ratio changes again. If you have human capital left, that also affects the optimal ratio.

I'm sorry that I'm not able to give you a simple answer, once you stray off the beaten path there are no simple answers. At this point I don't even have enough information to say whether you should be considering leveraged funds at all.
Uncorrelated wrote: Mon Jan 20, 2020 3:25 pm
danielc wrote:
Uncorrelated wrote: Mon Jan 20, 2020 10:17 am Is this your entire portfolio or just a part? It is not a good idea to buy leveraged funds if you are only using it as a part of your portfolio. If you have an emergency fund, you might want to eliminate that too.
Currently it is 0% of my portfolio, but I am currently thinking of putting maybe 10% of my portfolio in HEDGEFUNDIE's adventure portfolio (i.e. risky bet; might win big) and maybe 10-20% of my portfolio in something similar to what I showed here ---- a leveraged portfolio with lower volatility than then unleveraged portfolio that it replaces.
That's a really bad idea. Don't think of your portfolio in buckets, think of your portfolio as a whole. It is likely that your goals are better served by replacing some bonds of your existing portfolio with equities, usage of factor funds, reverse mortgage or replacing one part of your portfolio (say, US stocks) with leveraged equities and keeping the rest the same.
Up to this point I have always treated my entire portfolio as a whole. I don't do buckets, and I haven't made changes yet. Hedgefundie's adventure portfolio is a bit of a gamble and I definitely want to separate it from my main investment; so maybe that's a bucket, or maybe it's "play money".

But I am completely open to looking at everything else as a whole. Now, I don't have a house, and I don't have a lot of bonds that I can practically move to stocks. Most of my bonds are my emergency fund. Even if I had a lot of bonds, I am not sure that replacing bonds with stocks is better than leverage. Bonds (other than very short T-Bills) are themselves risky assets with a relatively low correlation with equities. An optimal portfolio should be the result of finding the right mix of stocks and bonds and then either add T-Bills or leverage to adjust the risk and return.
The optimal portfolio is not a fixed mix of stocks and bonds unless you have free leverage, which you do not have.

It might be useful to read lifecycle investing (in particular, read the paper) to figure out whether you want any leverage at all. Afterwards we can look at which combinations of funds and/or leverage is most suited for your purpose. Due to the complexities involved most folks seem to get their leverage on the equity side with options.
Cpadave
Posts: 278
Joined: Wed Nov 22, 2017 10:53 am

Re: Critique this leveraged portfolio

Post by Cpadave »

I lost more than half of net worth in 2008 investment in leveraged closed end fund. As the market went down, they had to sell out the stocks inside the fund. So when market finally recovered, I couldn’t even get back to where I was. I wished they just banned these types of investment vehicles. They are just tools for gambling.
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Re: Critique this leveraged portfolio

Post by danielc »

Uncorrelated wrote: Tue Jan 21, 2020 5:36 am
danielc wrote: Now I see your point. Maybe it is better to replace leveraged ITT with unleveraged LTT. I recently discovered EDV which has a very long duration. Has a mean duration of 25 years :shock: . Maybe I'll grab that. One advantage of your suggestion is a lower expense ratio, on top of the point you already made that borrowing costs money.
It is not a good idea to use EDV due to a bet-against-beta anomaly in bonds. The longer the duration goes, the less you get compensated for your risk. According to simba's bactesting spreadsheet, ITT7-index has a return of 1.87% per year (+ risk free) and standard deviation of 6.36%. EDV has a return of 3.23% (+ risk free rate) and standard deviation of 24.40%. That is twice the return for 4 times as much risk. According to the same spreadsheet 3x leveraged ITT has higher returns and lower risk than EDV.
Thanks. How likely is it that a leveraged ITT gets to borrow a rate close enough to the risk free rate that 3xITT still comes out ahead compared to EDV. Let's see... I need:

(1.87 + (1.87 + risk_free_rate - margin_rate)*2) / (6.36*3) > 3.23 / 24.40

If I got that formula right, then historically I would have needed

margin_rate < risk_free_rate + 1.542

That doesn't look too hard for a futures contract. Today the 3-month LIBOR is 1.84 and I believe I read somewhere that leveraged ETFs pay something close to that.

