Lifecycle Investing - Leveraging when young

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
guyinlaw
Posts: 94
Joined: Wed Jul 03, 2019 9:54 am

Re: Lifecycle Investing - Leveraging when young

Post by guyinlaw » Thu Nov 07, 2019 2:32 pm

305pelusa wrote:
Thu Nov 07, 2019 12:19 pm

Have you even read the paper ?
Please add the authors website in the original post. The previous links seems to be dead.

http://www.lifecycleinvesting.net/

User avatar
Topic Author
305pelusa
Posts: 903
Joined: Fri Nov 16, 2018 10:20 pm

Re: Lifecycle Investing - Leveraging when young

Post by 305pelusa » Thu Nov 07, 2019 2:53 pm

guyinlaw wrote:
Thu Nov 07, 2019 2:32 pm
305pelusa wrote:
Thu Nov 07, 2019 12:19 pm

Have you even read the paper ?
Please add the authors website in the original post. The previous links seems to be dead.

http://www.lifecycleinvesting.net/
True. Will update the op with new link.

I didn't find their website tremendously useful personally. But the paper is truly excellent.

https://poseidon01.ssrn.com/delivery.ph ... 23&EXT=pdf

Uncorrelated
Posts: 105
Joined: Sun Oct 13, 2019 3:16 pm

Re: Lifecycle Investing - Leveraging when young

Post by Uncorrelated » Fri Nov 08, 2019 6:41 am

Lee_WSP wrote:
Wed Nov 06, 2019 3:36 pm
I think he may well be fine because nearly all his leverage is in an un-callable family loan. But to do this with margin would probably be violating the Kelly Criterion.
The kelly criterion is only relevant if you have a logarithmic utility function (constant risk aversion coefficient of 1). This is almost never the case.

See the third image in my recent thread for evidence it is rational to employ higher leverage than the maximum one suggested by the kelly criterion:
viewtopic.php?f=10&t=293469

305pelusa might also be interested in reading this. I get where you are coming from when you are targeting a lifetime constant equity exposure, but that doesn't look very optimal to me. The paper on lifecycle investing and leveraging uses a simple model. The paper from Forsyth and Vetzal, "Robust Asset Allocation for Long-Term Target-Based Investing" does something similar, but in a continuous model with continuously updating asset allocation. My model (see thread above) is discrete but allows for the specification of arbitrary utility functions. Both models reach the same conclusion.

Uncorrelated
Posts: 105
Joined: Sun Oct 13, 2019 3:16 pm

Re: Lifecycle Investing - Leveraging when young

Post by Uncorrelated » Fri Nov 08, 2019 10:13 am

Just wanted to add a few thoughts to this discussion.
EfficientInvestor wrote:
Wed Nov 06, 2019 2:21 pm
Note that my backtests below for Portfolio 2 just uses the oldest bond mutual fund I could find a ticker symbol for. I implement my actual strategy with treasury futures.
Unfortunately that is not nearly enough data. See this chart from ERN in particular:

Image

Since 1980, bonds have done nothing but go up. There is no information in a backtest like this. Only huge potential for misinformation.


I have done some research on expected bond performance with a factor model (as opposed to backtesting). According to data from Fama & French 1963~1991, exposure to the TERM factor (20 years treasury) provides monthly returns of 0.06 with a standard deviation of 3.02. The mv-optimal ratio between stocks and TERM is approximately 77:23, this decreases the standard deviation of your portfolio by a grand total of 1% compared to a portfolio of stocks and cash with the same expected returns. According to own calculations on bond data since 1870, the TERM factor provides monthly returns of 0.12 with a standard deviation of 3.12. In this case, the optimal ratio between stocks and TERM is around 63:37 and such a portfolio decreases your standard deviation compared to a stocks + cash portfolio for the same expected return by around 10%. This is all without leverage.

Cash in this context means the return of RF in Fama & French factor model. Stocks means MKT-RF, TERM means the return of 20 years government bonds. The above might be a little hard to interpret, so here is a simple conclusion: based on data from 1963~1991, leveraging stocks/bonds in the optimal ratio (which can only be known after the fact) provides 1% higher sharpe ratio than a 100% stocks portfolio with the same expected return. Based on data since 1870, leveraging stocks/bonds in the optimal ratio (which can only be known after the fact) provides 10% higher sharpe ratio than a portfolio of 100% stocks.

This analysis assumes a correlation of 0 between stocks and bonds and that leverage occurs at the risk free rate, both of which I consider to be in favor of bonds. This analysis also assumes that the model captures all the variation on bond returns, this is likely a false assumption. Fama & French in "Common risk factors in the returns on stocks and bonds" seem to imply that treasuries with a duration lower than 10 years have statistically significant alpha of around 1% per year but they don't elaborate further. (if this is true, then leveraging long term treasuries is just plain stupid).

For reference SmB and HmL have approximately the same standard deviation as TERM, but the return of SmB is around 1.5 and 3.5 times higher and HmL around 3 to 7 times higher. The conclusion here is that according to Fama & French factor model, leveraging bonds is of little use even if you ignore the existence of SmB and HmL.
EfficientInvestor wrote:
Wed Nov 06, 2019 5:40 pm
The image below is based on historical data since 1955 and we can both agree that the future won't look like the past. So let's just use this image to discuss theory. I'm not trying to imply that the future will look like the past.
As 305Pelusa has mentioned, a risk party portfolio is not the same as MVO. Risk parity is a buzzword that doesn't seem to mean anything useful.

Everybody agrees that leveraging the MVO portfolio results in lower risk for the same return compared to any other portfolio. The problem in practice is that:
  • Gold returns are not statistically different from zero, but you're still using them anyway. Classic case of survivorship bias.
  • bond data since 1955 in no way represents future returns from bonds. There is a 60-year period where TLL (10 year) returned nothing. If you use that period as a basis for your analysis, you would get a completely MVO portfolio. This is less of an issue with stocks since stocks have performed well over basically any 60 year period.
  • Assumptions about the equity risk premium are much more fundamental than assumptions abound gold and bonds, and in my opinion have far lower chance of blowing up in your face.
  • Leverage is not free.
At the end of the day, the optimal investment policy depends on your assumptions. I would like to know what your return assumptions are.
EfficientInvestor wrote:
Wed Nov 06, 2019 10:29 pm
305pelusa wrote:
Wed Nov 06, 2019 9:50 pm
As for your retirement accounts, is that around 14x leverage? How does it compare returns-wise to the more brute, less efficient 229% that I'm using?
Yes, it's around 14x. But you have to remember that 12x of that 14x is in an asset that has had a max drawdown of -4% or so since 1955
I assume that drawdown is in nominal terms. According to PV data the real drawdown is around 12% for CASH and 18% for ITT (1977~1980). I don't have more data for ITT, but I know that CASH has a maximum drawdown of 48% between 1933 and 1952. Conveniently before the start of your backtesting period. One has to wonder how ITT fared in that time period.

I'm not saying that leveraging ITT is a bad idea, but It does have that distinct smell of data mining and recency bias. A leveraged stock portfolio appears to be a lot less susceptive to poor assumptions.

EfficientInvestor
Posts: 234
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: Lifecycle Investing - Leveraging when young

Post by EfficientInvestor » Fri Nov 08, 2019 10:47 am

Uncorrelated wrote:
Fri Nov 08, 2019 10:13 am
Just wanted to add a few thoughts to this discussion.
I really appreciate all your thoughts. I’ll provide a more detailed response when I get time to do so over the weekend. For now, I want to point out that I’m not promoting the use of this kind of leverage on ITT. If I were to use ITT, I would use much less leverage. I’m promoting the use of this leverage on STT. I prefer STT because I believe it will continue to have higher sharpe ratios than ITT or LTT over extended periods of time.

Edit/addition: I’ll also point out that I decrease my amount of leverage as interest rates rise. So if I’m currently 10x right now, my borrow rate is 1.5%, and an unleveraged max drawdown of STT is 4%, then that would be a leveraged max drawdown on my bond side of 10*5.5% = 55%. This is kind of a crude calculation, but it gets me in the ballpark. This is still less than my max drawdown that occur on my stock side. If borrow rates were currently 10%, I would probably be around a max 4x leverage. This would result in a max drawdown of bonds of 4*14% = 56%.
Last edited by EfficientInvestor on Fri Nov 08, 2019 11:02 am, edited 1 time in total.

