Time Diversification - bottom line, what do you do?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
investorguy1
Posts: 543
Joined: Mon Nov 24, 2014 7:13 pm

Time Diversification - bottom line, what do you do?

Post by investorguy1 » Wed Aug 17, 2016 7:24 pm

My understanding is that some people argue that time horizon doesn't reduce equity risk. I think when it comes down to it this argument doesn't have much particle value. Any feedback on this topic would be appreciated. I'd also like to know if you decrease your equity exposure as you get older and how do you do it and by how much (annually, every few years, never)?

When in or near retirement you pretty much have to reduce risk in your portfolio (unless you have tons of money) people on both sides of the debate would agree on this. If stocks crash in short term you are in trouble.

The question really comes into play when you are younger. In this case you are left with two options. If you say stocks are just as risky long term and therefore you shouldn't hold more your only option left is saving more. If you are already saving as much as you can this may not really be an option or be a very difficult one to achieve any how. So you are left with investing more aggressively. Could the market do poorly in the long term? Sure it can. But if you don't invest more aggressively you almost certainly won't reach your goal. So practically speaking I think most people have to invest more aggressively when they are younger.

Additionally when you are younger you probably have less money invested so you are putting less dollars at risk. This would be true even for someone who saved a lot when very young and isn't going to save much more. So even if stocks are risky long term the average guy doesn't have a choice. Would the person who doesn't believe in time diversification suggest being more aggressive later in life is equally risky? I just don't see it.

protagonist
Posts: 6004
Joined: Sun Dec 26, 2010 12:47 pm

Re: Time Diversification - bottom line, what do you do?

Post by protagonist » Wed Aug 17, 2016 8:45 pm

Being more aggressive in retirement is more risky for the obvious simple reason- you do not have the safety net of employment and a life of income ahead of you.

A 30 year old with, say, a secure, steady 50K/year job and 100K, or even 10K or 1K, in investments is still fine if he is 100% in stocks and the stock market loses 90% of his value. He will still probably make 50K/year or more for years to come (plus probably SS when he retires). There is plenty of time for him to potentially experience a market recovery. He could also train in other fields if he loses his job.

On the other hand, a retiree with 500K in investments, if 100% in stocks, will be down to his last 50K if the stock market loses 90%, and may have to sell his remaining stock for living expenses. All that is left is SS for income.

User avatar
Topic Author
investorguy1
Posts: 543
Joined: Mon Nov 24, 2014 7:13 pm

Re: Time Diversification - bottom line, what do you do?

Post by investorguy1 » Wed Aug 17, 2016 9:06 pm

protagonist wrote:Being more aggressive in retirement is more risky for the obvious simple reason
Agreed. But there are those who argue that over the long term stocks are just as risky.

http://www.vanguard.com/pdf/icrtd.pdf?2210045172

staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 9:40 am

Re: Time Diversification - bottom line, what do you do?

Post by staythecourse » Wed Aug 17, 2016 10:18 pm

Okay before someone quotes the raondom Norad cyber piece I'll throw my two cent in. Yes if you look at that random paper mathamatically since stocks have a higher dispersion of returns in the short run and you carry out those different possible returns over a 5, 10, or 20 yr. periods there will be compounding and further amplificiation of that dispersion. Meaning a 2% return is compouned out after 20 yrs. will be significantly lower then if a 8% return is compounded out for 20 yrs. Vice versa bonds which have a more narrow dispersin of returns will thus have a lower range of returns after compounding over those same 5, 10, or 20 yr. periods. That is just the laws of math. So a larger dispersion means larger S.D. Again just the laws of math.

Now a young investor shoiuld be asking is how often is a high equity portfolio going to outperform a more balanced or bond heavy portfolio.

For example: Lets say stocks after 20 yrs. have a higher S.D. then bonds but outperform the bond heavy portfolio 95% of the time. Mathamatically, the stocks are riskier (wider dispersion/ S.D.), but I am more then happy as they have a higher probability of outperforming. I have crunched the numbers and the chances of a high equity portfolio (80/20) outperforming a balanced portfolio (50/50) increases over 5, 10, 15, 20 yr. periods. Interestingly, the times they do underperform they lag less and less over longer time horizons. I did the crunching using DJ and Treasury notes from 1926 to current.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

freyj6
Posts: 216
Joined: Thu Jun 19, 2014 2:09 am

Re: Time Diversification - bottom line, what do you do?

Post by freyj6 » Wed Aug 17, 2016 11:00 pm

The argument generally made against holding 100% stocks when you're young is that it's psychologically challenging and a fair number of people are likely to lose a lot of money selling at bottom.

Psychology aside, there is little reason to believe that bonds will outperform stocks over the next 20 or 30 years.

Like others have mentioned, being 100% stocks near retirement is foolish unless you have so much money that it doesn't matter. This is because distribution portfolios are extremely vulnerable to volatility.

There's no set-in-stone path to decreasing equity allocation.

The most common is to decrease equity holdings gradually with age (you could look at the Vanguard Target Retirement portfolios).

You could also go off of market conditions. Personally if I was within 5 years of retirement I'd probably go 50/50 right now, but if we were 3 years into a bear market I'd be less likely to reduce stock allocation.

In other words, dramatically reduce risk if you have a high point within a few years of retirement.

Swelfie
Posts: 200
Joined: Mon Mar 14, 2016 12:54 am

Re: Time Diversification - bottom line, what do you do?

Post by Swelfie » Wed Aug 17, 2016 11:15 pm

I follow market_timer's philosophy in his infamous thread, but with a MUCH more cautious approach. My asset allocation is what I intend to keep in retirement, but I utilize modest leverage. I have modeled my standard deviation and reduced volatility in my portfolio as much as possible and come up with a maximum leverage I am comfortable with. I also reduce that leverage by not investing dividends or new contributions at any of a laundry list of negative economic indicators and ramp it slowly back up in the absence of those. As I approach my goal valuation for retirement I reduce this leverage. When my notional valuation hits my goal I will simply buy out of leverage at that point. When I am at my goal with 1x gearing, I'm ready to retire. This greatly increases risk in the early phases of investing, but the low monetary investment reduces the risk. At the later phases, the low leverage and low volatility portfolio greatly reduces risk, but the high valuation increases it. My goal is to have a constant risk factor over time, which isn't strictly possible because risk will increase prior to the notional value being attained, but from that point forward risk remains somewhat constant.

Understand though that this introduces a new type of risk: risk of ruin. That type of risk is not present in an unleveraged portfolio. This risk diminished as the portfolio value increases however, so were ruin to occur, it would most likely occur prior to the goal notional valuation when the portfolio is still small.

I do not recommend my approach. Read market_timer's thread and see how this can go very, very wrong. It takes a steady hand, massive research, an emergency strategy, constant monitoring, and balls of steel. And I still might go broke and have to start over.

freyj6
Posts: 216
Joined: Thu Jun 19, 2014 2:09 am

Re: Time Diversification - bottom line, what do you do?

Post by freyj6 » Thu Aug 18, 2016 12:49 am

Swelfie wrote:
Understand though that this introduces a new type of risk: risk of ruin. That type of risk is not present in an unleveraged portfolio. This risk diminished as the portfolio value increases however, so were ruin to occur, it would most likely occur prior to the goal notional valuation when the portfolio is still small.