Uncorrelated wrote: Tue Jan 21, 2020 5:36 am I'm not a fan of the target duration approach. Not all durations are created equal. A portfolio with total bond market has a substantially higher sharpe ratio than a portfolio of t-bills and long-term bonds calibrated to the same duration exposure
Thanks for saying that. I was feeling uncomfortable with 305pelusa's idea of target duration because, for example, the yield curve is not a straight line, and not all bonds have the same convexity.

Uncorrelated wrote: Tue Jan 21, 2020 5:36 am That is what I'd call bucketing your portfolio. But what we call it is not important, the point is that it is impossible to judge your portfolio because you only gave a part of it.
I am not looking for help with my portfolio. I am looking for a critique of the idea of replacing a simple balanced portfolio, like VASGX, with a leveraged portfolio that seems to have higher return and less risk. If the idea is sound, that might mean that a lot of people using 3-fund or 4-fund BH portfolios could be better off with some leveraged portfolio instead. One of the fruits of that discussion has been the discussion of the merits of leveraged bonds vs extended duration bonds. @305pelusa seems very opposed to borrowing and suggested that one use longer duration bonds instead, and you have made a good argument against that idea. At the same time, you said earlier that the sample portfolio has a bond exposure that's too high.

Uncorrelated wrote: Mon Jan 20, 2020 3:25 pm The problem with leveraged ITT instead of leveraged LTT is that is it less space efficient. If you use ITT to get the same duration exposure as with LTT you must take on more leveraged funds on the equities side to keep the appropriate ratio of equities and bonds.
What is the appropriate ratio of equities and bonds, and why? Most people tune their stock/bond allocation as the main way to adjust the risk of the portfolio. Are they wrong to do that?
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Critique this leveraged portfolio

Post by willthrill81 »

klaus14 wrote: Mon Jan 20, 2020 12:21 pm Do you monitor stock-bond correlations? Will you change your alloc as correlations change?
That's the advantage of a target volatility approach: it adjusts allocations based on recent volatility and not volatility from years prior (although you could use long-term volatility, I'm not aware of anyone who does or why they would).
The Sensible Steward
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Critique this leveraged portfolio

Post by Steve Reading »

danielc wrote: Tue Jan 21, 2020 10:56 am
Uncorrelated wrote: Tue Jan 21, 2020 5:36 am I'm not a fan of the target duration approach. Not all durations are created equal. A portfolio with total bond market has a substantially higher sharpe ratio than a portfolio of t-bills and long-term bonds calibrated to the same duration exposure
Thanks for saying that. I was feeling uncomfortable with 305pelusa's idea of target duration because, for example, the yield curve is not a straight line, and not all bonds have the same convexity.
Well enlighten me then. I’m here to learn as much as the next guy.

I get that the curve isn’t straight. So there would seem to be easy money in just looking for the tangent bond in the curve and then leveraging/mixing with cash ala CAPM to achieve your desired duration, with higher return.

This is something every retail investor could do with Futures. So either it’s a free lunch for everyone (and somehow hasn’t been arbitraged away even though literally any one can) OR there’s some additional risk aside from volatility (which comes from duration) that creates the anomaly.

I’m no bond expert so I lean towards the latter. Care to explain? Cause in this subject, all I see are people simulating portfolios from the past and finding the best choice. That’s nowhere near good enough to pass my sniff test. You seem knowledgeable, how does it pass yours?
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Critique this leveraged portfolio

Post by Steve Reading »

danielc wrote: Tue Jan 21, 2020 10:56 am@305pelusa seems very opposed to borrowing and suggested that one use longer duration bonds instead, and you have made a good argument against that idea. At the same time, you said earlier that the sample portfolio has a bond exposure that's too high.
BTW, I’m not opposed to borrowing (for stocks). I do it and a lot. I’mjust wary of these ideas of borrowing for bonds. It’s borrowing to lend money. I know you can abstract assets into neat black boxes with behaviors and just forget what they actually are but common sense gives me pause.

