The Glidepath Illusion (it doesn't work)

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

The Glidepath Illusion (it doesn't work)

Post by Browser »

Interesting article by Rob Arnott concludes you're probably better off maintaining a static allocation to stocks and bonds instead of investing in a Target Date fund.
Young adults should buy stocks; mature adults should favor bonds. Or so we’re taught. It makes intuitive sense. Young people have modest savings and lots of time to recover losses from any bear markets. People approaching retirement have more to lose and less time to recover from bear markets. Typically, they want greater certainty as to how much they can safely spend in retirement and less risk that a decline in the value of their investments will demolish their retirement plans.

This type of logic has spawned a huge retirement planning industry, with a wide array of target-date strategies whose Glidepath mechanisms systematically ramp down portfolio risk as an investor approaches retirement. These products are, for many people, the default option in their 401(k) and other defined contribution pension portfolios. Shockingly, the basic premise upon which these billions are invested is flawed.
Read more.
We don't know where we are, or where we're going -- but we're making good time.
User avatar
Mel Lindauer
Moderator
Posts: 35782
Joined: Mon Feb 19, 2007 7:49 pm
Location: Daytona Beach Shores, Florida
Contact:

Re: The Glidepath Illusion (it doesn't work)

Post by Mel Lindauer »

Browser wrote:Interesting article by Rob Arnott concludes you're probably better off maintaining a static allocation to stocks and bonds instead of investing in a Target Date fund.
Young adults should buy stocks; mature adults should favor bonds. Or so we’re taught. It makes intuitive sense. Young people have modest savings and lots of time to recover losses from any bear markets. People approaching retirement have more to lose and less time to recover from bear markets. Typically, they want greater certainty as to how much they can safely spend in retirement and less risk that a decline in the value of their investments will demolish their retirement plans.

This type of logic has spawned a huge retirement planning industry, with a wide array of target-date strategies whose Glidepath mechanisms systematically ramp down portfolio risk as an investor approaches retirement. These products are, for many people, the default option in their 401(k) and other defined contribution pension portfolios. Shockingly, the basic premise upon which these billions are invested is flawed.
Read more.
Surprise, surprise! This from someone who stands to lose business if folks invest in the Target Retirement funds.
Best Regards - Mel | | Semper Fi
BYUvol
Posts: 120
Joined: Sat Mar 24, 2012 3:06 pm
Location: KY

Re: The Glidepath Illusion (it doesn't work)

Post by BYUvol »

Mel Lindauer wrote:
Browser wrote:Interesting article by Rob Arnott concludes you're probably better off maintaining a static allocation to stocks and bonds instead of investing in a Target Date fund.
Young adults should buy stocks; mature adults should favor bonds. Or so we’re taught. It makes intuitive sense. Young people have modest savings and lots of time to recover losses from any bear markets. People approaching retirement have more to lose and less time to recover from bear markets. Typically, they want greater certainty as to how much they can safely spend in retirement and less risk that a decline in the value of their investments will demolish their retirement plans.

This type of logic has spawned a huge retirement planning industry, with a wide array of target-date strategies whose Glidepath mechanisms systematically ramp down portfolio risk as an investor approaches retirement. These products are, for many people, the default option in their 401(k) and other defined contribution pension portfolios. Shockingly, the basic premise upon which these billions are invested is flawed.
Read more.
Surprise, surprise! This from someone who stands to lose business if folks invest in the Target Retirement funds.
He wouldn't be losing any more or less business if someone chooses a LifeStrategy rather than a Target Retirement fund at Vanguard.

This is actually pretty interesting to me, because our fiduciary at Merrill Lynch for my 401k adamantly refuses to allow target date funds in our plan at my request, because he says they have "consistently lagged benchmarks." I thought he was crazy, but maybe this is what he was referring to, I'll definitely need to do more research. My Roth IRA is 100% VFORX, and I've often wondered what would happen if the increase in bond % coincides with rising rates. I've got long enough that I'm comfortable with 90% equities, but I don't think I always will be. If you truly get better expected return with a 50/50 constant allocation, seems like that's a no-brainer.
User avatar
nisiprius
Advisory Board
Posts: 52219
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: The Glidepath Illusion (it doesn't work)

Post by nisiprius »

Same-old, same-old. It's just a risk/reward proposition, with the presenter telling us that we should not trust our own risk tolerance, but should accept more risk in order to get more reward.