I do not recommend my approach. Read market_timer's thread and see how this can go very, very wrong. It takes a steady hand, massive research, an emergency strategy, constant monitoring, and balls of steel. And I still might go broke and have to start over.
I think that's the key point here: deep risk.

If I were able to leverage stocks as to just have twice the volatility and twice the return, I'd do it in a second.

But with leverage, if you miscalculate, you can lose a big chunk of money permanently.

Ari
Posts: 538
Joined: Sat May 23, 2015 6:59 am

Re: Time Diversification - bottom line, what do you do?

Post by Ari » Thu Aug 18, 2016 2:39 am

investorguy1 wrote:When in or near retirement you pretty much have to reduce risk in your portfolio (unless you have tons of money) people on both sides of the debate would agree on this.
Well, you'd probably want to reduce risk if you're 100% stocks before retirement, sure. But I think rules of thumb like "Age minus ten in bonds" are needlessly conservative. I don't want to give advice to others on this, as people should set an asset allocation they are comfortable with, but personally I would not want to be less than maybe 75-80% stocks in retirement.

Here's an article by Wade Pfau: http://www.forbes.com/sites/wadepfau/20 ... b0f47d6924

I'll quote some relevant bits:
the general idea is that higher stock allocations tended to support higher withdrawal rates, with little in the way of downside risk. The SAFEMAX does not appear to be that much lower with higher stock allocations, though the potential for upside with higher stock allocations is quite striking as higher sustainable withdrawal rates are possible with all but the worst-case outcomes.
there is a sweet spot between about 35% stocks and 80% stocks where higher stock allocations have no discernable impact on the SAFEMAX. A 4% withdrawal rate tended to work no matter what stock allocation was chosen in this range. On the downside, retirees would have been just as well-off with 80% stocks as with 35% stocks.
Looking at the charts, it seems a 100% stock allocation could support a 3.5% SWD. That's hardly "tons of money", at least compared to the 4% which is usually suggested.
All in, all the time.

protagonist
Posts: 6004
Joined: Sun Dec 26, 2010 12:47 pm

Re: Time Diversification - bottom line, what do you do?

Post by protagonist » Thu Aug 18, 2016 12:01 pm

investorguy1 wrote:
protagonist wrote:Being more aggressive in retirement is more risky for the obvious simple reason
Agreed. But there are those who argue that over the long term stocks are just as risky.
I agree completely with that. Stocks are, of course, risky over the long term, since as a complex, nonlinear and chaotic system with high sensitivity to initial conditions ("butterfly effect"), the future becomes murkier and murkier as a function of time. That is definitely a factor. "Past performance" becomes less and less relevant with an increasing time frame. I have no idea what the market will look like 25 or 50 years from now, and even less so 100 or 1000 years from now, if a market even still exists. I can only guess and hope for the course of my lifetime.

But that said, the issue is not so much the compounding of risks in the market with time. It is the ability of a trained worker with skills and a profession to weather a huge financial loss, given a (hopefully) reliable income that will cover his expenses, and lots of time to recuperate his loss with earnings. This is not the case for a retiree dependent on his life savings to get by. That is a concrete reality. It is not so much of an issue for a retiree with a (hopefully) reliable pension large enough to cover his needs if his investments tank- that person can afford to take large risks with his investments.

Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: Time Diversification - bottom line, what do you do?

Post by Rodc » Thu Aug 18, 2016 6:30 pm

freyj6 wrote:The argument generally made against holding 100% stocks when you're young is that it's psychologically challenging and a fair number of people are likely to lose a lot of money selling at bottom.

Psychology aside, there is little reason to believe that bonds will outperform stocks over the next 20 or 30 years.
That is one point.

Another is many people end up needing to raid their retirement accounts early because of a prolonged job lose, which also tends to happen when stocks are down (say in a recession). You can mitigate this to some degree with a healthy emergency fund, but of course then you are not super heavy in stocks. You might be super heavy stocks in your retirement accounts, but if you have a couple of years in an emergency fund (cash, bonds, etc.) overall you have a stock/fixed income portfolio.

It is also worth noting that from starting to save for retirement, say age 25, to retirement, say age 65, if you put away the same amount each year the first year of saving is held for 40 years, the second for 39, etc. On average the holding period is 20 years. Given inflation and hopefully you have some career growth, most people are putting in larger amounts later in life so the average holding period is rather less than 20 years. Of course when you retire you do not stop investing so this is a simplification, but most have by then long since started selling off stocks to tilt more towards bonds, so perhaps not so far off. In the end your average holding period may be rather shorter than it at first may seem. So the issue is not so much will stocks out perform bonds over 30 or 40 years, but will they out perform bonds over maybe 15 years or even less (say considering someone age 40 trying to figure out an allocation, not someone 25).
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.

Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: Time Diversification - bottom line, what do you do?

Post by Rodc » Thu Aug 18, 2016 6:37 pm

investorguy1 wrote:My understanding is that some people argue that time horizon doesn't reduce equity risk. I think when it comes down to it this argument doesn't have much particle value. Any feedback on this topic would be appreciated. I'd also like to know if you decrease your equity exposure as you get older and how do you do it and by how much (annually, every few years, never)?

When in or near retirement you pretty much have to reduce risk in your portfolio (unless you have tons of money) people on both sides of the debate would agree on this. If stocks crash in short term you are in trouble.

The question really comes into play when you are younger. In this case you are left with two options. If you say stocks are just as risky long term and therefore you shouldn't hold more your only option left is saving more. If you are already saving as much as you can this may not really be an option or be a very difficult one to achieve any how. So you are left with investing more aggressively. Could the market do poorly in the long term? Sure it can. But if you don't invest more aggressively you almost certainly won't reach your goal. So practically speaking I think most people have to invest more aggressively when they are younger.

Additionally when you are younger you probably have less money invested so you are putting less dollars at risk. This would be true even for someone who saved a lot when very young and isn't going to save much more. So even if stocks are risky long term the average guy doesn't have a choice. Would the person who doesn't believe in time diversification suggest being more aggressive later in life is equally risky? I just don't see it.
First, there is no meaningful conversation until you carefully define what you personally mean by risk.

I would also note that when young you have little invested so your allocation is hardly here or there - it is not going to drive success or failure. Your savings rate is the big driver. Over time allocation matters more and more, and by the time it matters a ton you are starting to run out of time for "time diversification".

One thing you seem to be missing is that when young you have more time to recover if risk shows up by saving more, working longer, trying to get a better job or work two jobs. Regardless of what one thinks of so-called time diversification this is likely a rather bigger issue.

These last two issues are likely rather more important than any so-called time diversification, it seems to me.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.

User avatar
Dutch
Posts: 1277
Joined: Thu Jun 27, 2013 2:12 pm

Re: Time Diversification - bottom line, what do you do?

Post by Dutch » Thu Aug 18, 2016 6:50 pm

Rodc wrote:First, there is no meaningful conversation until you carefully define what you personally mean by risk.
Under-performing a riskless asset. That's the only risk definition that makes sense in this case. The infamous "Norstadt paper" used standard deviation as the definition of risk, and I never understood why that would apply.

Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: Time Diversification - bottom line, what do you do?