User avatar
Topic Author
305pelusa
Posts: 903
Joined: Fri Nov 16, 2018 10:20 pm

Re: Lifecycle Investing - Leveraging when young

Post by 305pelusa » Fri Nov 08, 2019 10:49 am

Uncorrelated wrote:
Fri Nov 08, 2019 6:41 am
305pelusa might also be interested in reading this. I get where you are coming from when you are targeting a lifetime constant equity exposure, but that doesn't look very optimal to me. The paper on lifecycle investing and leveraging uses a simple model. The paper from Forsyth and Vetzal, "Robust Asset Allocation for Long-Term Target-Based Investing" does something similar, but in a continuous model with continuously updating asset allocation. My model (see thread above) is discrete but allows for the specification of arbitrary utility functions. Both models reach the same conclusion.
I read that paper. Let me summarize what I took from it and you tell me where/if I'm incorrect please:
- A constant % allocation MV dominates any other strategy that varies allocation in time in a predetermined manner. It's just good old myopic portfolio choice.
- They propose a wealth adaptive allocation. This one changes % allocation along the way based on its relationship to that final goal (high leverage when far, little or no leverage once getting close). This allocation was shown to have similar end wealth targets, but with much less variability about that target.
- This model does not include income. It is only a lump sum. They leave that for future work (darn!).


The way I think you deal with income is similar to how I'm doing here. You discount it to the present and invest accordingly. So I discount it to the present and invest as a constant %. Using a wealth adaptive strategy, you discount to the present and invest like they say. Would that be a fair conclusion?

Here's a couple of thoughts on this:
- First of all, their model is complex, no doubt. They are estimating various parameters from data and it's not totally clear that I could do this correctly, error-proof myself. This isn't my field of study. So I do gather plenty of utility from a simpler model, even if not as good. This sounds trivial but is actually important to me personally.
- I'm not necessarily sure I vibe with their quadratic loss penalty. If I target, say, 1M final wealth and I'm at 990k, their model would tell me to basically go all cash/bond because I can get to 1M on purely cash. I believe these are Deep Risks (as Bernstein would put it) that stocks protect against, like long term high inflation (a la Germany or Latin America). These aren't captured by models like this.
In other words, any model that puts me in extremes (no cash because it's theoretically fine, or Max leverage because theoretically optimal, like EfficientInvestor) is a hard no for me.
- Related to that point ^, when I look at their graphs, their probability of shortfall is much lower. That's a good thing. But their probability of being above is essentially zero (by their own mode construction of course). That's a hard no for me. Call me greedy or what have you but I'm willing to increase shortfall risk to boost my chances of very high wealth tremendously. I can always deal with shortfall by working longer.

This might sound incompatible with Lifecycle Investing (where I sacrifice truly great results to avoid truly bad ones). But the trade off here, looking at the graphs, looks like they sacrifice ALL truly great results (above target) to decrease a bit more of that bad ones. Not sure that's a trade off I like.

I do think I have Constant RRA and power utility. So even at my retirement target, I'm still willing to keep risking for more wealth if consistent with CRRA. Their model would not; once you hit the target, no more risk taking.

With all of that^ in mind, I think I prefer to stay the course. Perhaps I'll change my mind once I see your response.

Uncorrelated
Posts: 105
Joined: Sun Oct 13, 2019 3:16 pm

Re: Lifecycle Investing - Leveraging when young

Post by Uncorrelated » Fri Nov 08, 2019 1:28 pm

EfficientInvestor wrote:
Fri Nov 08, 2019 10:47 am
Uncorrelated wrote:
Fri Nov 08, 2019 10:13 am
Just wanted to add a few thoughts to this discussion.
Edit/addition: I’ll also point out that I decrease my amount of leverage as interest rates rise. So if I’m currently 10x right now, my borrow rate is 1.5%, and an unleveraged max drawdown of STT is 4%, then that would be a leveraged max drawdown on my bond side of 10*5.5% = 55%. This is kind of a crude calculation, but it gets me in the ballpark. This is still less than my max drawdown that occur on my stock side. If borrow rates were currently 10%, I would probably be around a max 4x leverage. This would result in a max drawdown of bonds of 4*14% = 56%.
I glossed over that part because I wasn't sure what to say. It's almost as if you're making the assumption that term risk is higher when inflation is high. Is that a fair statement?

You have a chart that displays the yield on the X axis and the standard deviation in dollars on the Y axis, which can be used to determine the right amount of leverage for your risk level. How much would that chart change if the Y axis was expressed in real dollar standard deviation rather than nominal?

If STT provide higher sharpe ratio than ITT and LTT, it seems that there is an arbitrage opportunity that can be exploited by longing STT and shorting LTT. But curiously most users appear to leverage LTT, why is that?

User avatar
Topic Author
305pelusa
Posts: 903
Joined: Fri Nov 16, 2018 10:20 pm

Re: Lifecycle Investing - Leveraging when young

Post by 305pelusa » Fri Nov 08, 2019 2:18 pm

Uncorrelated wrote:
Fri Nov 08, 2019 6:41 am

The kelly criterion is only relevant if you have a logarithmic utility function (constant risk aversion coefficient of 1). This is almost never the case.
Just to be clear:
The Kelly criterion in this context refers to the maximum leverage you should take, if all you have is financial capital (no income), to maximize your expected return. This also happens to maximize your utility IF you have CRA of 1.0. So only THOSE people would leverage as high as Kelly Criterion.

To the extent you're more risk averse than 1.0, you'd use LESS leverage than Kelly Criterion on a lump sum.

I say this because:
Uncorrelated wrote:
Fri Nov 08, 2019 6:41 am
See the third image in my recent thread for evidence it is rational to employ higher leverage than the maximum one suggested by the kelly criterion:
viewtopic.php?f=10&t=293469
The reason leveraging past the Kelly Criterion can make sense has nothing to do with your utility function here. It's because you have a future income stream.

So I claim the Kelly Criterion has nothing to do with this thread not because of my utility function or RRA. It's because there's a presence of future income.

So I disagree with Lee just like you. Leveraging past KC is sensible. But I don't agree that it's because of utility functions. Hence, even if all of my debt was callable, it would STILL make sense to leverage past the Kelly Criterion.

User avatar
Topic Author
305pelusa
Posts: 903
Joined: Fri Nov 16, 2018 10:20 pm

Re: Lifecycle Investing - Leveraging when young

Post by 305pelusa » Fri Nov 08, 2019 2:24 pm

Uncorrelated wrote:
Fri Nov 08, 2019 10:13 am



I have done some research on expected bond performance with a factor model (as opposed to backtesting). According to data from Fama & French 1963~1991, exposure to the TERM factor (20 years treasury) provides monthly returns of 0.06 with a standard deviation of 3.02. The mv-optimal ratio between stocks and TERM is approximately 77:23, this decreases the standard deviation of your portfolio by a grand total of 1% compared to a portfolio of stocks and cash with the same expected returns. According to own calculations on bond data since 1870, the TERM factor provides monthly returns of 0.12 with a standard deviation of 3.12. In this case, the optimal ratio between stocks and TERM is around 63:37 and such a portfolio decreases your standard deviation compared to a stocks + cash portfolio for the same expected return by around 10%. This is all without leverage.

Cash in this context means the return of RF in Fama & French factor model. Stocks means MKT-RF, TERM means the return of 20 years government bonds. The above might be a little hard to interpret, so here is a simple conclusion: based on data from 1963~1991, leveraging stocks/bonds in the optimal ratio (which can only be known after the fact) provides 1% higher sharpe ratio than a 100% stocks portfolio with the same expected return. Based on data since 1870, leveraging stocks/bonds in the optimal ratio (which can only be known after the fact) provides 10% higher sharpe ratio than a portfolio of 100% stocks.