Either way, to be clear, I recommend buying and holding long term bonds to maturity. This is effectively riskless. In fact the long term bond held to maturity IS the riskless asset for the long term investor. What Uncorrolated simulates are portfolios of bond funds that rebalance duration to maintain a constant duration (like BND or EDV). As an ardent supporter of bond immunization, that’s just not how I would do my bonds. So there’s a bit of context you were missing with my recommendation.

Also, you’re saying Uncorrolated has given good arguments against what I’m saying? Huh, name one good argument that isn’t “well in the past it’s been better”. I’m looking for some intuition here not just backtesting.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
langlands
Posts: 1093
Joined: Wed Apr 03, 2019 10:05 pm

Re: Critique this leveraged portfolio

Post by langlands »

@Uncorrelated

Regarding your point of reminding us that retail investors don't have access to "free leverage," is the idea very roughly that a less risk-averse investor should lean a bit towards taking on more risky tilts instead of relying entirely on leverage (since there's an associated cost) to get the extra return? So it might make more sense to take a 100% stock portfolio rather than a levered up 80/20 stock/bond portfolio. If I recall correctly, you also suggested factor tilts before taking on more leverage (if you believe in that sort of thing).

I'm a little confused about where you stand exactly regarding bonds. It seems you dislike long term treasuries because of the low beta anomaly (bad risk reward proposition). You also say:
The problem with leveraged ITT instead of leveraged LTT is that is it less space efficient. If you use ITT to get the same duration exposure as with LTT you must take on more leveraged funds on the equities side to keep the appropriate ratio of equities and bonds.
I don't understand why you'd need to "take on more leveraged funds on the equities side to keep the appropriate ratio of equities and bonds." Isn't the duration exposure effectively a measure of how much bonds you have, just as market beta is a metric of how much equity exposure you have (certainly if I own a utilities ETF for instance, I count that as less equity exposure than if I owned a tech ETF)

Under what circumstances does your optimizer recommend bonds? Basically I'm trying to get an intuitive feel for what bonds can add to the portfolio of a rather risk seeking investor.
User avatar
Topic Author
danielc
Posts: 1079
Joined: Sun Dec 10, 2017 3:48 am
Location: Iowa, USA
Contact:

Re: Critique this leveraged portfolio

Post by danielc »

305pelusa wrote: Tue Jan 21, 2020 11:22 am BTW, I’m not opposed to borrowing (for stocks). I do it and a lot. I’mjust wary of these ideas of borrowing for bonds. It’s borrowing to lend money. I know you can abstract assets into neat black boxes with behaviors and just forget what they actually are but common sense gives me pause.
I am not a bond expert. Someone in the forum once said that long term bonds are not a good deal for most individual investors because that market is dominated by pension funds with very long term liabilities, and the risk/reward calculation is different for them than it is for regular people. I don't know if that is true, but I do know that the yield curve is very rarely a simple straight line, and there are even bond strategies like "rolling down the yield curve" that rely on that fact.
305pelusa wrote: Tue Jan 21, 2020 11:22 am Either way, to be clear, I recommend buying and holding long term bonds to maturity. This is effectively riskless.
What you said feels intuitive, and I confess that I often think of bonds that way. But the Bogleheads wiki explicitly says this is an error in thinking. I kind of understand the argument, but I've never been completely convinced.
305pelusa wrote: Tue Jan 21, 2020 11:22 am In fact the long term bond held to maturity IS the riskless asset for the long term investor.
A rolling short term bonds should have less inflation risk than a long term bond.

305pelusa wrote: Tue Jan 21, 2020 11:22 am Also, you’re saying Uncorrolated has given good arguments against what I’m saying? Huh, name one good argument that isn’t “well in the past it’s been better”. I’m looking for some intuition here not just backtesting.
That would require that I give a good argument why the yield curve is usually not a straight line. I'm not sure I can do that. But if you were to accept the premise that most often the yield curve has a negative curvature, it would follow that there must be a positive interest rate at which it is better to leverage intermediate bonds than to buy long bonds.
ThrustVectoring
Posts: 771
Joined: Wed Jul 12, 2017 2:51 pm

Re: Critique this leveraged portfolio

Post by ThrustVectoring »