Post by Rodc » Thu Aug 18, 2016 7:26 pm

Dutch wrote:
Rodc wrote:First, there is no meaningful conversation until you carefully define what you personally mean by risk.
Under-performing a riskless asset. That's the only risk definition that makes sense in this case. The infamous "Norstadt paper" used standard deviation as the definition of risk, and I never understood why that would apply.
I would think you would want to take into account how far under the riskless asset; missing by $1M is not the same as missing by $1K. Or maybe vs a low risk asset as one may know they have no interesting a riskless asset. And which riskless asset? Cash, nominal 30 year bonds, 30 year TIPS; each are riskless in different contexts.

I really think one needs to carefully consider what they mean.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.

User avatar
Dutch
Posts: 1277
Joined: Thu Jun 27, 2013 2:12 pm

Re: Time Diversification - bottom line, what do you do?

Post by Dutch » Thu Aug 18, 2016 8:16 pm

Rodc wrote:
Dutch wrote:
Rodc wrote:First, there is no meaningful conversation until you carefully define what you personally mean by risk.
Under-performing a riskless asset. That's the only risk definition that makes sense in this case. The infamous "Norstadt paper" used standard deviation as the definition of risk, and I never understood why that would apply.
I would think you would want to take into account how far under the riskless asset; missing by $1M is not the same as missing by $1K. Or maybe vs a low risk asset as one may know they have no interesting a riskless asset. And which riskless asset? Cash, nominal 30 year bonds, 30 year TIPS; each are riskless in different contexts.

I really think one needs to carefully consider what they mean.
The relative amount of under-performance is the measure of risk of course.

As far as which riskless asset: Just pick the highest performing one, in hindsight. You can only do this analyses on historical data after all.

Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: Time Diversification - bottom line, what do you do?

Post by Rodc » Thu Aug 18, 2016 9:03 pm

Dutch wrote:
Rodc wrote:
Dutch wrote:
Rodc wrote:First, there is no meaningful conversation until you carefully define what you personally mean by risk.
Under-performing a riskless asset. That's the only risk definition that makes sense in this case. The infamous "Norstadt paper" used standard deviation as the definition of risk, and I never understood why that would apply.
I would think you would want to take into account how far under the riskless asset; missing by $1M is not the same as missing by $1K. Or maybe vs a low risk asset as one may know they have no interesting a riskless asset. And which riskless asset? Cash, nominal 30 year bonds, 30 year TIPS; each are riskless in different contexts.

I really think one needs to carefully consider what they mean.
The relative amount of under-performance is the measure of risk of course.

As far as which riskless asset: Just pick the highest performing one, in hindsight. You can only do this analyses on historical data after all.
We invest forward, not backward.

Point remains. If OP or anyone else wants a meaningful answer they need to provide a carefully defined question.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.

lemonPepper
Posts: 89
Joined: Wed Feb 18, 2015 7:43 pm

Re: Time Diversification - bottom line, what do you do?

Post by lemonPepper » Fri Aug 19, 2016 4:34 pm

do you mean time diversification as a way to reduce sequence of returns risk?

User avatar
Topic Author
investorguy1
Posts: 543
Joined: Mon Nov 24, 2014 7:13 pm

Re: Time Diversification - bottom line, what do you do?

Post by investorguy1 » Fri Aug 19, 2016 5:14 pm

Rodc wrote: If OP or anyone else wants a meaningful answer they need to provide a carefully defined question.
My question is, should I gradually reduce the equity portion of my portfolio over time? Or should I do what I am doing now which is, maintain my asset allocation which is based on the maximum I'd be comfortable with losing in any given year (estimated at 50% of the equities in my portfolio).

There has been some discussion about the definition of risk. I don't have an exact definition of risk but I think it includes the probability, frequency and magnitude of losing money. It also includes the chance of a complete loss or long term loss. I think under performance compared to other assets, inflation and other investors are also important factors. But I think the most important factors for me is to reduce the chances of big short term loses or long term poor performance.

User avatar
Topic Author
investorguy1
Posts: 543
Joined: Mon Nov 24, 2014 7:13 pm

Re: Time Diversification - bottom line, what do you do?

Post by investorguy1 » Fri Aug 19, 2016 5:25 pm

lemonPepper wrote:do you mean time diversification as a way to reduce sequence of returns risk?
"Time diversification, the phrase used to refer to the concept that investments in stocks are less risky over longer periods than
shorter ones"-http://www.vanguard.com/pdf/icrtd.pdf?2210045172

Most people believe that holding stocks for a longer period of time is safer that holding them for a shorter period of time. There are those who argue that stocks are not less risky over longer periods of time. The conclusion I would draw from this is to maintain a constant allocation to equities. However there would still be a factor which would lead me to gradually reduce equity over time which is hopefully having more money later than I do now. In which case even though I would have the same allocation I would have more dollars at risk. From a risk tolerance stand point I don't really see myself needing to lower my risk though.

Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: Time Diversification - bottom line, what do you do?

Post by Rodc » Fri Aug 19, 2016 5:29 pm

investorguy1 wrote:
Rodc wrote: If OP or anyone else wants a meaningful answer they need to provide a carefully defined question.
My question is, should I gradually reduce the equity portion of my portfolio over time? Or should I do what I am doing now which is, maintain my asset allocation which is based on the maximum I'd be comfortable with losing in any given year (estimated at 50% of the equities in my portfolio).

There has been some discussion about the definition of risk. I don't have an exact definition of risk but I think it includes the probability, frequency and magnitude of losing money. It also includes the chance of a complete loss or long term loss. I think under performance compared to other assets, inflation and other investors are also important factors. But I think the most important factors for me is to reduce the chances of big short term loses or long term poor performance.
I think every few years you should carefully examine your goals, ability to handle risk both emotionally and as far as saving more if needed to get back on track, consider how your career is going and how viable your job is, do you think income will go up, down stay the same, and do you want to try to retire earlier or are you having fun and maybe work longer and any other assumptions. Adjust goals and plans as needed.

Then set an appropriate allocation. You can not know today even simple things like will you be well ahead of where you want to be (could dial back risk) or if you think you will be forced into early retirement (dial back risk), or if the market is down (maybe increase stocks or savings rate, get a second job).

If young and far from retirement, maybe just steady as she goes. If closing in on retirement and have as much as you need, dial back.

I don't think generic answers are the right answers.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.

User avatar
Topic Author
investorguy1
Posts: 543
Joined: Mon Nov 24, 2014 7:13 pm

Re: Time Diversification - bottom line, what do you do?

Post by investorguy1 » Sun Aug 21, 2016 1:44 pm

Rodc wrote:
I think every few years you should carefully examine your goals, ability to handle risk both emotionally and as far as saving more if needed to get back on track, consider how your career is going and how viable your job is, do you think income will go up, down stay the same, and do you want to try to retire earlier or are you having fun and maybe work longer and any other assumptions. Adjust goals and plans as needed.

Then set an appropriate allocation. You can not know today even simple things like will you be well ahead of where you want to be (could dial back risk) or if you think you will be forced into early retirement (dial back risk), or if the market is down (maybe increase stocks or savings rate, get a second job).

If young and far from retirement, maybe just steady as she goes. If closing in on retirement and have as much as you need, dial back.

I don't think generic answers are the right answers.
Those are some excellent points and gave me a lot of food for thought. Lets say all other factors stay the same other than being one year closer to retirement should I lower my equity exposure or not?

Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: Time Diversification - bottom line, what do you do?