This analysis assumes a correlation of 0 between stocks and bonds and that leverage occurs at the risk free rate, both of which I consider to be in favor of bonds. This analysis also assumes that the model captures all the variation on bond returns, this is likely a false assumption. Fama & French in "Common risk factors in the returns on stocks and bonds" seem to imply that treasuries with a duration lower than 10 years have statistically significant alpha of around 1% per year but they don't elaborate further. (if this is true, then leveraging long term treasuries is just plain stupid).
I just had a chance to read this. This is FANTASTIC information, truly. It's the thing that I keep coming up against EfficientInvestor: I don't like bond backtest because all they have is gone up. When you look more at the factor and fundamental level, I don't think I'm missing much. This clearly is more rigurous however.

Dude, do you live in the Northeastern area? If so, we should totally hang out. You're like a treasure chest of knowledge.

Lee_WSP
Posts: 1167
Joined: Fri Apr 19, 2019 5:15 pm
Location: Arizona

Re: Lifecycle Investing - Leveraging when young

Post by Lee_WSP » Fri Nov 08, 2019 2:55 pm

Just to be clear. I actually do not understand the Kelly Criterion as it pertains to investing. My own understanding is that there is some bet that is too risky to take even though the expected value is positive. And that is my only point.

rascott
Posts: 995
Joined: Wed Apr 15, 2015 10:53 am

Re: Lifecycle Investing - Leveraging when young

Post by rascott » Fri Nov 08, 2019 4:04 pm

Uncorrelated wrote:
Fri Nov 08, 2019 1:28 pm
EfficientInvestor wrote:
Fri Nov 08, 2019 10:47 am
Uncorrelated wrote:
Fri Nov 08, 2019 10:13 am
Just wanted to add a few thoughts to this discussion.
Edit/addition: I’ll also point out that I decrease my amount of leverage as interest rates rise. So if I’m currently 10x right now, my borrow rate is 1.5%, and an unleveraged max drawdown of STT is 4%, then that would be a leveraged max drawdown on my bond side of 10*5.5% = 55%. This is kind of a crude calculation, but it gets me in the ballpark. This is still less than my max drawdown that occur on my stock side. If borrow rates were currently 10%, I would probably be around a max 4x leverage. This would result in a max drawdown of bonds of 4*14% = 56%.
I glossed over that part because I wasn't sure what to say. It's almost as if you're making the assumption that term risk is higher when inflation is high. Is that a fair statement?

You have a chart that displays the yield on the X axis and the standard deviation in dollars on the Y axis, which can be used to determine the right amount of leverage for your risk level. How much would that chart change if the Y axis was expressed in real dollar standard deviation rather than nominal?

If STT provide higher sharpe ratio than ITT and LTT, it seems that there is an arbitrage opportunity that can be exploited by longing STT and shorting LTT. But curiously most users appear to leverage LTT, why is that?


This was discussed in the HF thread.... different levels of STT leverage depending upon if rates are low/ high to get duration equivalent to LTT

viewtopic.php?f=10&t=288192&p=4764748&h ... n#p4764748

EfficientInvestor
Posts: 234
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: Lifecycle Investing - Leveraging when young

Post by EfficientInvestor » Fri Nov 08, 2019 6:08 pm

Uncorrelated wrote:
Fri Nov 08, 2019 10:13 am
Just wanted to add a few thoughts to this discussion.
You have lots of great stuff to respond to. Let me start by saying that I agree that only doing a backtest since 1980 is a horrible idea. My first backtest I showed was limited in length because I wanted to include a bond fund that had data that everyone could see on PortfolioVisualizer. I then provided a follow-up backtest using my data since 1955 (the furthest back I could find monthly data and the furthest back PV goes). This backtest at least includes a period of time of rising interest rates from 1955-1980. In both cases, the main point of the backtest was to show that using 200%+ leverage on stocks alone has shown the capacity to lose all your money and I don't think it is wise to invest in that manner.

I have run backtests even further back in time using the Simba spreadsheet (viewtopic.php?f=10&t=2520&hilit=simba+s ... t#p4379369). This spreadsheet contains data going back to 1871. I assume this is the same Shiller data that the chart you posted uses. However, the CASH dataset prior to 1934 uses the 1-year interest rate instead of the T-bill (3-month) rate that is used from 1934 onward. So in looking at the use of leverage, it really doesn't make sense to look at the data prior to 1934 since the rates that you are borrowing at are much higher than they really would have been. For instance, the data point for 1934 shows that the 1-year bond returned 1.01% while the T-bill (a much better approximation of risk-free/borrow rate), was only 0.32%. This difference of around 70 bps is about average across all available 1-yr vs 3-mo data.

With that said, I ran a backtest in Simba going back to 1934. However, in order to not cherry pick and in order to make sure I include multiple bond bear markets (like the one from 1915-1920 that is shown in your chart), I went back further to 1910, even though this will include 23 years of much higher leverage cost. For this backtest, I use 8X STT because that provided a similar return/SD ratio to 100% stocks over the time period. Here are allocations:

Image

Here are results:

Image

As I discuss in my article (https://theleveragedindexer.com/2019/08 ... ged-bonds/) about trying to flip 2 investing coins instead of just 1, the goal of leveraging up the bonds is to create a second coin that is similar in return and risk to the first coin. This backtest shows that this has held true since 1910. When you split your returns between coins 1 (P1 - 100% stock) and 2 (P2 - 800% STT), your result (P3) is a return that is similar to both coins but a much lower SD. P4 shows the outcome of using additional leverage so you can have 100% stock and 800% STT. Portfolio 5 shows using 200% stock. I am not trying to point at these results and say, "look...this worked really well and I hope it continues to do so". I'm trying to say, "look...there is a theoretical diversification benefit to using two uncorrelated assets that have shown to have similar return and risk profiles over a really long period of time and these results seem to prove the theory."
Last edited by EfficientInvestor on Mon Nov 11, 2019 9:15 am, edited 1 time in total.

User avatar
Topic Author
305pelusa
Posts: 903
Joined: Fri Nov 16, 2018 10:20 pm

Re: Lifecycle Investing - Leveraging when young

Post by 305pelusa » Fri Nov 08, 2019 6:42 pm

EfficientInvestor wrote:
Fri Nov 08, 2019 6:08 pm
Uncorrelated wrote:
Fri Nov 08, 2019 10:13 am
Just wanted to add a few thoughts to this discussion.
You have lots of great stuff to respond to. Let me start by saying that I agree that only doing a backtest since 1980 is a horrible idea. My first backtest I showed was limited in length because I wanted to include a bond fund that had data that everyone could see on PortfolioVisualizer. I then provided a follow-up backtest using my data since 1955 (the furthest back I could find monthly data and the furthest back PV goes). This backtest at least includes a period of time of rising interest rates from 1955-1980. In both cases, the main point of the backtest was to show that using 200%+ leverage on stocks alone has shown the capacity to lose all your money and I don't think it is wise to invest in that manner.

I have run backtests even further back in time using the Simba spreadsheet. This spreadsheet contains data going back to 1871. I assume this is the same Shiller data that the chart you posted uses. However, the CASH dataset prior to 1934 uses the 1-year interest rate instead of the T-bill (3-month) rate that is used from 1934 onward. So in looking at the use of leverage, it really doesn't make sense to look at the data prior to 1934 since the rates that you are borrowing at are much higher than they really would have been. For instance, the data point for 1934 shows that the 1-year bond returned 1.01% while the T-bill (a much better approximation of risk-free/borrow rate), was only 0.32%. This difference of around 70 bps is about average across all available 1-yr vs 3-mo data.

With that said, I ran a backtest in Simba going back to 1934. However, in order to not cherry pick and in order to make sure I include multiple bond bear markets (like the one from 1915-1920 that is shown in your chart), I went back further to 1910, even though this will include 23 years of much higher leverage cost. For this backtest, I use 8X STT because that provided a similar return/SD ratio to 100% stocks over the time period. Here are allocations:

Image

Here are results:

Image

As I discuss in my article (https://theleveragedindexer.com/2019/08 ... ged-bonds/) about trying to flip 2 investing coins instead of just 1, the goal of leveraging up the bonds is to create a second coin that is similar in return and risk to the first coin. This backtest shows that this has held true since 1910. When you split your returns between coins 1 (P1 - 100% stock) and 2 (P2 - 800% STT), your result (P3) is a return that is similar to both coins but a much lower SD. P4 shows the outcome of using additional leverage so you can have 100% stock and 800% STT. Portfolio 5 shows using 200% stock. I am not trying to point at these results and say, "look...this worked really well and I hope it continues to do so". I'm trying to say, "look...there is a theoretical diversification benefit to using two uncorrelated assets that have shown to have similar return and risk profiles over a really long period of time and these results seem to prove the theory."
Just to be really clear so it does not get lost on the rhetoric:

There's two reasons to leverage. There is leverage to further increase the risk and return of an asset and (hopefully) create a more efficient portfolio via asset diversification. This is your second coin toss and the backbone of CAPM. Then there's leverage to temporally diversify your future income stream (this is what Lifecycle Investing and wealth adaptive strategies do). This is temporal diversification.