You're paying a lot more for leverage than you should need to spend. Put 60% of your portfolio in straight equities instead of 30% in 2x levered equities, and purchase Treasury futures contracts for leverage. These sorts of futures contracts are what the 2x daily ETFs are using anyhow, they're just marking things up by something like 90 basis points.
Current portfolio: 60% VTI / 40% VXUS
EfficientInvestor
Posts: 580
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: Critique this leveraged portfolio

Post by EfficientInvestor »

305pelusa wrote: Tue Jan 21, 2020 11:22 am BTW, I’m not opposed to borrowing (for stocks). I do it and a lot. I’mjust wary of these ideas of borrowing for bonds. It’s borrowing to lend money. I know you can abstract assets into neat black boxes with behaviors and just forget what they actually are but common sense gives me pause.

Either way, to be clear, I recommend buying and holding long term bonds to maturity. This is effectively riskless. In fact the long term bond held to maturity IS the riskless asset for the long term investor. What Uncorrolated simulates are portfolios of bond funds that rebalance duration to maintain a constant duration (like BND or EDV). As an ardent supporter of bond immunization, that’s just not how I would do my bonds. So there’s a bit of context you were missing with my recommendation.

Also, you’re saying Uncorrolated has given good arguments against what I’m saying? Huh, name one good argument that isn’t “well in the past it’s been better”. I’m looking for some intuition here not just backtesting.
In regards to looking for intuition that doesn't rely on backtesting...have you tried to think about leveraged bonds strictly in terms of portfolio insurance? Let's say you would otherwise be 100% stock. Instead, you take 10% of that and put it towards leveraged bonds in an amount that you think would help balance the 90% stock exposure. Let's just stick with the 30/70 example from the original post and say the bonds used are the 5 year treasury futures. So that means you would have 210% in bonds. Feel free to ramp that up or down based on expected volatility of 5-year treasuries, but I think that's close enough for now. If you had a $100k portfolio, you would have $90k in stock ETFs and you would buy 2 contracts (about $210k notional) of /ZF (total maintenance margin of $1,400). Under normal circumstances, you would earn a bit of an ongoing spread by borrowing at the 3-month to invest in the 5-year. Therefore, you are getting some insurance that pays for itself and earning some return on the 10% that isn't in stocks.

However, that's not the case at the moment due to the inversion. The current rate on the 5-year is 1.57% and the current 3-month LIBOR is about 1.84%. However, this doesn't mean it's a bad idea to still purchase the insurance. You are just having to pay out some premium at the moment to get the insurance. In the event of a downturn in stocks, it is likely that interest rates will go down further and you will wish you had some bond exposure. The other issue to consider is what happens if rates rise and your insurance works against you. As long as rates don't rise too quickly, it shouldn't hurt you too much.

However, if rates rise quickly, it would sting the bonds badly (this is one reason why I actually prefer the shorter duration 2-year treasuries over the 5, but I'm using 5 for the sake of this example). Therefore, given that inflation is the main thing that would cause rates to rise so quickly as to make the bonds suffer badly, you would also want to use some of the 10% you set aside to purchase some inflation insurance. This is where a gold futures contract could be handy. Perhaps 2 contracts of /MGC which would have a notional value of around $31k and require margin of $1,000.

When it's all said and done, you still have 90% exposure to stocks and the other 10% is invested in some "insurance" in a way that may work against you in any shorter time period (especially when stocks are outperforming expectation) but should at least have a positive expected return over a longer period of time. In other words, you aren't just using 10% of your portfolio to pay insurance premium.
Last edited by EfficientInvestor on Tue Jan 21, 2020 1:33 pm, edited 2 times in total.
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Critique this leveraged portfolio

Post by Steve Reading »

danielc wrote: Tue Jan 21, 2020 12:39 pm
305pelusa wrote: Tue Jan 21, 2020 11:22 am BTW, I’m not opposed to borrowing (for stocks). I do it and a lot. I’mjust wary of these ideas of borrowing for bonds. It’s borrowing to lend money. I know you can abstract assets into neat black boxes with behaviors and just forget what they actually are but common sense gives me pause.
I am not a bond expert. Someone in the forum once said that long term bonds are not a good deal for most individual investors because that market is dominated by pension funds with very long term liabilities, and the risk/reward calculation is different for them than it is for regular people. I don't know if that is true, but I do know that the yield curve is very rarely a simple straight line, and there are even bond strategies like "rolling down the yield curve" that rely on that fact.
Meh I don’t really buy that. First that assumes those institutions can’t leverage shorter term bonds for the same purpose. Since those portfolios are managed by asset managers, I’m going to assume they can.