In table 1, every measure of dispersion shows that the balanced allocation provides both a greater mean outcome and a greater dispersion of outcome.

In table 2, in which--prudently or foolishly--Arnott substitutes a personal pessimistic prediction of the future for an extrapolation of the past, it is still the same result. The static allocation has a greater dispersion of results than the glide path.

The word "risk tolerance" appears nowhere in the paper.

The rationale for a glide path is simple: if we have lower risk tolerance at age 65 than we do at age 30, then our stock allocation should be lower at age 65 than it was at 30.

The data we need to know to judge the suitability of a glide path allocation has nothing to do with projections of "3-D hurricanes." What the data we need is simple: how does the average investor's risk tolerance change with age? If average investor's risk tolerance declines with age, then a glide slope allocation is more suitable for the average investor than a steady allocation would be.

Notice that it doesn't matter what the reason for the change in risk tolerance is. Maybe it's Homo economicus intuitively doing calculus on declining human capital, maybe it's declining testosterone levels, it doesn't matter--if our risk tolerance declines, so should our stock allocation.

I am still figuring out how I can courteously approach a Balanced Barbara of my acquaintance and get her story; what I know that she uses or used a fee-only advisor who really liked Wellington, and I only just learned from her husband that "well, you know, she always invested more aggressively than I did--and during 2008-2009 she sold and locked in her loss."
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
Epsilon Delta
Posts: 8090
Joined: Thu Apr 28, 2011 7:00 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Epsilon Delta »

Do some research into why somebody would calculate the standard deviation and then use the range as a measure of dispersion. Very odd.
ourbrooks
Posts: 1575
Joined: Fri Nov 13, 2009 3:56 pm

Re: The Glidepath Illusion (it doesn't work)

Post by ourbrooks »

I think we need to distinguish between an individual's risk tolerance and their risk preference. My risk preference is for zero risk, but I'm even more afraid of running out of money before I die and glidepath strategies may increase the odds of my running out of money!!

Here's why I think that's so: All of the "4%" studies - Bengen, Trinity 1, Trinity 2, Wade Pfau, etc. used fixed allocations. Wade Pfau has promised to look at glidepaths, but he hasn't done so yet. Is there reason to believe that the 4% rate wouldn't work with a glidepath strategy? Yes, there is. All of these studies show that when the percentage of bonds gets high enough, you can no longer safely withdraw at 4%. This seems to happen at about 80% or 90% bonds.

It may be the case that if you reach that percentage late enough in life, it doesn't have much effect on the safe withdrawal rate, but I wouldn't want to be the one to find out.
So here's my horrible choice: I can do a glidepath strategy and reduce my withdrawal rate and end up old and poor or I can do a glidepath strategy, keep my withdrawal rate the same, and take the chance of ending up completely broke or I can do a fixed allocation, toss my monthly statements in the paper shredder when the market is down, and end up neither poor nor broke.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: The Glidepath Illusion (it doesn't work)

Post by Rodc »

http://home.comcast.net/~rodec/finance/ ... nBonds.pdf

Figure 3 is most germane. This is something I did back about 4 years ago.

Conclusion is age in bonds vs fixed allocation with the same bond exposure give same results within measurement error.

Age in bond though arguably leaves you more relaxed as you are older, and less likely to have less regret should markets tank near retirement.

Sort of person A invests for decades and have a more or less smooth trajectory to portfolio $X is likely happier than person B whose nest egg shot up to $1.5X then had it drop to $X. Same end point, not same happiness.
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.
ourbrooks
Posts: 1575
Joined: Fri Nov 13, 2009 3:56 pm

Re: The Glidepath Illusion (it doesn't work)

Post by ourbrooks »

Rodc wrote:http://home.comcast.net/~rodec/finance/ ... nBonds.pdf

Figure 3 is most germane. This is something I did back about 4 years ago.