Post by Rodc » Sun Aug 21, 2016 3:27 pm

investorguy1 wrote:
Rodc wrote:
I think every few years you should carefully examine your goals, ability to handle risk both emotionally and as far as saving more if needed to get back on track, consider how your career is going and how viable your job is, do you think income will go up, down stay the same, and do you want to try to retire earlier or are you having fun and maybe work longer and any other assumptions. Adjust goals and plans as needed.

Then set an appropriate allocation. You can not know today even simple things like will you be well ahead of where you want to be (could dial back risk) or if you think you will be forced into early retirement (dial back risk), or if the market is down (maybe increase stocks or savings rate, get a second job).

If young and far from retirement, maybe just steady as she goes. If closing in on retirement and have as much as you need, dial back.

I don't think generic answers are the right answers.
Those are some excellent points and gave me a lot of food for thought. Lets say all other factors stay the same other than being one year closer to retirement should I lower my equity exposure or not?
How far from retirement? If you are a long way from retirement your ability to bounce back, adjust plans, know your ultimate goals hardly change year to year. Consider if you are 30 years out. In reality you don't know it is really 30, might end up being 25 or 35 (unless you are something like an air traffic controller with well defined age out policies). A year later you are at 29+/-5 or so. Plus or minus 5 years is like +/- 17% in uncertainly in one very important variable. In contrast being one year closer is about a 3% change. Not really much change as far as ability to adjust, how well defined your ultimate income goals, etc. Personally I would just hold a constant plan. Others adjust stocks by 1% (which is meaningless in that one year, but does add up if you continue to change 1% every year for a couple of decades). The decision to hold steady or shift by a percent or something similar just does not really matter. Use whichever one appeals to you.

On the other hand if you are 5 years out, now you likely have a very good idea of what income you want. You are really dialing in the year you will retire. You have very little room to make adjustments (though you could depending on the situation simply keep working, ie change 5 years to something else). One year is 20% of the time remaining. If markets crash now you have very little time to hope they bounce back. In this range you should be very carefully evaluating the situation fairly often. If nothing changed, I would still think the plan you had last year should be pretty good this year so would not need a huge change year to year.

In the end though, I stick with what I said. Instead of hypothetical "what if nothing changed" situations to make a plan years ahead of time, just reevaluate as needed. If young and nothing much is changing maybe rework the plan every 5 or even every 10 years. Middle aged maybe every 3-5 years. If nearing retirement keep an eye on things yearly. If nothing much is changing you may not make any changes after checking the plan. Also if you have a big life event you might want to check the plan. You get downsize or think you might - could time to reevaluate. Have an extra kid. Start a new career.

And this check up is more than allocation. Do you have enough liquid savings? Do you have enough saved for college for the kids? Do you have the right insurance and the right amount? Are you spending your money on the things that are important to you, or do you need to adjust what you are spending money on?
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.

protagonist
Posts: 6004
Joined: Sun Dec 26, 2010 12:47 pm

Re: Time Diversification - bottom line, what do you do?

Post by protagonist » Sun Aug 21, 2016 6:52 pm

risk
risk/Submit
noun
1.
a situation involving exposure to danger.
"flouting the law was too much of a risk"
verb
1.
expose (someone or something valued) to danger, harm, or loss.

To me, "risk" in investing is the sum of many different things at the same time. There is the risk of losing a substantial portion of my money, due to a market crash, a cyberattack, political unrest, whatever. There is the risk of having to sell stock in a down market. There is the risk of the anxiety caused by volatility. There is the risk of missing out on something better. There is the risk of not keeping up with inflation. There is the risk of dying and never getting to enjoy your money. There are probably other risks as well.

So when I assess "risk" in investing and whether a particular course of action justifies it, I try to take everything into account, just as I would if I was assessing the risk of a new job, a move, a marriage, or any other action. A broad definition is appropriate, I would think, for most investors.

Unfortunately, many of the factors that comprise investing risk are not calculable....just as with a job or a marriage. So assessing risk is, to some extent (i would argue to a large extent), guesswork. You can sometimes assess relative risk.....eg the relative risk of being 100% in stocks if you are retired with no other income is greater than if you are 30 years old and a working physician with no debt. And the relative risk of losing money is greater if you are paying 5% annual management fees than if you are paying 0.05%.

User avatar
vitaflo
Posts: 1170
Joined: Sat Sep 03, 2011 3:02 pm

Re: Time Diversification - bottom line, what do you do?

Post by vitaflo » Sun Aug 21, 2016 8:24 pm

investorguy1 wrote: My question is, should I gradually reduce the equity portion of my portfolio over time? Or should I do what I am doing now which is, maintain my asset allocation which is based on the maximum I'd be comfortable with losing in any given year (estimated at 50% of the equities in my portfolio).

There has been some discussion about the definition of risk. I don't have an exact definition of risk but I think it includes the probability, frequency and magnitude of losing money. It also includes the chance of a complete loss or long term loss. I think under performance compared to other assets, inflation and other investors are also important factors. But I think the most important factors for me is to reduce the chances of big short term loses or long term poor performance.
Many people write about risk and how it influences numbers ("big short term loses or long term poor performance"). But there are other risks, to me the largest is behavioral risk. If your current AA is "based on the maximum I'd be comfortable with losing in any given year", then you're properly managing behavioral risk. If you go more into stocks, you not only take on more equity risk, but also more behavioral risk. It's not just about numbers, we're not robots.

The other thing I think people forget about, and Rdoc sorta pointed this out, is that other form of equity, sweat equity. You have more control over your work, how much of it you do and how far you want to advance, than you do the market. You also have more control over your spending then you do the market. Both of these things are just as important (perhaps more so) than any AA you can come up with.

I think the reason we post so much about AA's, numbers, et al, is because we have so little control over them. We're trying to will them to our command somehow. Instead focus on the things you do have control over, your behavior, your sweat equity and your spending levels. IMO they will have a much more profound impact.

KlangFool
Posts: 13931
Joined: Sat Oct 11, 2008 12:35 pm

Re: Time Diversification - bottom line, what do you do?

Post by KlangFool » Sun Aug 21, 2016 8:57 pm

OP,

A very simplistic view of this problem.

1) Risk = not having enough money to spend when you need it.

This could be any time before, during, and after retirement. For example, if recession shows up tomorrow, you lose your job and income plus stock market drop 50%, can you survive for at least 2 years? Will you lock in your loss permanently by selling all your stock since you need the money?

2) Stock = you can lose 50% of it at any time.

3) We have no idea when we will be long-term unemployed or under-employed. Aka, we do not know when we will be forced to retire.

4) 50% of stock = X number of years of saving. Do we have enough time to save and replace the money? Given (3), how do we know how much time that we have?

KlangFool

protagonist
Posts: 6004
Joined: Sun Dec 26, 2010 12:47 pm

Re: Time Diversification - bottom line, what do you do?

Post by protagonist » Sun Aug 21, 2016 10:08 pm

vitaflo wrote: I think the reason we post so much about AA's, numbers, et al, is because we have so little control over them. We're trying to will them to our command somehow. Instead focus on the things you do have control over, your behavior, your sweat equity and your spending levels. IMO they will have a much more profound impact.
+1

I would also add debt. That is something that can be controlled, and is the downfall of so many people.

Rob Bertram
Posts: 823
Joined: Mon May 05, 2014 12:15 pm

Re: Time Diversification - bottom line, what do you do?