For many reasons, I believe temporal diversification is superior and far far more important to accomplish. I will touch leverage to diversify temporally but don't care enough to make a portfolio more efficient. I can achieve risk-parity like portfolios via factors, riskier stocks, and longer term bonds (think Larry portfolio).

The implication is that if I didn't have an income, I wouldn't be leveraged. And I'm not planning on being leveraged as a retiree either. You imply that you would be and presumably plan to be leveraged for your whole life (just less so once retired since you don't need so much leverage to also diversify temporally).


I thought you understood this but I bring it up because:
EfficientInvestor wrote:
Fri Nov 08, 2019 6:08 pm
In both cases, the main point of the backtest was to show that using 200%+ leverage on stocks alone has shown the capacity to lose all your money and I don't think it is wise to invest in that manner.
You're confusing the two here. Why are you comparing a 200% leveraged strategy in terms of efficiency to your leveraged bonds from 1910 to 2018? I'm so confused. Again, I leverage to diversify temporally. That means two things:
- Leverage is not constantly that high. Only for the first 10-15 years.
- There must be an income stream.

The diversification benefit will be seen as less variability about a terminal wealth, NOT less variability of returns (which is what you're looking at in your spreadsheets). I just don't think you're thinking about this correctly.

If you want to judge how much better leveraged bonds are over pure stocks as a temporal diversification, then you need to do the analysis Nalebuff did. You need to assume some income stream, with some real growth, with a working life of X years, retirement at Y age and an allocation that deleverages as time goes by. Then you need to standardize the mean terminal wealth of both only leveraged stocks and leveraged stocks/bonds, and check out which one had the lowest St Dev for that terminal wealth. This isn't what you're showing at all. I do venture a guess that leveraged bonds helps, but I'm not really convinced that it helps that much.

What you seem to be showing is the efficiency in terms of returns (Sharpe) of your leveraged portfolio. In that case, compare apples to apples to my Samuelson allocation of 30% stocks, NOT 200%. It's just that the 30% gets applied to my discounted future income, coming up with 200% of current financial wealth. So P5 on your images should be 30% stocks, 70% LT Bonds (my unleveraged allocation in the presence of no income stream).

Now does that allocation really do that badly compared to your leveraged, retirement allocation in terms of efficiency/Sharpe?

rascott
Posts: 995
Joined: Wed Apr 15, 2015 10:53 am

Re: Lifecycle Investing - Leveraging when young

Post by rascott » Fri Nov 08, 2019 7:12 pm

EfficientInvestor wrote:
Fri Nov 08, 2019 6:08 pm
Uncorrelated wrote:
Fri Nov 08, 2019 10:13 am
Just wanted to add a few thoughts to this discussion.
You have lots of great stuff to respond to. Let me start by saying that I agree that only doing a backtest since 1980 is a horrible idea. My first backtest I showed was limited in length because I wanted to include a bond fund that had data that everyone could see on PortfolioVisualizer. I then provided a follow-up backtest using my data since 1955 (the furthest back I could find monthly data and the furthest back PV goes). This backtest at least includes a period of time of rising interest rates from 1955-1980. In both cases, the main point of the backtest was to show that using 200%+ leverage on stocks alone has shown the capacity to lose all your money and I don't think it is wise to invest in that manner.

I have run backtests even further back in time using the Simba spreadsheet (viewtopic.php?f=10&t=2520&hilit=simba+s ... t#p4379369). This spreadsheet contains data going back to 1871. I assume this is the same Shiller data that the chart you posted uses. However, the CASH dataset prior to 1934 uses the 1-year interest rate instead of the T-bill (3-month) rate that is used from 1934 onward. So in looking at the use of leverage, it really doesn't make sense to look at the data prior to 1934 since the rates that you are borrowing at are much higher than they really would have been. For instance, the data point for 1934 shows that the 1-year bond returned 1.01% while the T-bill (a much better approximation of risk-free/borrow rate), was only 0.32%. This difference of around 70 bps is about average across all available 1-yr vs 3-mo data.

With that said, I ran a backtest in Simba going back to 1934. However, in order to not cherry pick and in order to make sure I include multiple bond bear markets (like the one from 1915-1920 that is shown in your chart), I went back further to 1910, even though this will include 23 years of much higher leverage cost. For this backtest, I use 8X STT because that provided a similar return/SD ratio to 100% stocks over the time period. Here are allocations:

Image

Here are results:

Image

As I discuss in my article (https://theleveragedindexer.com/2019/08 ... ged-bonds/) about trying to flip 2 investing coins instead of just 1, the goal of leveraging up the bonds is to create a second coin that is similar in return and risk to the first coin. This backtest shows that this has held true since 1910. When you split your returns between coins 1 (P1 - 100% stock) and 2 (P2 - 800% STT), your result (P3) is a return that is similar to both coins but a much lower SD. P4 shows the outcome of using additional leverage so you can have 100% stock and 800% STT. Portfolio 5 shows using 200% stock. I am not trying to point at these results and say, "look...this worked really well and I hope it continues to do so". I'm trying to say, "look...there is a theoretical diversification benefit to using two uncorrelated assets that have shown to have similar return and risk profiles over a really long period of time and these results seem to prove the theory."


I know that you hold two year futures..... but how much are you actually leveraging your total portfolio? Is it more P3 or you going big (P4)?

EfficientInvestor
Posts: 234
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: Lifecycle Investing - Leveraging when young

Post by EfficientInvestor » Fri Nov 08, 2019 7:44 pm

305pelusa wrote:
Fri Nov 08, 2019 6:42 pm
Just to be really clear so it does not get lost on the rhetoric:

There's two reasons to leverage. There is leverage to further increase the risk and return of an asset and (hopefully) create a more efficient portfolio via asset diversification. This is your second coin toss and the backbone of CAPM. Then there's leverage to temporally diversify your future income stream (this is what Lifecycle Investing and wealth adaptive strategies do). This is temporal diversification.

For many reasons, I believe temporal diversification is superior and far far more important to accomplish. I will touch leverage to diversify temporally but don't care enough to make a portfolio more efficient. I can achieve risk-parity like portfolios via factors, riskier stocks, and longer term bonds (think Larry portfolio).
Maybe I should come up to the northeast and hang out with you so we can hash this out in person. I'm really enjoying this exchange. :sharebeer

I understand what you are saying about the differences between asset diversification and temporal diversification. I am in complete agreement that temporal diversification is important and perhaps even more important than asset diversification in the same way that savings rate is more important than asset diversification. However, I also believe that asset diversification can provide a huge benefit on top of temporal diversification. I am trying to say that if you combine them together and do BOTH, it should be much better than just doing one or the other. Looking back to my 1910 backtest...P3 is how someone might invest if you are just doing asset diversification. P5 is how you are currently investing while in the earlier stages of your temporal spectrum. P4 is BOTH. Why would you not invest in something like P4 (or a more leveraged version) now and then ramp it down to 30/70? Like I said earlier, why not start at the green dot (or maybe even at the intersection of the orange line and 35% SD) and ramp down the orange line instead of starting at the blue dot and ramping down the black dashed line?

Image

When I find time, I will try to do the analysis you mention about determining variability of terminal net worth. I have a very strong feeling that spending 20-30 years traveling from the green dot to the red dot will have a better outcome than traveling from the blue dot to the red dot.
Last edited by EfficientInvestor on Fri Nov 08, 2019 8:32 pm, edited 2 times in total.