Conclusion is age in bonds vs fixed allocation with the same bond exposure give same results within measurement error.

Age in bond though arguably leaves you more relaxed as you are older, and less likely to have less regret should markets tank near retirement.

Sort of person A invests for decades and have a more or less smooth trajectory to portfolio $X is likely happier than person B whose nest egg shot up to $1.5X then had it drop to $X. Same end point, not same happiness.
From the above link:
Given the limitations in the data I think the most we can say is that historically there is no evidence to support Age in Bonds vs simply holding the simpler fixed 60/40 portfolio. A few years ago when I was writing Monte Carlo simulators and simple consumption smoothing models (for using in accumulation through retirement) for my own use, I could find nothing to support the superiority of any simple shifting of bonds as we age paradigm, or based on retirement dates (for example, 60/40 to retirement and 40/60 afterwards). Perhaps there is value in monitoring your investment progress towards “enough” and shifting to a more conservative capital preservation allocation at that point, but I have yet to see such a study and did not try to do that here. In fact if that is what you want to do, one might do better to but some money into an annuity rather than shifting to more bonds
Currently, there's another thread active about a study by Wade Pfau in which he shows that in times of low expected returns (now), annuities + stocks beat stocks + bonds. What the study did not evaluate was an annuity ladder, annuitizing increasing amounts with increasing age. My suspicion is that it beats stocks + bonds even when expected returns are high.
User avatar
nisiprius
Advisory Board
Posts: 52219
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: The Glidepath Illusion (it doesn't work)

Post by nisiprius »

ourbrooks wrote:I think we need to distinguish between an individual's risk tolerance and their risk preference. My risk preference is for zero risk, but I'm even more afraid of running out of money before I die and glidepath strategies may increase the odds of my running out of money!!
It's risk/return again. And it's not the glide path that's making the difference, it's just the higher stock allocation.

If you pick an arbitrary number, and decide you will spend that number no matter what, then if you pick a high number and consider only the chances of exhausting the portfolio, the results you will get will always favor a high-risk, high-return portfolio. But you baked a preordained conclusion into your analysis when you picked that high number. If you pick a number that is higher than a low-risk portfolio can deliver, then you will probably run out of money if you use a low-risk portfolio, and you have a better chance of lucking out if you use a high-risk portfolio. If you in fact have a high risk tolerance that leads you prefer taking a shot with a high withdrawal rate, then the high-risk portfolio suits you.

Once you have decided that you are going to jump Snake Canyon, you are safer on a rocket-powered Evel Knievel Skycycle, which might make it, then on a Ducati, which can't possibly. The Ducati has a higher change of crashing into the bottom of Snake Canyon. That doesn't mean a Skycycle is safer than a Ducati. It only means a Skycycle is safer than a Ducati if you are a daredevil to begin with. You baked in an implied high risk tolerance when you decided to jump Snake Canyon.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
ourbrooks
Posts: 1575
Joined: Fri Nov 13, 2009 3:56 pm

Re: The Glidepath Illusion (it doesn't work)

Post by ourbrooks »

nisiprius wrote: Insist on a 6% withdrawal rate and you'll conclude that you should be 100% stocks. And you'll conclude that any glide path is deleterious because it cuts down on your stock allocation.
To paraphrase slightly, "Insist on a 4% withdrawal rate and you'll conclude that you should follow a fixed allocation with at least 30% stocks."
Even a 4% safe withdrawal rate bakes in a fairly substantial degree of risk tolerance. If you really, really can't stand volatility, reduce your withdrawal rate; don't fool yourself that a glidepath means you can spend as much as you did and still avoid risk.

What I'm really concerned about is running out of money before I die or of having to reduce my spending to uncomfortable levels. That's the only risk I'm worried about; what happens to my stocks in the meantime is just noise. There's no evidence whatsoever from any of the safe withdrawal rate studies that holding betweemn 40% and 60% stocks increases my risk by any measurable amount. There is evidence in all of the studies that bond allocations above 80% DO increase my risk of running out of money. For the risk I'm most worried about, a portfolio with an intermediate stock allocation is safer - less chance of a damaging loss - than a portfolio with too high a bond percentage.
pkcrafter
Posts: 15461
Joined: Sun Mar 04, 2007 11:19 am
Location: CA
Contact:

Re: The Glidepath Illusion (it doesn't work)

Post by pkcrafter »

I think it's a very interesting article, but very misdirected in the supposed analysis of target date funds. It's not really about target date funds; it's about asset allocation. Anyone who's familiar with John Norstad's article on dispersion of returns should recognize the pattern.