Post by Rob Bertram » Mon Aug 22, 2016 11:41 am

I am surprised that no one has yet mentioned Ayres and Nalebuf's article Mortgage Your Retirement and book Lifecycle Investing: A New, Safe, and Audacious Way to Improve the Performance of Your Retirement Portfolio. Our fellow forum member Market Timer was a case study in the book. Ayres and Nalebuff argue that young investors should take on deep risk early in their investing/accumulation phase and slowly reduce risk as they approach their "number".

While I disagree with their approach (starting with 2x leverage on stocks with no bonds), I believe that their reasoning is sound. And I believe that we Bogleheads often tell young investors the same thing -- that any asset allocation they pick (including all stocks) will be dwarfed by their savings rate in the beginning -- and that watching $10k drop by half to $5k is very different emotionally from watching $2m drop to $1m.

Similar to Swelfie, I have a leveraged balanced portfolio. And I would argue that anyone who wants to take on more risk should keep a balanced portfolio and not tilt towards a specific market (stocks), sector, or factor.

Swelfie
Posts: 200
Joined: Mon Mar 14, 2016 12:54 am

Re: Time Diversification - bottom line, what do you do?

Post by Swelfie » Mon Aug 22, 2016 1:56 pm

Rob Bertram wrote:Similar to Swelfie, I have a leveraged balanced portfolio. And I would argue that anyone who wants to take on more risk should keep a balanced portfolio and not tilt towards a specific market (stocks), sector, or factor.
I have been heavily influenced by those papers, as well as market_timer (more as a case study of how NOT to do this, but amazingly educational), yourself Rob, and Larry Swedrow. I do now strongly believe that a leveraged low volatility portfolio is the best way to balance temporal risk. I do not like the idea of risking piddly amounts early in life and huge amounts later. Extreme volatility or even risk of ruin with a $30k portfolio is nothing compared with 20% fluctuations in a $5MM portfolio. Modifying AA over time is one way to do this with no ROR but I dont see it as being nearly as clean or efficient or effective as simply employing leverage.

I do however employ tilt. I see factor tilt as a way of diversifying non correlated returns to reduce volatility. To that end I tilt rather mildly towards small, quality, momentum, tips and commodities and moderately towards value. Back testing shows that I can significantly decrease volatility and get a better Sorting ratio with tilts, which helps me sleep at night with a leveraged portfolio.

But thank you Rob for your enlightening little experiment. Have learned more from your musings than most anything on this board, save Larry.

Day9
Posts: 820
Joined: Mon Jun 11, 2012 6:22 pm

Re: Time Diversification - bottom line, what do you do?

Post by Day9 » Mon Aug 22, 2016 2:41 pm

Theoretically choosing between leveraging a balanced portfolio and using factor tilting is not an either/or decision.

But practically it may be more of an either/or decision than it is in theory because perhaps the best way to implement the leveraged strategy would be with S&P 500 futures as the domestic, or entire stock allocation in your portfolio.

I would like to explore the idea of getting your leverage on the bond side by using treasury futures (like Rob Bertram does) while holding factor tilting stock funds and ETFs. Maybe something like 90% of your portfolio in factor tilted stock funds and ETFs and 10% as the collateral for a leveraged treasury bond position using futures?
I'm just a fan of the person I got my user name from

Rob Bertram
Posts: 823
Joined: Mon May 05, 2014 12:15 pm

Re: Time Diversification - bottom line, what do you do?

Post by Rob Bertram » Mon Aug 22, 2016 4:26 pm

For moderate levels of leverage, you can likely achieve a factor tilt by using an ETF while keeping S&P 500 futures for your stock position and treasury futures for your bonds. You do, however, get into practical limitations where factor ETFs have higher maintenance margin than your futures.

In order to have a productive conversation around leverage, we really need to get back to basics and look at tilting from a different perspective. In most of the traditional tilts, people are trying to increase returns by adding a volatile and risky asset to the portfolio -- like REITs or SCV. It lowers the risk-adjusted return of the portfolio, but the investor is reaching for return and is willing to take on more risk.

When we allow leverage in our portfolio, we can dial risk up to taste without needing to take uncompensated risk. So when we start to develop a portfolio to lever, I would start with the market portfolio (both stocks and bonds) and then tilt in ways that reduce down-side volatility but keeps the same risk-adjusted return.

Day9
Posts: 820
Joined: Mon Jun 11, 2012 6:22 pm

Re: Time Diversification - bottom line, what do you do?

Post by Day9 » Mon Aug 22, 2016 6:04 pm

Rob Bertram wrote:... In order to have a productive conversation around leverage, we really need to get back to basics and look at tilting from a different perspective. In most of the traditional tilts, people are trying to increase returns by adding a volatile and risky asset to the portfolio -- like REITs or SCV. It lowers the risk-adjusted return of the portfolio, but the investor is reaching for return and is willing to take on more risk.

When we allow leverage in our portfolio, we can dial risk up to taste without needing to take uncompensated risk. ...
Tilters argue just the opposite. Diversifying your sources of risk improves the risk-adjusted return of the portfolio. Swedroe famously argues that exposing your portfolio to more than just beta market risk and gaining exposure to other factors such as small and value can allow you to lower your overall stock allocation and keep the same expected return and improve risk-adjusted returns. He shows evidence that by some metrics, the value premium is even more reliable than beta market premium. I apologize if this rehashes discussion that appears on this forum over and over again.
For moderate levels of leverage, you can likely achieve a factor tilt by using an ETF while keeping S&P 500 futures for your stock position and treasury futures for your bonds. You do, however, get into practical limitations where factor ETFs have higher maintenance margin than your futures.
Thanks for this. I didn't think of the difference in maintenance margin requirements between S&P 500 futures and factor ETFs. I was thinking of a scenario where you keep your factor ETFs in a separate cash account, and hold treasury futures in another margin account, and for sake of clarity of my example, at a different broker. I am a novice at this but I believe the flaw is that you would have to add more money to your treasury future margin position more often to avoid margin call because there would be no stock investments in that account to get a diversification benefit.
I'm just a fan of the person I got my user name from

Swelfie
Posts: 200
Joined: Mon Mar 14, 2016 12:54 am

Re: Time Diversification - bottom line, what do you do?

Post by Swelfie » Mon Aug 22, 2016 11:02 pm

Day9 wrote:
Rob Bertram wrote:... In order to have a productive conversation around leverage, we really need to get back to basics and look at tilting from a different perspective. In most of the traditional tilts, people are trying to increase returns by adding a volatile and risky asset to the portfolio -- like REITs or SCV. It lowers the risk-adjusted return of the portfolio, but the investor is reaching for return and is willing to take on more risk.

When we allow leverage in our portfolio, we can dial risk up to taste without needing to take uncompensated risk. ...
Tilters argue just the opposite. Diversifying your sources of risk improves the risk-adjusted return of the portfolio. Swedroe famously argues that exposing your portfolio to more than just beta market risk and gaining exposure to other factors such as small and value can allow you to lower your overall stock allocation and keep the same expected return and improve risk-adjusted returns. He shows evidence that by some metrics, the value premium is even more reliable than beta market premium. I apologize if this rehashes discussion that appears on this forum over and over again.
For moderate levels of leverage, you can likely achieve a factor tilt by using an ETF while keeping S&P 500 futures for your stock position and treasury futures for your bonds. You do, however, get into practical limitations where factor ETFs have higher maintenance margin than your futures.
Thanks for this. I didn't think of the difference in maintenance margin requirements between S&P 500 futures and factor ETFs. I was thinking of a scenario where you keep your factor ETFs in a separate cash account, and hold treasury futures in another margin account, and for sake of clarity of my example, at a different broker. I am a novice at this but I believe the flaw is that you would have to add more money to your treasury future margin position more often to avoid margin call because there would be no stock investments in that account to get a diversification benefit.
I am one of those that argue the opposite. I do not believe that my factors are increasing my risk to provide higher return. Rather, I believe they are diversifying the source of my returns away from pure market beta.