And even if they can’t, the bond market is the most efficient market in this planet. So efficient even Swensen just buys index funds here. I simply refuse to believe there’s some egregious inneficiency in the yield curve like that, where even a retail investor could exploit.
“Rolling down the yield curve”, I thought wasn’t actually a strategy that works consistently (just like other Wall Street axioms like “cut your losers short” or “sell in May and go away”).
danielc wrote: Tue Jan 21, 2020 12:39 pm
What you said feels intuitive, and I confess that I often think of bonds that way. But the Bogleheads wiki explicitly says this is an error in thinking. I kind of understand the argument, but I've never been completely convinced.
No, there’s no error in thinking here and we’re both correct. Read the wiki again. A zero-coupon treasury bond is completely risk free for an investor that buys and holds. No risk of capital and no uncertainty on what return you’ll get. The wiki agrees.

However, if the bonds throws coupons, then you have reinvestments. I wrote that on my post 0_o. In high rate environments, the reinvestment could be significant. We’re in a pretty low rate environment so not as significant. Still it isn’t perfect like a zero-coupon. The duration doesn’t decrease linearly.

But a bond fund (or the simulations Uncorrolated does) are completely different because they maintain a somewhat constant duration. At the very least buying and holding a coupon bearing bond with a low coupon (like nowadays) is much closer in behavior to a STRIP than a bond fund.

I recommend investors buy and hold LT bonds if they have a LT horizon. OR use bond funds but make sure you sell a little and keep more in cash every year to make it look like a bond. These aren’t things Uncorrolated is simulating, but it is precisely what makes the LT bond effectively risk free (or close).
danielc wrote: Tue Jan 21, 2020 12:39 pm
That would require that I give a good argument why the yield curve is usually not a straight line. I'm not sure I can do that. But if you were to accept the premise that most often the yield curve has a negative curvature, it would follow that there must be a positive interest rate at which it is better to leverage intermediate bonds than to buy long bonds.
Huh? No no. The yield curve is unequivocally not straight. I know that. So it would appear there’s a free lunch here. I claim there isn’t and that there must be something else (be it convexity, or market expectations in the exact time period, or whatever else) that is unaccounted for by simply looking at duration and volatility.

I’m no bond expert so I can’t figure it out. But I just don’t see how it could be any other way.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Critique this leveraged portfolio

Post by Steve Reading »

EfficientInvestor wrote: Tue Jan 21, 2020 1:24 pm
305pelusa wrote: Tue Jan 21, 2020 11:22 am BTW, I’m not opposed to borrowing (for stocks). I do it and a lot. I’mjust wary of these ideas of borrowing for bonds. It’s borrowing to lend money. I know you can abstract assets into neat black boxes with behaviors and just forget what they actually are but common sense gives me pause.

Either way, to be clear, I recommend buying and holding long term bonds to maturity. This is effectively riskless. In fact the long term bond held to maturity IS the riskless asset for the long term investor. What Uncorrolated simulates are portfolios of bond funds that rebalance duration to maintain a constant duration (like BND or EDV). As an ardent supporter of bond immunization, that’s just not how I would do my bonds. So there’s a bit of context you were missing with my recommendation.