EfficientInvestor
Posts: 234
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: Lifecycle Investing - Leveraging when young

Post by EfficientInvestor » Fri Nov 08, 2019 7:45 pm

rascott wrote:
Fri Nov 08, 2019 7:12 pm
I know that you hold two year futures..... but how much are you actually leveraging your total portfolio? Is it more P3 or you going big (P4)?
I'm currently just above P4. 120% stock, 1,200% 2-year treasury, 40% gold.

User avatar
Topic Author
305pelusa
Posts: 903
Joined: Fri Nov 16, 2018 10:20 pm

Re: Lifecycle Investing - Leveraging when young

Post by 305pelusa » Fri Nov 08, 2019 10:36 pm

EfficientInvestor wrote:
Fri Nov 08, 2019 7:44 pm

Maybe I should come up to the northeast and hang out with you so we can hash this out in person. I'm really enjoying this exchange. :sharebeer
Bro hit me up any time you're in Boston :sharebeer
EfficientInvestor wrote:
Fri Nov 08, 2019 7:44 pm
I am trying to say that if you combine them together and do BOTH, it should be much better than just doing one or the other.
1) I think it will only be a little better since most of the gains from temporal diversification (eliminating sequence of returns risk) is already solved with purely stocks. That's just my guess.
2) More importantly, I don't see how you can actually do both. The amount of leverage I would need to fully diversify temporally with a bond-stock portfolio (especially ST bonds) is almost incomprehensible. Right now I consider myself at 30% stock/70% cash (applied to my entire lifetime wealth). If I just wanted to be 30% stock/70% ST bonds, I'd need to short about 1.5M in cash right now. So I'd need to buy 8 ST Treasury contracts right now.

And that's only to get to 30% stock/70% ST bonds. You're suggesting P3 which is 50% Stocks / 450% ST bonds / -400% cash (I'm assuming the -390% was a typoe on your screenshot). So your retirement portfolio has 4x leverage. So I'd need to then grab the 2.29 leverage I have on stocks via options and debt, the 8 ST futures contracts, and leverage THAT 4 times more!. For an amazing 9.16 times leverage on stocks and 32 ST Treasury contracts!

That's how much leverage I would need to fully diversify temporally with a portfolio like P3. That's how much cash I need to short to completely short all of my future cash income four times over (!).

Consider me skeptical but even though I don't know your age, labor income, retirement age, etc, I'm willing to bet you're not anywhere nearly leveraged enough to diversify temporally. And temporal diversification is far more useful because the correlation of returns between years is far lower than correlation among assets. Diversifying temporally is like investing in 30+ different assets with low correlation to each other.

Which leads me to:
3) I could simply go long a few ST treasury contracts right now in addition to my 2.29 stock leverage right? Won't be anywhere near risk parity but it's a step in the right direction right? True. That's something I've certainly considered.

But achieving your green dot with full temporal diversification? It's certainly beyond possible for me right now.

EfficientInvestor
Posts: 234
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: Lifecycle Investing - Leveraging when young

Post by EfficientInvestor » Fri Nov 08, 2019 11:10 pm

305pelusa wrote:
Fri Nov 08, 2019 10:36 pm
EfficientInvestor wrote:
Fri Nov 08, 2019 7:44 pm

Maybe I should come up to the northeast and hang out with you so we can hash this out in person. I'm really enjoying this exchange. :sharebeer
Bro hit me up any time you're in Boston :sharebeer
EfficientInvestor wrote:
Fri Nov 08, 2019 7:44 pm
I am trying to say that if you combine them together and do BOTH, it should be much better than just doing one or the other.
1) I think it will only be a little better since most of the gains from temporal diversification (eliminating sequence of returns risk) is already solved with purely stocks. That's just my guess.
2) More importantly, I don't see how you can actually do both. The amount of leverage I would need to fully diversify temporally with a bond-stock portfolio (especially ST bonds) is almost incomprehensible. Right now I consider myself at 30% stock/70% cash (applied to my entire lifetime wealth). If I just wanted to be 30% stock/70% ST bonds, I'd need to short about 1.5M in cash right now. So I'd need to buy 8 ST Treasury contracts right now.

And that's only to get to 30% stock/70% ST bonds. You're suggesting P3 which is 50% Stocks / 450% ST bonds / -400% cash (I'm assuming the -390% was a typoe on your screenshot). So your retirement portfolio has 4x leverage. So I'd need to then grab the 2.29 leverage I have on stocks via options and debt, the 8 ST futures contracts, and leverage THAT 4 times more!. For an amazing 9.16 times leverage on stocks and 32 ST Treasury contracts!

That's how much leverage I would need to fully diversify temporally with a portfolio like P3. That's how much cash I need to short to completely short all of my future cash income four times over (!).

Consider me skeptical but even though I don't know your age, labor income, retirement age, etc, I'm willing to bet you're not anywhere nearly leveraged enough to diversify temporally. And temporal diversification is far more useful because the correlation of returns between years is far lower than correlation among assets. Diversifying temporally is like investing in 30+ different assets with low correlation to each other.

Which leads me to:
3) I could simply go long a few ST treasury contracts right now in addition to my 2.29 stock leverage right? Won't be anywhere near risk parity but it's a step in the right direction right? True. That's something I've certainly considered.

But achieving your green dot with full temporal diversification? It's certainly beyond possible for me right now.
My wife's uncle lives in Worcester and we want to make it back up there in the next few years to see him. I'll have to remember to reach out to you whenever that happens.

As for your questions...if I remember correctly, you currently have a little over $400k in equity exposure. Let's just call it $400k for round numbers. You currently have it all in stocks (coin #1). But forget stocks and bonds for the time being. Let's just talk in terms of coins. Based on your calculations, you need $400k at risk at the moment to achieve your temporal diversification. You currently have that $400k completely dependent each year upon the flip of coin #1. But let's assume coin #2 is uncorrelated to coin #1 and has the same return and risk profile. Is there any reason why you can't split your funds at risk between the two coins and have $200k in each? By doing so, you could receive the full asset diversification benefit in addition to the temporal diversification. Now, let's bring types of assets back into the fold. That means you would reduce your stock position to $200k and put the other $200k toward a bond position that has similar risk/return to stock. I'm claiming that would currently be 10X your stock balance (due to lower than average interest rates), so $2 million or about 10 /ZT contracts. Or, if you just stick to the backtest that is shown, you could do 8X and be at $1.6 million or 8 /ZT contracts. But you would have plenty of funds to buy the 8 /ZT contracts (coin #2) by backing off on some of the exposure in stocks (coin #1).

User avatar
Topic Author
305pelusa
Posts: 903
Joined: Fri Nov 16, 2018 10:20 pm

Re: Lifecycle Investing - Leveraging when young

Post by 305pelusa » Fri Nov 08, 2019 11:49 pm

EfficientInvestor wrote:
Fri Nov 08, 2019 11:10 pm
As for your questions...if I remember correctly, you currently have a little over $400k in equity exposure. Let's just call it $400k for round numbers. You currently have it all in stocks (coin #1). But forget stocks and bonds for the time being. Let's just talk in terms of coins. Based on your calculations, you need $400k at risk at the moment to achieve your temporal diversification. You currently have that $400k completely dependent each year upon the flip of coin #1. But let's assume coin #2 is uncorrelated to coin #1 and has the same return and risk profile. Is there any reason why you can't split your funds at risk between the two coins and have $200k in each? By doing so, you could receive the full asset diversification benefit in addition to the temporal diversification. Now, let's bring types of assets back into the fold. That means you would reduce your stock position to $200k and put the other $200k toward a bond position that has similar risk/return to stock. I'm claiming that would currently be 10X your stock balance (due to lower than average interest rates), so $2 million or about 10 /ZT contracts. Or, if you just stick to the backtest that is shown, you could do 8X and be at $1.6 million or 8 /ZT contracts. But you would have plenty of funds to buy the 8 /ZT contracts (coin #2) by backing off on some of the exposure in stocks (coin #1).
Ah ok I was double counting. I thought you were saying P3 is how you'd invest purely for asset diversification (with no income present). Just a misunderstanding then.
You're effectively saying I switch from 30% stock/70% cash, to 15% stock/150% ST Treas./-65% cash.