I also do not hold my futures in separate accounts from my ETFs, nor do I worry about one account's margin vs. another. I may go to extreme margin in one account with no margin at all in another. I don't have a problem with this because if I were faced with a margin call I would simply close a position in my overly margined account and reopen it in the low margin account. I do maintain some balance here though to avoid excess trading.

Personally, I use:
^ZF/^ZT for treasuries. I park at the steepest slope of the yield curve and roll quarterly. I supplement with SCHR and SCHO. Mostly in traditional 401k.
^ES/^RLV/^EMD/^SMC for a TSM with small and value tilt. I supplement with VTI. Mostly in ROTH.
^MXEA for developed. I supplement with VEA. Mostly in ROTH.
^MXEF for emerging. I supplement with VWO or VWO single stock futures. Mostly in ROTH.
^ZB for long term treasuries. I Supplement with TLO. Mostly in traditional.
^AIGCI for commodities. No supplementation since the futures only control $8k, I just round off. Mostly in ROTH.
SCHP for TIPS. Traditional.
MTUM for Momentum. ROTH/HSA.
QUAL for Quality. Traditional.
VSS for international small tilt. ROTH/HSA.

I only hold VEA and VTI in taxable. Futures are all in tax deferred to avoid realizing gains and missing out on TLH. I use reg-t margin accounts but I will switch to portfolio when I hit the thresholds. My tax deferred are reg-t, and I won't likely switch to portfolio because I take on no debt in those accounts (all leverage is futures).

I park any remaining money that is covering futures leverage, except for a buffer, in BIL, SHV or SCHO depending on where the yield curve is steepest. I count this as cash and not against my leverage and will sell this if needed to meet margin before moving contracts around. I'll also use 2x leveraged ETFs if I need to fit under my rebalance bands and the contracts are too large (^ZT and ^ZB are pretty large contracts) and rebalance frequently to avoid a lot of tracking error.

Rob Bertram
Posts: 823
Joined: Mon May 05, 2014 12:15 pm

Re: Time Diversification - bottom line, what do you do?

Post by Rob Bertram » Tue Aug 23, 2016 11:51 am

We might have to agree to disagree about factor tilting. William Sharpe proved that under simplified market assumptions the market portfolio (cap weighted) is the one and only super-efficient portfolio from a mean-vairance perspective. This is the core reason why Bogleheads advocate total market index funds. (Simplicity is another reason.) Sharpe used these same assumptions to build the Capital Asset Pricing Model (CAPM) which is the framework for modeling a leveraged portfolio.

Now, it is fair to reject his assumptions or claim that variance is not a true measure of risk. (In my mind, both are justifiable arguments.) However, doing so means that you need to build a different framework to model your leveraged portfolio. From what I understand of Fama and French's factor model, they take issue with Sharpe's assumptions (markets are not very efficient, thus certain factors are incorrectly priced). You also need to take their warning about factor investing -- that it could take years or decades for a tilt to beat the total market. (And as Nisiprius often points out, small-cap value in some cases underperformed the market for 15 years.)

I don't necessarily have a problem with factor models. I, however, have a strong word of caution about mixing models. You need to understand the differences between the models that you use -- for example, taking a 5-factor optimized portfolio and plugging it into CAPM to leverage. You may seriously underestimate the risk you are taking.

Now, switching the conversation back to leverage. There are some practical issues that you need to consider. For starters, things like margin call are potential risks that will force you to realize losses long before your portfolio goes to zero. This forces you to look at and better understand your left-tail (downside) risk. The standard deviation of bond returns tends to be positively skewed (upside), while stocks tend to be negatively skewed (downside). Also, keeping a floor (target) on your leverage ratio will allow you to compound your gains, while sitting at a loss will prevent compounding. So things like small modest gains and faster recovery from a drop have a major impact on your returns. This favors things like short-term bonds.

Swelfie
Posts: 200
Joined: Mon Mar 14, 2016 12:54 am

Re: Time Diversification - bottom line, what do you do?

Post by Swelfie » Tue Aug 23, 2016 12:50 pm

Rob Bertram wrote:So things like small modest gains and faster recovery from a drop have a major impact on your returns. This favors things like short-term bonds.
Yes, I do disagree with you on factor investing. I see the factors as their own asset classes, each with their own characteristics. I prefer to diversify across them to not attain superior return (I'm doing that through leverage) but rather to get a diversification of return through less correlated sources. This should reduce overall volatility (supported through a lot of my back testing), but of course, it's all theory and the market has a way of misbehaving. I also am, like you, extremely bond heavy (73%) for the same reasons. I'm not willing to go anywhere near the leverage you are though because of my increase on intermediate and long term bonds over your portfolio, not to mention I have more than twice the equities and commodities that you do.

One thing I'm not quite clear on with your portfolio, and maybe this belongs in the other thread. Why high leverage on short term treasuries instead of moderate on more intermediate term? As I understand it, and you made a statement to this affect, leveraging Treasuries effectively increases their duration, amplifying the effect of yield changes. Would not any Treasury do then? This is why I choose to switch it up and chase curve roll yield and vary my leverage, hoping to gain some extra resilience against rate hikes by hitting the steepest slope and roll down rather than park at 2 years (which I'm doing anyway right now since the 2 year slope is better than the 3, 5, 7 or 10). Do you feel that a leveraged 2 year will recover faster than a less leveraged 5?

Swelfie
Posts: 200
Joined: Mon Mar 14, 2016 12:54 am

Re: Time Diversification - bottom line, what do you do?

Post by Swelfie » Tue Aug 23, 2016 3:23 pm

And also, wouldn't the cost of carry eat up a smaller proportion of your yield since both your yield is higher and you are less leveraged in a longer term Treasury?

Rob Bertram
Posts: 823
Joined: Mon May 05, 2014 12:15 pm

Re: Time Diversification - bottom line, what do you do?

Post by Rob Bertram » Tue Aug 23, 2016 4:20 pm

Swelfie wrote:One thing I'm not quite clear on with your portfolio, and maybe this belongs in the other thread. Why high leverage on short term treasuries instead of moderate on more intermediate term? As I understand it, and you made a statement to this affect, leveraging Treasuries effectively increases their duration, amplifying the effect of yield changes. Would not any Treasury do then?
Nisiprius shows in this thread that from a mean-variance optimization perspective, bond duration doesn't matter. He demonstrates that stocks/long-term bonds have the same optimal Sharpe ratio as stocks/intermediate term bonds. By extension, any increased returns from stocks/LT bonds can be obtained by leveraging a stock/IT bond or stock/ST bond portfolio.