Also, you’re saying Uncorrolated has given good arguments against what I’m saying? Huh, name one good argument that isn’t “well in the past it’s been better”. I’m looking for some intuition here not just backtesting.
In regards to looking for intuition that doesn't rely on backtesting...have you tried to think about leveraged bonds strictly in terms of portfolio insurance? Let's say you would otherwise be 100% stock. Instead, you take 10% of that and put it towards leveraged bonds in an amount that you think would help balance the 90% stock exposure. Let's just stick with the 30/70 example from the original post and say the bonds used are the 5 year treasury futures. So that means you would have 210% in bonds. Feel free to ramp that up or down based on expected volatility of 5-year treasuries, but I think that's close enough for now. If you had a $100k portfolio, you would have $90k in stock ETFs and you would buy 2 contracts (about $210k notional) of /ZF (total maintenance margin of $1,400). Under normal circumstances, you would earn a bit of an ongoing spread by borrowing at the 3-month to invest in the 5-year. Therefore, you are getting some insurance that pays for itself and earning some return on the 10% that isn't in stocks.

However, that's not the case at the moment due to the inversion. The current rate on the 5-year is 1.57% and the current 3-month LIBOR is about 1.84%. However, this doesn't mean it's a bad idea to still purchase the insurance. You are just having to pay out some premium at the moment to get the insurance. In the event of a downturn in stocks, it is likely that interest rates will go down further and you will wish you had some bond exposure. The other issue to consider is what happens if rates rise and your insurance works against you. As long as rates don'r rise too quickly, it shouldn't hurt you too much.

However, if rates rise quickly, it would sting the bonds badly (this is one reason why I actually prefer the shorter duration 2-year treasuries over the 5, but I'm using 5 for the sake of this example). Therefore, given that inflation is the main thing that would cause rates to rise so quickly as to make the bonds suffer badly, you would also want to use some of the 10% you set aside to purchase some inflation insurance. This is where a gold futures contract could be handy. Perhaps 2 contracts of /MGC which would have a notional value of around $31k and require margin of $1,000.

When it's all said and done, you still have 90% exposure to stocks and the other 10% is invested in some "insurance" in a way that may work against you in any shorter time period (especially when stocks are outperforming expectation) but should at least have a positive expected return over a longer period of time. In other words, you aren't just using 10% of your portfolio to pay insurance premium.
What you said makes sense but I have a hard time connecting it with what I’m actually asking.

I’m saying if your horizon is 20 years, buy yourself a a bond with a duration of approximately 20 years. That’s bond immunization 101.

What Uncorrolated and others are saying is “instead leverage 10Y bonds x2. Same duration, same capital, but higher risk adjusted returns historically”. The only explanation I’ve heard why that’s better is that historically it’s been better.

The only thing close to intuition as to why that should be better is that LT bonds are a “bet against beta” as Uncorrolated put it. Perhaps he’ll expand on that.

My logic is that if leveraging x2 10Y bonds (or x10, 2Y bonds or whatever else) is produces higher risk adjusted returns, then the market would short the 20Y and buy up the 10Y or 2Y.

We see the yield curve isn’t linear. So either the market is inefficient and leaving this excellent opportunity to any one who wants (even retail can trade futures) or there is some OTHER risk of leveraging the bonds that simply isn’t represented by yield curve or duration (but the market is pricing in regardless). Something hedge funds and money managers know but us mere mortals don’t. The latter makes more sense to me.

With this said, was your post answering that^? Because if so I kinda didn’t follow 0_o
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
EfficientInvestor
Posts: 580
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: Critique this leveraged portfolio

Post by EfficientInvestor »

305pelusa wrote: Tue Jan 21, 2020 1:39 pm
EfficientInvestor wrote: Tue Jan 21, 2020 1:24 pm
305pelusa wrote: Tue Jan 21, 2020 11:22 am BTW, I’m not opposed to borrowing (for stocks). I do it and a lot. I’mjust wary of these ideas of borrowing for bonds. It’s borrowing to lend money. I know you can abstract assets into neat black boxes with behaviors and just forget what they actually are but common sense gives me pause.

Either way, to be clear, I recommend buying and holding long term bonds to maturity. This is effectively riskless. In fact the long term bond held to maturity IS the riskless asset for the long term investor. What Uncorrolated simulates are portfolios of bond funds that rebalance duration to maintain a constant duration (like BND or EDV). As an ardent supporter of bond immunization, that’s just not how I would do my bonds. So there’s a bit of context you were missing with my recommendation.