That could be achieved with 2.5 times leverage on the stocks (80k of equity) and the remaining 120k for the 2M in contracts (a hefty 16x leverage).

This is what I thought you were saying since the beginning, and something I don't feel comfortable with because it's a lot of leverage. Yes, I know it's a lot of leverage on something very stable. But leverage is very unforgiving.

I also don't feel comfortable with the idea of being leveraged as a retiree, for many reasons (psychological, practical, etc).

Personally (and I've mentioned this before already), I think the best compromise is to shift my retirement allocation from 30% stocks/70% cash, to 30% stocks/70% bonds (most likely IT bonds to use less leverage). This would require I simply go a long a few contracts (maybe just 5 of them), which IMO is far more reasonable and manageable. And it's something I do continue to ponder.

BTW, I don't actually plan to retire with 70% in cash. I would use IT bonds then. It's just that for purposes of temporal diversification, I count it as cash.

It's something to think about. You're really just telling me to diversify my bond exposure temporally as well ("why are you planning to buy IT bonds when retired but aren't buying any now?"). And my answer continues to be the same one: "Because I can diversify my stock exposure temporally with far less leverage, so it's an easier gimme. The leverage I need to diversify term exposure temporally would be quite high".

Once more, it's something I do consider and have answered in this thread before already. I just haven't pursued it because it would require more leverage than I'n using right now.

Uncorrelated
Posts: 105
Joined: Sun Oct 13, 2019 3:16 pm

Re: Lifecycle Investing - Leveraging when young

Post by Uncorrelated » Sat Nov 09, 2019 6:56 pm

EfficientInvestor will have to wait a bit longer on a response so I have time to review some literature.
305pelusa wrote:
Fri Nov 08, 2019 10:49 am
Uncorrelated wrote:
Fri Nov 08, 2019 6:41 am
305pelusa might also be interested in reading this. I get where you are coming from when you are targeting a lifetime constant equity exposure, but that doesn't look very optimal to me. The paper on lifecycle investing and leveraging uses a simple model. The paper from Forsyth and Vetzal, "Robust Asset Allocation for Long-Term Target-Based Investing" does something similar, but in a continuous model with continuously updating asset allocation. My model (see thread above) is discrete but allows for the specification of arbitrary utility functions. Both models reach the same conclusion.
I read that paper. Let me summarize what I took from it and you tell me where/if I'm incorrect please:

- This model does not include income. It is only a lump sum. They leave that for future work (darn!).
My model supports arbitrary functions and can also simulate such a model with income. Should not be hard to do.

(a few hours later) The following figures assume CRRA (isoelastic utility) with a risk coefficient of gamma = 1, 2, 3 or 5. The model attempts to optimize the terminal utility function under the assumption of saving 10k per year. To be clear, the model does not optimize for each individual timestep, but optimizes for the sum of all future utility. The Y axis indicates your current wealth, the X axis indicates the amount of years left before you measure your utility. Color indicates the asset allocation in stocks. The contour lines indicate the 'expected' path (the numbers are not important).

gamma 1.0
gamma 2.0
gamma 3.0
gamma 5.0

If I simulate with annual savings equal to zero, the resulting asset allocation is a constant. In particular, with gamma = 1 the resulting asset allocation is the kelly criterion since gamma = 1 means a logarithmic utility function. With gamma < 1 (e.g. 0.5) the resulting allocation appears to favor a boatload of leverage.

What we see here is similar to the lifecycle investing paper. When retired, determine your portfolio according to your risk tolerance. Before retirement, use a more aggressive asset allocation. Even if you are extremely risk averse (gamma = 5, normally around 30% stocks), you should use leverage eyarly in your career. It is possible to improve the model by deriving utility from spending during retirement rather than at an arbitrary end date, but my optimizer cannot do that... yet. aacalc.com has a paper (Human Capital, Social Security, and Asset Allocation) that does that. (I didn't full read the paper yet, but it sounds interesting. Calculation method is identical to my model).

If I have enough time I'll do a showdown between this model and the model from lifecycle investing.
The way I think you deal with income is similar to how I'm doing here. You discount it to the present and invest accordingly. So I discount it to the present and invest as a constant %. Using a wealth adaptive strategy, you discount to the present and invest like they say. Would that be a fair conclusion?
I don't really think of it that way. I think in terms of a goal and then try to determine the asset allocation that maximizes the probability of hitting that goal. The advantage of this model is that you avoid the difficulty of specifying one's risk tolerance. The disadvantage is that specifying your goals in a way that results in robust asset allocations can be difficult.

According to aacalc.com the lifetime wealth approach is a good approximation of the optimal asset allocation (https://www.aacalc.com/docs/simple_rules).
- First of all, their model is complex, no doubt. They are estimating various parameters from data and it's not totally clear that I could do this correctly, error-proof myself. This isn't my field of study. So I do gather plenty of utility from a simpler model, even if not as good. This sounds trivial but is actually important to me personally.
That is true. However A) the model depends on little more than the existence of the equity risk premium. B) your simpler model (from lifecycle investing paper) also uses similar estimation parameters (the expectancy and variance of stock and bond returns are needed to calculate the optimal asset allocation). Therefore I am not really sure whether the chance of estimation errors differs between the two models.

One big difference between the two models is that the paper I linked is optimal given the assumptions, and all assumptions are given. The solution, by design, can't do market timing. An empirical model gives less certainty. For example, how do you know the lifecycle investing model doesn't exploit long-term mean reversion by accident? A more complex model doesn't have to be less robust.

Uncorrelated
Posts: 105
Joined: Sun Oct 13, 2019 3:16 pm

Re: Lifecycle Investing - Leveraging when young

Post by Uncorrelated » Sun Nov 10, 2019 7:22 am

rascott wrote:
Fri Nov 08, 2019 4:04 pm
This was discussed in the HF thread.... different levels of STT leverage depending upon if rates are low/ high to get duration equivalent to LTT

viewtopic.php?f=10&t=288192&p=4764748&h ... n#p4764748
I don't think that is the answer. I did some research and it appears that long term bonds suffer from a low beta anomaly. An LTT (20 years) bond has similar TERM exposure as a 10x leveraged short term (2 years) bond. The returns of the long term bond are almost completely explained by the exposure to the TERM factor, but the 10x short term bond portfolio has residual volatility that is about 2 times as large as expected from just the TERM exposure. This result is consistent when comparing short term bonds to immediate term bonds, immediate term bonds to long term bonds, and long term bonds to EDV.

This means that duration-matching a long bond with a leveraged short term bond is a horrible idea. See also the remainder of this post.

(note that duration matching is distinct from leveraging in general. I specifically mean that getting exposure to the TERM factor with leveraged short term bonds is a bad idea. However, that is not the only reason to leverage STT).
EfficientInvestor wrote:
Fri Nov 08, 2019 6:08 pm
Uncorrelated wrote:
Fri Nov 08, 2019 10:13 am
Just wanted to add a few thoughts to this discussion.
You have lots of great stuff to respond to.
Thanks for your response. The first thing that I did was to verify your results with simba's backtesting spreadsheet. I notice something odd, namely that the STT has much higher returns than one would expect based on TERM exposure. This effect is also visible with LTT. Curiously, EDM shows the opposite effect (lower returns than expected based on TERM exposure). This appears to indicate a low beta anomaly in bonds.

I reviewed some literature about this effect. The first paper is The Short-Term Corporate Bond Anomaly. This paper investigates corporate bonds in the period 1993-2003 (yuk) and tests the low volatility anomaly under four different factor models, all fail to explain the performance of short term bonds. This piece from Swedroe claims the low volatility appears in most/all bond markets outside the US (also government) between 1997-2012: https://www.etf.com/sections/index-inve ... nopaging=1. Unfortunately I wasn't able to find research on US government bonds and/or longer periods of time, but I have seen anecdotic evidence of this effect pop up in a few places.

To conclude, existing factor models don't do a good job of explaining the performance of short term bonds. My prior analysis with Fama & French model does a good job of explaining the expected returns of leveraged long term bonds, but not for short term bonds. Provided that the low beta anomaly persists over time and that borrowing costs are less than the low beta anomaly, this can definitely work. I'm not so optimistic about long term bond strategies. The flip side of the low beta anomaly is that high-beta (high TERM) bonds have lower returns than expected, this explains why EDV has a negative alpha.