However, variance is only one aspect of the portfolio. Increasing leverage magnifies variance, but it doesn't change the bond duration characteristics of the notional portfolio. The portfolio with short-term bonds is still less sensitive to interest rate changes. And this plays back into the bit I said about small modest gains are very important when leverage is involved as those gains are also leveraged and compounded. At least, with my leverage strategy, I buy more assets when the portfolio appreciates so that I get back to my target leverage ratio.
Swelfie wrote:This is why I choose to switch it up and chase curve roll yield and vary my leverage, hoping to gain some extra resilience against rate hikes by hitting the steepest slope and roll down rather than park at 2 years (which I'm doing anyway right now since the 2 year slope is better than the 3, 5, 7 or 10). Do you feel that a leveraged 2 year will recover faster than a less leveraged 5?
Well, I can't predict what the yield curve or (the Fed) will do. If it inverts, leverage gets ugly. But in general, duration does not change when you leverage bonds. Futures financing gets a little tricky. Expectations are already factored into the contract price. My analysis on treasury futures suggest that 2-year treasury futures recover faster that 5- and 10-year futures.

Rob Bertram
Posts: 823
Joined: Mon May 05, 2014 12:15 pm

Re: Time Diversification - bottom line, what do you do?

Post by Rob Bertram » Tue Aug 23, 2016 4:58 pm

Swelfie wrote:And also, wouldn't the cost of carry eat up a smaller proportion of your yield since both your yield is higher and you are less leveraged in a longer term Treasury?
It nets out to not making a difference because they both lie on the same capital asset line. Since both portfolios have the same Sharpe ratio (risk-adjusted return), CAPM tells me that leveraging one portfolio to match the standard deviation of the other will also result in identical returns after borrowing costs.

Swelfie
Posts: 200
Joined: Mon Mar 14, 2016 12:54 am

Re: Time Diversification - bottom line, what do you do?

Post by Swelfie » Tue Aug 23, 2016 4:58 pm

Rob Bertram wrote:The portfolio with short-term bonds is still less sensitive to interest rate changes.
This is the part I question. A $1000 notional bond of a specific duration geared at 1x will be less effected by a change in yield than a $2000 notional bond geared at 2x. Obviously the hit from a yield hike will be twice as hard and make the bond act as if it had a longer duration. What you are implying is that as you increase leverage you increase the yield of the bond faster than the effective duration, so that if the duration of the bond in yield change sensitivity made a 2 year bond act like a 5 year bond due to leverage, the actual yield from that 5 year bond would be more akin to a 7 year bond. I'm not sure I buy that, otherwise why would anyone buy 5 year bonds? We would just by 4 week Treasuries and leverage them to act like 30 year with a guaranteed better return than 30 year Treasuries. I may have to sit down and do some calculations. What I might buy is that you are doing something similar to me and taking advantage of a lump in the yield curve. There is one at 2 years now and has been for a bit (actually at 6 months, but it spills over into 2 years) and you might be seeing superior returns over five years because your roll yield is so much better at 2, so as long as this shape persists a 2 year levered to look like a 5 will yield better than a 5.
Rob Bertram wrote:Well, I can't predict what the yield curve or (the Fed) will do. If it inverts, leverage gets ugly. But in general, duration does not change when you leverage bonds. Futures financing gets a little tricky. Expectations are already factored into the contract price. My analysis on treasury futures suggest that 2-year treasury futures recover faster that 5- and 10-year futures.
I can't predict what the yield curve will do, but I can predict what it looks like right now and roll into the new yield curve in 90 days (or sooner if I feel like it). I can, at the same time, adjust my leverage so my yield matches any maturity I wish. If, as I postulated above, yield change sensitivity scales with leverage at the same rate as yield, then it simply makes sense for me to pick the steepest point on the yield curve and lever to the duration I like until either the contract expires or the yield curve changes shape to give another point that is steep enough to justify the transaction cost to switch positions. This maximizes roll yield as the bonds mature. I could also look for maximum backwardation, but I don't think there is much in these contracts beyond the implied yield. If my premise is wrong and yield increases faster than duration under leverage, then yeah, it would kind if be dumb to ever invest in anything but short term treasuries for the purpose of yield unless you wanted to capitalize on yield volatility itself, but that should have no expected return.

I'll also gave to think about recovery time for shorter duration. It might be that even if leverage changes your effective duration for change in principal value from yield changes, it certainly doesn't change term. So... I don't know. I have to go math now.

Rob Bertram
Posts: 823
Joined: Mon May 05, 2014 12:15 pm

Re: Time Diversification - bottom line, what do you do?

Post by Rob Bertram » Tue Aug 23, 2016 7:27 pm

Swelfie wrote:
Rob Bertram wrote:The portfolio with short-term bonds is still less sensitive to interest rate changes.
This is the part I question. A $1000 notional bond of a specific duration geared at 1x will be less effected by a change in yield than a $2000 notional bond geared at 2x. Obviously the hit from a yield hike will be twice as hard and make the bond act as if it had a longer duration. What you are implying is that as you increase leverage you increase the yield of the bond faster than the effective duration, so that if the duration of the bond in yield change sensitivity made a 2 year bond act like a 5 year bond due to leverage, the actual yield from that 5 year bond would be more akin to a 7 year bond. I'm not sure I buy that, otherwise why would anyone buy 5 year bonds?
So maybe a different way to look at it would be like this: To match the 5-year yield, you need less than 2.5x leverage on a 2-year bond.

In my massive leverage thread, someone posted an article that showed shorter-term bonds have higher risk-adjusted returns. The author suggested that the smarter investor should leverage short-term bonds. I'll see if I can dig up that link.

Swelfie
Posts: 200
Joined: Mon Mar 14, 2016 12:54 am

Re: Time Diversification - bottom line, what do you do?

Post by Swelfie » Wed Aug 24, 2016 12:17 am

Rob Bertram wrote:To match the 5-year yield, you need less than 2.5x leverage on a 2-year bond.
I think that is because yield is a factor of duration, not term:
http://www.investopedia.com/university/ ... dbond5.asp

So what I'm suspecting (and I haven't done the math yet) is that leverage stretches duration. And I would suspect that the effect of the duration stretching both increases yield and increases sensitivity to yield changes at an equivalent rate, mainly because the sensitivity to changing yield is a direct factor of the bond's yield.

So what I suspect is that a 2 year STRIP leveraged to 2.5x might be equivalent to a 5 year STRIP, but that is only because term and duration are the same thing in a coupon-free bond. But once you introduce duration I suspect that it is not the term leverage is increasing, but the duration, and the duration of a 5yr coupon treasury is likely less than 2.5x a 2yr coupon treasury. So when you leverage the 2yr treasury by 2.5x you are getting something closer to a 15yr treasury in both return and sensitivity to yield changes (which would be important to know with your high degree of leverage). Again, I haven't actually done the math yet and I'm too tired to go at that hairy equation tonight.

But, I also think that bond duration and returns assumes a linear yield curve. However, the yield curve is never perfectly linear. Most of the time it's an S-curve, with a low slope at 4 weeks, a gradually increasing slope out to 2-5 years, and then a gradually decreasing slope out to 30 years. Right now it's lumpy, starting slightly steep, rapidly steepening to a max slope at 6 months, then decreasing slope out to 3 years, then increasing slope to 5 years, gradually decreasing out to 20 years, then slightly increasing out to 30. The slope at 6 months is by far the maximum right now though.