Also, you’re saying Uncorrolated has given good arguments against what I’m saying? Huh, name one good argument that isn’t “well in the past it’s been better”. I’m looking for some intuition here not just backtesting.
In regards to looking for intuition that doesn't rely on backtesting...have you tried to think about leveraged bonds strictly in terms of portfolio insurance? Let's say you would otherwise be 100% stock. Instead, you take 10% of that and put it towards leveraged bonds in an amount that you think would help balance the 90% stock exposure. Let's just stick with the 30/70 example from the original post and say the bonds used are the 5 year treasury futures. So that means you would have 210% in bonds. Feel free to ramp that up or down based on expected volatility of 5-year treasuries, but I think that's close enough for now. If you had a $100k portfolio, you would have $90k in stock ETFs and you would buy 2 contracts (about $210k notional) of /ZF (total maintenance margin of $1,400). Under normal circumstances, you would earn a bit of an ongoing spread by borrowing at the 3-month to invest in the 5-year. Therefore, you are getting some insurance that pays for itself and earning some return on the 10% that isn't in stocks.

However, that's not the case at the moment due to the inversion. The current rate on the 5-year is 1.57% and the current 3-month LIBOR is about 1.84%. However, this doesn't mean it's a bad idea to still purchase the insurance. You are just having to pay out some premium at the moment to get the insurance. In the event of a downturn in stocks, it is likely that interest rates will go down further and you will wish you had some bond exposure. The other issue to consider is what happens if rates rise and your insurance works against you. As long as rates don'r rise too quickly, it shouldn't hurt you too much.

However, if rates rise quickly, it would sting the bonds badly (this is one reason why I actually prefer the shorter duration 2-year treasuries over the 5, but I'm using 5 for the sake of this example). Therefore, given that inflation is the main thing that would cause rates to rise so quickly as to make the bonds suffer badly, you would also want to use some of the 10% you set aside to purchase some inflation insurance. This is where a gold futures contract could be handy. Perhaps 2 contracts of /MGC which would have a notional value of around $31k and require margin of $1,000.

When it's all said and done, you still have 90% exposure to stocks and the other 10% is invested in some "insurance" in a way that may work against you in any shorter time period (especially when stocks are outperforming expectation) but should at least have a positive expected return over a longer period of time. In other words, you aren't just using 10% of your portfolio to pay insurance premium.
What you said makes sense but I have a hard time connecting it with what I’m actually asking.

I’m saying if your horizon is 20 years, buy yourself a a bond with a duration of approximately 20 years. That’s bond immunization 101.

What Uncorrolated and others are saying is “instead leverage 10Y bonds x2. Same duration, same capital, but higher risk adjusted returns historically”. The only explanation I’ve heard why that’s better is that historically it’s been better.

The only thing close to intuition as to why that should be better is that LT bonds are a “bet against beta” as Uncorrolated put it. Perhaps he’ll expand on that.

My logic is that if leveraging x2 10Y bonds (or x10, 2Y bonds or whatever else) is produces higher risk adjusted returns, then the market would short the 20Y and buy up the 10Y or 2Y.

We see the yield curve isn’t linear. So either the market is inefficient and leaving this excellent opportunity to any one who wants (even retail can trade futures) or there is some OTHER risk of leveraging the bonds that simply isn’t represented by yield curve or duration (but the market is pricing in regardless). Something hedge funds and money managers know but us mere mortals don’t. The latter makes more sense to me.

With this said, was your post answering that^? Because if so I kinda didn’t follow 0_o
Whoops...I missed the mark. I thought you were asking for intuition about using leveraged bonds that doesn't rely on a backtest. If you are willing to change subjects slightly and provide your thoughts on what I have said, I would be curious to know your thoughts on this idea of viewing bonds and commodities as insurance.

As for the question at hand...I agree that holding a zero coupon bond has no default or reinvestment risk. However, the reinvestment risk swings both ways. In the event of inflation and rising rates, I would rather have more leverage on shorter term treasuries as opposed to no leverage on longer term because you can now reinvest at higher rates.