I did a quick search in HEDGEFUNDIE's threads but was unable to find discussion on the low beta anomaly in bonds. Simba's backtesting spreadsheet suggests that shorting EDV until the TERM exposure of your portfolio reaches zero improves the portfolio sharpe ratio, essentially forming a long-short bond low beta strategy. This is consistent with the theory from the research literature above.
305pelusa wrote:
Fri Nov 08, 2019 10:36 pm
1) I think it will only be a little better since most of the gains from temporal diversification (eliminating sequence of returns risk) is already solved with purely stocks. That's just my guess.
2) More importantly, I don't see how you can actually do both. The amount of leverage I would need to fully diversify temporally with a bond-stock portfolio (especially ST bonds) is almost incomprehensible. Right now I consider myself at 30% stock/70% cash (applied to my entire lifetime wealth). If I just wanted to be 30% stock/70% ST bonds, I'd need to short about 1.5M in cash right now. So I'd need to buy 8 ST Treasury contracts right now.
I believe this is a gross misinterpretation of the theory behind the lifecycle investment paper. Let me explain.

The lifecycle investment paper makes the assumption that you can choose between two different assets. Stocks and bonds. They model stocks with a random variable with expected return of X and standard deviation of Y. Bonds are modeled by the risk free return rate. Stocks can be leveraged.

At the end of the paper, the authors conclude that leveraging stocks early in your career results in higher utility. But that is not what they actually mean. What they actually mean is that *if* you had access to an asset with higher risk and higher returns than stocks, it is rational to choose that asset even if it would exceed your personal risk tolerance. Leveraged stocks are one such asset with higher risk and reward, but there are many others (leveraged STT might be one). You should not keep a constant amount of your lifetime wealth in stocks. You should choose your portfolio in such a way that the combination of your human capital and portfolio maximizes your future utility. In a simple model with only leveraged stocks, this can be approximated with the concept of lifetime wealth in stocks.

My model also uses leveraged stocks, but that is a simplification. It doesn't mean that you should use leveraged stocks, just that it is rational to choose an asset allocation that is more aggressive than 100% stocks.
305pelusa wrote:
Fri Nov 08, 2019 2:24 pm
Dude, do you live in the Northeastern area? If so, we should totally hang out. You're like a treasure chest of knowledge.
I live around 5000 km away. Probably not going to happen.

EfficientInvestor
Posts: 234
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: Lifecycle Investing - Leveraging when young

Post by EfficientInvestor » Mon Nov 11, 2019 9:20 am

305pelusa wrote:
Fri Nov 08, 2019 10:36 pm
You're suggesting P3 which is 50% Stocks / 450% ST bonds / -400% cash (I'm assuming the -390% was a typoe on your screenshot).
I updated my screenshots. I had been playing around with a few different allocations and forgot to change it back. I also updated P4 to be 100/900 (instead of 100/800) so that it would have the same ratios as P3.

Image
Image

UberGrub
Posts: 77
Joined: Wed Aug 21, 2019 3:47 pm

Re: Lifecycle Investing - Leveraging when young

Post by UberGrub » Fri Nov 15, 2019 6:07 pm

I'm considering another set of options. I have done very well with this first set. However, the markets have gone up a lot thus far so I'm afraid of buying at the peak (yeah yeah, I know that's market timing but still).

Any thoughts 305pelusa?

ebot
Posts: 4
Joined: Thu Nov 07, 2019 5:48 pm

Re: Lifecycle Investing - Leveraging when young

Post by ebot » Fri Nov 15, 2019 6:48 pm

Can you provide some insight into which specific LEAPS you are buying (which S&P ETF, how far out, strike price, etc)? Are you using any stop limits to in case of a large drop?

User avatar
Topic Author
305pelusa
Posts: 903
Joined: Fri Nov 16, 2018 10:20 pm

Re: Lifecycle Investing - Leveraging when young

Post by 305pelusa » Fri Nov 15, 2019 8:17 pm

UberGrub wrote:
Fri Nov 15, 2019 6:07 pm
I'm considering another set of options. I have done very well with this first set. However, the markets have gone up a lot thus far so I'm afraid of buying at the peak (yeah yeah, I know that's market timing but still).

Any thoughts 305pelusa?
Buy additional market exposure to hedge your ideosyncratic lifetime sequence of returns risk. It has nothing to do with current prices.

I like MT's advice:

"Take your exposure to risky asset classes on your current allocation path for the next 10-15 years, multiply it by, say, 1.01, and fund this by reducing exposure to cash and bonds. Reduce exposure in real dollar terms to these risky asset classes by the same amount after 10-15 years. If your means allow, increase the multiplier by 1.02, 1.03, etc. It's a free lunch, thanks to intertemporal diversification."
ebot wrote:
Fri Nov 15, 2019 6:48 pm
Can you provide some insight into which specific LEAPS you are buying (which S&P ETF, how far out, strike price, etc)? Are you using any stop limits to in case of a large drop?
I did earlier. Here's a set I bought/sold:
viewtopic.php?p=4709613#p4709613

Here's the process for figuring them out:
viewtopic.php?p=4712716#p4712716

No stop limits. I recommend that if you leverage with margineable derivatives (like options and futures) you keep leverage low enough so that your salary/savings contributions can stabilize the portfolio against the harshest environments and avoid liquidation. I believe the one thing where this can go wrong is forced liquidation. It is also why I do not use Leveraged ETFs. The systematic buying high and selling low is effectively being long volatility, which is not a position that has been generally rewarded historically.
Just my 2 cents.

ebot
Posts: 4
Joined: Thu Nov 07, 2019 5:48 pm

Re: Lifecycle Investing - Leveraging when young

Post by ebot » Fri Nov 15, 2019 8:41 pm

Thanks for the info. I picked up the book and am starting to go through it. I did stumble across a few other boglehead posts about this same topic. Have you seen this one:
viewtopic.php?t=88487&start=100

In the end, it looks like the OP decided the risk of irreversable losses made it not worth it to leverage.

User avatar
Topic Author
305pelusa
Posts: 903
Joined: Fri Nov 16, 2018 10:20 pm

Re: Lifecycle Investing - Leveraging when young

Post by 305pelusa » Fri Nov 15, 2019 9:21 pm

ebot wrote:
Fri Nov 15, 2019 8:41 pm
Thanks for the info. I picked up the book and am starting to go through it. I did stumble across a few other boglehead posts about this same topic. Have you seen this one:
viewtopic.php?t=88487&start=100

In the end, it looks like the OP decided the risk of irreversable losses made it not worth it to leverage.
It's a good thread. I learned a lot from MT's posts.

The OP's concept of an irreversible loss has to do with the fact that with leverage, it's possible to loss everything today. Ex: you buy a 2:1 leveraged LEAP and if the market drops 50%, the call is worthless. You lose everything related to that option at that point.

I believe the correct mindset is to think in terms of your lifetime. I could lose all of my financial capital today and be fine. Most of my capital is human capital, to be earned and saved in the future. So I'm not concerned with % losses ("will I lose 100%?") but with real dollar losses ("will I lose 200k?"). I want to make sure I keep a somewhat steady level of real dollar risk (the latter) throughout my life, varying my equity exposure as needed to achieve that. This means leverage when financial assets are low.

What is crucial is that you can continue to maintain the same notional exposure to the market so you can actually participate in the recoveries. That's key. That's why this eliminates sequence of returns risk. So the person who buys a LEAP that is about to expire worthless better saved enough money during the life of the option to immediately purchase another one, maintaining exposure.

UberGrub
Posts: 77
Joined: Wed Aug 21, 2019 3:47 pm

Re: Lifecycle Investing - Leveraging when young

Post by UberGrub » Sun Nov 17, 2019 12:08 pm

305pelusa wrote:
Fri Nov 15, 2019 8:17 pm
UberGrub wrote:
Fri Nov 15, 2019 6:07 pm
I'm considering another set of options. I have done very well with this first set. However, the markets have gone up a lot thus far so I'm afraid of buying at the peak (yeah yeah, I know that's market timing but still).