So, there are 3 returns from bonds (4 from bond futures). The first is has an expected return of zero. When yield goes up on new bonds of the same duration, the bonds you hold lose money and vice versa. This is obviously because no one wants your low yield bonds when there are higher yield ones available now. The second return is coupon yield. We all know that one as it is the most obvious. I'll skip to the fourth, because I want to dwell more on the third. The fourth is backwardation or contago, when the new contract you are about to roll into is lower priced or higher priced then it logically should be based on the projected returns. I haven't measured for bond contracts, but they are so liquid I doubt this is a big factor. I know it's tiny in ^ES. It's more of a factor in commodities. The 3rd is kind of weird. Since a 5 year bond in 2 years becomes a 3 year bond, and with a normal yield curve the yield on 3 year bonds is lower than that of 5 year bonds, the price of the bond will go up for no other reason that you waited two years because the bonds in the 3 year environment have lower yields than this new three year bond you hold that used to be a 5 year bond. So as long as the yield curve is not inverted, your bonds should increase in value just by waiting. The slope of the yield curve at your bond's term is the rate at which this third return happens. The steeper the slope at the remaining term of the bond, the faster the bond appreciates in value.

So what I'm getting at is that the if I can control the degree of volatility from 1 and the yield from 2 by changing leverage, it doesn't matter at all, given a linear yield curve, what term treasury I choose, I can turn any treasury into any other. But the yield curve is not linear, and leverage changes the duration in my hypothesis, but not the term. So, if I choose, from my risk profile, that my preferred bond holding is a 10 year bond, but the steepest part of the yield curve is a 2 year bond, I should lever the 2 year bond to have the duration of a 10 year bond, but it will retain the roll yield of a 2 year bond as it's remaining term decreases. So, in my opinion, rather than simply levering 2 year bonds always, sometimes I may want to reduce leverage and roll to a 5 year bond because this quarter the best roll yield is on 5 year bonds. I can change term every time I roll contracts, chasing the best roll yield, but keeping the same coupon yield and rate change risk. I think it may even be more advantageous because as rates go up, the point of the steepest slope has ALREADY increased, and since the yield curve has a kind of reversion to the mean, it is less likely I will get as much of a magnitude change for those bonds for the amount I have stretched their duration with leverage.

Now, using my above hypothesis (again, alas, not having done the math) I think it is perfectly reasonable that this explains why shorter term treasuries to have a better sharpe ratio than long term treasuries. The reason would be that the normal shape of the curve give steeper curves at 2-5 years than normally exist at 20-30 years and that the difference in sharpe ratio is completely explained by the average relative slope of the yield curve at these terms. I would also suspect that the 4 week and 3 month treasuries have worse sharpe ratios than 2-5 year treasuries for the same reason, since the yield curve tends to flatten at extremely short terms.

So, my hypothesis that I will get around to backtesting soon but I've been busy lately, is that the optimal sharpe ratio for bonds is to pick a duration, say it's the duration of the 10 year bond, then always buy the future at the highest slope of the yield curve (unfortunately you are kind of stuck with only a few points due to low liquidity in say 3 year treasuries) and lever that treasury so that the duration is equivalent to your reference bond. I think that will give you a higher sharpe ratio than any term bond you could choose to buy and hold, so it is that bond that I want to hold and lever. Of course, you might optimize further by trading more often than quarterly if the yield curve changes rapidly enough to make another term have a significantly higher slope that justifies the cost of additional trades, or trying to target backwardation in the future roll prices, but I suspect those optimizations are much smaller than just moving into the best future every quarter.

And this strategy means that you wouldn't have a target leverage, but rather a dynamic leverage with each roll and a target effective duration.

Rob Bertram
Posts: 823
Joined: Mon May 05, 2014 12:15 pm

Re: Time Diversification - bottom line, what do you do?

Post by Rob Bertram » Wed Aug 24, 2016 12:16 pm

I think we might be going down two different paths. My comments are around a balanced portfolio of stocks and bonds. Transitioning a portfolio from intermediate-term bonds to short-term bonds is an exercise in aligning the Sharpe ratio. I might have confused you, the goal was not to synthesize a higher-duration bond by leveraging lower-duration bonds.

Leverage definitely increases the interest rate sensitivity linearly. For example, if I leveraged 2-year duration bonds by 3x, I would expect a 6% drop in value for every 1% increase in yield. That makes sense, right? The notional portfolio drops by 2%, and that is magnified by my 3x leverage. On the other hand, the notional portfolio still consists of 2-year duration bonds, so I would expect the point of indifference to still be 2 years (assuming no future change in yield), not 6 years.

Swelfie
Posts: 200
Joined: Mon Mar 14, 2016 12:54 am

Re: Time Diversification - bottom line, what do you do?

Post by Swelfie » Wed Aug 24, 2016 1:54 pm

Rob Bertram wrote:I think we might be going down two different paths. My comments are around a balanced portfolio of stocks and bonds. Transitioning a portfolio from intermediate-term bonds to short-term bonds is an exercise in aligning the Sharpe ratio. I might have confused you, the goal was not to synthesize a higher-duration bond by leveraging lower-duration bonds.

Leverage definitely increases the interest rate sensitivity linearly. For example, if I leveraged 2-year duration bonds by 3x, I would expect a 6% drop in value for every 1% increase in yield. That makes sense, right? The notional portfolio drops by 2%, and that is magnified by my 3x leverage. On the other hand, the notional portfolio still consists of 2-year duration bonds, so I would expect the point of indifference to still be 2 years (assuming no future change in yield), not 6 years.
Yeah, I know your intent wasn't to synthesize a longer duration bond by leverage, I'm just saying that's what you did ;). And I did some of math. The bond you synthesize when you leverage a 2 year term to maturity bond has the duration determined by the function (durationOfBondsCarried * ValueOfBondsCarried - durationOfNegativeBondPayingForLeverage * valueOfLeverage)/ValueOfUnleveragedBonds. Since the duration of a 2 year Treasury is about 1.96, if you lever it out 2.5x you get a 6-7 year bond. Close to 5 but it overshoots because you aren't scaling term to maturity linearly, you are scaling duration. Your 25x is basically turning those 2 year bonds into 40 year bonds.

Now my thought was that maybe it didn't make sense to floor leverage, because with a good yield roll, half your returns can be roll with the other half being coupon (and anything else just market timing luck from yield volatility). So maybe, when levering bonds, since your coupon you really have no control over, not do you have control over yield to maturity, you can have a better sharpe ratio be moving your term to maturity to an optimal yield roll, but keeping it a 40 year bond in duration. So maybe when you roll your contract, if your best yield roll is out at 10 year Treasuries this quarter, you might want to roll your contract there but drop your leverage from 25x to about 3.9x, which would give you the same synthetic position you are in now from a risk perspective but your expected returns would be better, giving you a better sharpe ratio. I mean, you are rolling anyway, just ask the spreadsheet where to roll to and what leverage level to retain your same position at a better expected return. But when I started playing with the maths i think I am seeing some anomalies and I think there might be something to this inherently higher sharpe ratios on short term that isn't explained by the roll yield (although that seems to be a competing factor). I'm mathing more and looking for that paper you mentioned now. I definitely need some complicated back testing on this to be sure.

Rob Bertram
Posts: 823
Joined: Mon May 05, 2014 12:15 pm

Re: Time Diversification - bottom line, what do you do?

Post by Rob Bertram » Wed Aug 24, 2016 2:37 pm

I believe that this is the paper I remember from this post.

edit: As we are drifting from the OP's topic, I've moved my comments to my thread on leverage.

Post Reply