EDIT - I'll also add that I have read up on the idea of leverage aversion and how you can do a little better by leveraging shorter term bonds instead of buying longer term bonds outright. This could potentially be due to the fact that many institutional investors (e.g. pension funds) are not allowed to use leverage. Therefore, if they want duration, they are forced to go further out on the curve. I wish I could prove this theory with math, but it sounds good enough in concept. I figure that by using the leverage on the shorter term, I can potentially take advantage of this edge, if it exists. However, by using leverage, I'm also subject to more borrowing costs that may exist. For instance, I am currently investing with a negative spread which is not sustainable over the long term. Perhaps the best bet is to just do and get exposure all along the yield curve. Anything beyond that is either cherry picking based on past data or active management based on whatever algorithm you can come up with. All that said...I am still on the short end due to the long history of overperformance, ability to reinvest as rates are rising, and the whole leverage aversion theory.
Last edited by EfficientInvestor on Tue Jan 21, 2020 2:05 pm, edited 1 time in total.
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Critique this leveraged portfolio

Post by Steve Reading »

EfficientInvestor wrote: Tue Jan 21, 2020 1:50 pm
Whoops...I missed the mark. I thought you were asking for intuition about using leveraged bonds that doesn't rely on a backtest.
Yes that’s effectively what I’m asking. But I’m not asking why have bonds at all. Rather why that 10% you dedicated to bonds was in leveraged ST or IT bonds and not simply unleveraged LT bonds.
EfficientInvestor wrote: Tue Jan 21, 2020 1:50 pm
As for the question at hand...I agree that holding a zero coupon bond has no default or reinvestment risk. However, the reinvestment risk swings both ways. In the event of inflation and rising rates, I would rather have more leverage on shorter term treasuries as opposed to no leverage on longer term because you can now reinvest at higher rates.
First of all, if the inflation causes a parallel shift in the yield curve, then the zero coupon bond would produce the same losses as the leveraged STTs provided they have similar duration (which is what we’re assuming here).

But if anything, a rise in inflation typically hits the short end of the curve more than the long term no? I would think holding 2Y bonds leveraged x10 would produce more losses from an inflationary increase than just a 20Y bond.

I’m actually not totally sure, just speculating here.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
EfficientInvestor
Posts: 580
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: Critique this leveraged portfolio

Post by EfficientInvestor »

305pelusa wrote: Tue Jan 21, 2020 2:00 pm
EfficientInvestor wrote: Tue Jan 21, 2020 1:50 pm
Whoops...I missed the mark. I thought you were asking for intuition about using leveraged bonds that doesn't rely on a backtest.
Yes that’s effectively what I’m asking. But I’m not asking why have bonds at all. Rather why that 10% you dedicated to bonds was in leveraged ST or IT bonds and not simply unleveraged LT bonds.
EfficientInvestor wrote: Tue Jan 21, 2020 1:50 pm
As for the question at hand...I agree that holding a zero coupon bond has no default or reinvestment risk. However, the reinvestment risk swings both ways. In the event of inflation and rising rates, I would rather have more leverage on shorter term treasuries as opposed to no leverage on longer term because you can now reinvest at higher rates.
First of all, if the inflation causes a parallel shift in the yield curve, then the zero coupon bond would produce the same losses as the leveraged STTs provided they have similar duration (which is what we’re assuming here).

But if anything, a rise in inflation typically hits the short end of the curve more than the long term no? I would think holding 2Y bonds leveraged x10 would produce more losses from an inflationary increase than just a 20Y bond.

I’m actually not totally sure, just speculating here.
Check out the edit I made in my previous post. This may help answer some of this.
UberGrub
Posts: 160
Joined: Wed Aug 21, 2019 3:47 pm

Re: Critique this leveraged portfolio

Post by UberGrub »

It would seem to me like it’s an anomaly similar to the BAB phenomenon. Low volatility stocks have higher sharpe so leveraging them produces higher returns at the same risk as high beta stocks.

The Low Vol phenomenon occurs from barriers to leverage which makes people buy riskier stocks instead of just leveraging (among other things). But it appears mostly a behavioral, not a risk factor.

I agree it’d be kind of weird to see it in the bond market too but they’re both supposedly efficient markets.

TLDR: Maybe it’s just an anomaly that hasn’t quite been arbitraged.
Post Reply