Any thoughts 305pelusa?
Buy additional market exposure to hedge your ideosyncratic lifetime sequence of returns risk. It has nothing to do with current prices.

I like MT's advice:

"Take your exposure to risky asset classes on your current allocation path for the next 10-15 years, multiply it by, say, 1.01, and fund this by reducing exposure to cash and bonds. Reduce exposure in real dollar terms to these risky asset classes by the same amount after 10-15 years. If your means allow, increase the multiplier by 1.02, 1.03, etc. It's a free lunch, thanks to intertemporal diversification."
ebot wrote:
Fri Nov 15, 2019 6:48 pm
Can you provide some insight into which specific LEAPS you are buying (which S&P ETF, how far out, strike price, etc)? Are you using any stop limits to in case of a large drop?
I did earlier. Here's a set I bought/sold:
viewtopic.php?p=4709613#p4709613

Here's the process for figuring them out:
viewtopic.php?p=4712716#p4712716

No stop limits. I recommend that if you leverage with margineable derivatives (like options and futures) you keep leverage low enough so that your salary/savings contributions can stabilize the portfolio against the harshest environments and avoid liquidation. I believe the one thing where this can go wrong is forced liquidation. It is also why I do not use Leveraged ETFs. The systematic buying high and selling low is effectively being long volatility, which is not a position that has been generally rewarded historically.
Just my 2 cents.
Thank you for the response. I will acquire another set this upcoming week.

Zilvervloot
Posts: 11
Joined: Fri Jun 01, 2018 3:00 am

Re: Lifecycle Investing - Leveraging when young

Post by Zilvervloot » Mon Nov 18, 2019 4:52 am

comeinvest wrote:
Tue Sep 10, 2019 2:33 am
Zilvervloot wrote:
Thu Sep 05, 2019 7:22 am
For those who invest internationally, European futures might be interesting. Libor for the Euro is negative at the moment making financing costs quite low. The Eurostoxx 50 future (the European equivalent of the S&P 500) has quite low financing cost of around 1,0 to 1,5%.
E.g. the dec future trades at 3461 giving a discount of 18 points to the index of 3479. Taking into account a dividend of 3,5% (and around 40 points per quarterly future) this leads to an implied financing rate of 1,4%. Problem is that dividends are not always distributed quarterly in Europe, so if dividends are mostly distributed in the first half year you overestimate the implied financing rate.

Back to the options, do I understand correctly that for those without capital gains taxes, it is better to chose a higher strike price for the call/put option pair?
I think something must be wrong with your math as well as with your logic. Short-term risk-free rates in Europe are -0.5% or even less if I'm not mistaken. An implied financing rate of 1.4% would be ca. 2% above the risk-free rate and therefore a very bad deal. I believe your dividend estimate is far off.

Unfortunately I have not found any source for expected dividends, or for implied financing rates on Eurex futures.
Yes, my dividends were wrong. I finally found out that dividends are paid mostly between april and july for European stocks (https://www.stoxx.com/index-details?symbol=SX5ED. In my calculations I assumed an even payout during the year. After correction for that I indeed found an implied financing rate of -0,5%.

I think this would be a great option for European investors, or US investors wanting to diversify.

guyinlaw
Posts: 94
Joined: Wed Jul 03, 2019 9:54 am

Re: Lifecycle Investing - Leveraging when young

Post by guyinlaw » Mon Nov 18, 2019 12:38 pm

305pelusa wrote:
Fri Mar 01, 2019 12:11 am

Nov 2019
Stock Exposure = 412k
Effective Debt = 233k
Equity in the exposure = 179k
Leverage = 2.29
Excellent thread. Thank you posting and sharing your experience.
305pelusa wrote:
Fri Mar 01, 2019 12:11 am
If the market drops, my current savings (invested in stocks) goes down. But all of my future income (bond-like) is untouched.
You have mentioned many times in the thread that you have no bonds in your portfolio and consider future income as a bond. You do not buy bonds with leverage (treasury futures) because the leverage needed is quite high.

The concern I have with this argument is that if stocks drop by >15% ( Dec 2018) or by >50% you will not have the future cash earning available to rebalance.

Consider the two portfolios
P1 200% stock/-200% cash.
P2 180% stock/120% IT Treas./-300% cash.

Since Jan 2018 CAGR
P1 - 13.86%
P2 - 16.53%
PV Link since 2018-Rebalance quarterly

Since Jan 2007 CAGR
P1 - 13.08%
P2 - 17.89%
PV Link since 2007

The backtesting above is not fully accurate because you would be buying stocks every month. Added fixed contributions below.

Since Jan 2007 CAGR with adding $100 every month.
P1 - 21.37%
P2 - 24.46%
PV Link since 2007 adding $100 montly

User avatar
Topic Author
305pelusa
Posts: 903
Joined: Fri Nov 16, 2018 10:20 pm

Re: Lifecycle Investing - Leveraging when young

Post by 305pelusa » Mon Nov 18, 2019 1:13 pm

guyinlaw wrote:
Mon Nov 18, 2019 12:38 pm
305pelusa wrote:
Fri Mar 01, 2019 12:11 am

Nov 2019
Stock Exposure = 412k
Effective Debt = 233k
Equity in the exposure = 179k
Leverage = 2.29
Excellent thread. Thank you posting and sharing your experience.
305pelusa wrote:
Fri Mar 01, 2019 12:11 am
If the market drops, my current savings (invested in stocks) goes down. But all of my future income (bond-like) is untouched.
You have mentioned many times in the thread that you have no bonds in your portfolio and consider future income as a bond. You do not buy bonds with leverage (treasury futures) because the leverage needed is quite high.

The concern I have with this argument is that if stocks drop by >15% ( Dec 2018) or by >50% you will not have the future cash earning available to rebalance.

Consider the two portfolios
P1 200% stock/-200% cash.
P2 180% stock/120% IT Treas./-300% cash.

Since Jan 2018 CAGR
P1 - 13.86%
P2 - 16.53%
PV Link since 2018-Rebalance quarterly

Since Jan 2007 CAGR
P1 - 13.08%
P2 - 17.89%
PV Link since 2007

The backtesting above is not fully accurate because you would be buying stocks every month. Added fixed contributions below.

Since Jan 2007 CAGR with adding $100 every month.
P1 - 21.37%
P2 - 24.46%
PV Link since 2007 adding $100 montly
Don't you think those comparisons are unfair since interest rates went down over those time periods? I think it would be better to consider a period where rates and P/E multiple were similar at start and finish to control for those confounding, speculative factors. Oct 2015-Today is an example and it has a drawdown (Dec 2018), which is where you claim bonds shine by rebalancing into cheap stocks.

ebot
Posts: 4
Joined: Thu Nov 07, 2019 5:48 pm

Re: Lifecycle Investing - Leveraging when young

Post by ebot » Tue Nov 19, 2019 6:02 am

I think you mentioned previously that you don’t plan to rebalance. Are you planning to roll the leap at expiration? Also, how are you planning to manage the cash received from your day job? Are you keeping a cash cushion and re-evaluating every month / quarter, etc for when to purchase a new LEAP Option?

User avatar
Topic Author
305pelusa
Posts: 903
Joined: Fri Nov 16, 2018 10:20 pm

Re: Lifecycle Investing - Leveraging when young

Post by 305pelusa » Tue Nov 19, 2019 11:03 pm

ebot wrote:
Tue Nov 19, 2019 6:02 am
I think you mentioned previously that you don’t plan to rebalance. Are you planning to roll the leap at expiration? Also, how are you planning to manage the cash received from your day job? Are you keeping a cash cushion and re-evaluating every month / quarter, etc for when to purchase a new LEAP Option?
Whether I roll the LEAPs at expiration or not depends on whether I need leverage to hit my target exposure by Dec 2021 (expiration). I'm hoping that will not be the case but will see when I get there.
Cash received from my day job gets used to pay for my expenses. Whatever I save, I either invest in the market or use to pay down debt (deleverage).

I don't need any more leverage to hit my target stock exposure so I don't need to buy any more LEAPs for now. I don't keep a cash cushion. Any idling cash earns a better rate by paying back my debt. If I need money, I'll sell some holdings.

Post Reply