For those who concluded Arnott is saying a static AA is better than an adjustable one is correct to a point. A static AA of 50/50 will reduce dispersion vs one that has 90% stock, but Arnott actually concludes that an inverse glidepath will provide the best returns.
Connie rationalizes that if a static 50/50 strategy is better than a Glidepath strategy, an Inverse-Glidepath strategy might be more appropriate for meeting her goals than either of the “standard” options. It should come as no surprise that this counterintuitive strategy beats a static 50/50 portfolio by essentially the same margin that static 50/50 beats Glidepath.


This conclusion is very interesting because as far as I know such a strategy has not been examined before. I guess the theory is that later in life Connie has built up a enough assets to where she actually has some ability to take more risk.

The other interesting point is also clear--you are better off with lower equity allocations, even when just starting out--unless, of course, you get lucky.

Paul
When times are good, investors tend to forget about risk and focus on opportunity. When times are bad, investors tend to forget about opportunity and focus on risk.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Browser »

nisiprius wrote:Same-old, same-old. It's just a risk/reward proposition, with the presenter telling us that we should not trust our own risk tolerance, but should accept more risk in order to get more reward.

In table 1, every measure of dispersion shows that the balanced allocation provides both a greater mean outcome and a greater dispersion of outcome.

In table 2, in which--prudently or foolishly--Arnott substitutes a personal pessimistic prediction of the future for an extrapolation of the past, it is still the same result. The static allocation has a greater dispersion of results than the glide path.

The word "risk tolerance" appears nowhere in the paper.

The rationale for a glide path is simple: if we have lower risk tolerance at age 65 than we do at age 30, then our stock allocation should be lower at age 65 than it was at 30.

The data we need to know to judge the suitability of a glide path allocation has nothing to do with projections of "3-D hurricanes." What the data we need is simple: how does the average investor's risk tolerance change with age? If average investor's risk tolerance declines with age, then a glide slope allocation is more suitable for the average investor than a steady allocation would be.

Notice that it doesn't matter what the reason for the change in risk tolerance is. Maybe it's Homo economicus intuitively doing calculus on declining human capital, maybe it's declining testosterone levels, it doesn't matter--if our risk tolerance declines, so should our stock allocation.

I am still figuring out how I can courteously approach a Balanced Barbara of my acquaintance and get her story; what I know that she uses or used a fee-only advisor who really liked Wellington, and I only just learned from her husband that "well, you know, she always invested more aggressively than I did--and during 2008-2009 she sold and locked in her loss."
Actually, your definition of "risk" as absolute volatility seems a little off the mark to me. It seems to me that it is risk-adjusted returns we're after here. Both the 50/50 and 80->20 portfolios have the same average 50/50 stock-bond allocation over time; they differ in the temporal distribution of that allocation. However, the fixed 50/50 portfolio produces a 10.8% higher mean return over time for a 9.5% increase in average volatility (SD), so it has a higher risk-adjusted return; i.e, it is more efficient.

The glidepath investor is making the false assumption that risk is being systematically lowered over time, but that's only true if risk is equated with volatility. Furthermore, it requires accepting much higher volatility on the front end. If risk is viewed in terms of volatility-adjusted returns, the glidepath investor is actually skewing her portfolio risk profile negatively; that is, although she is generating lower volatility over time, it is at the cost of disproportionately reducing portfolio returns. In contrast, the fixed-allocation investor is maintaining the same risk-adjusted return profile over time, which is has a higher likelihood of generating higher terminal risk-adjusted returns over a given time period.

Now if one is concerned that the volatility of a given fixed allocation is unacceptably high, then one can simply reduce the allocation to stocks and maintain that fixed allocation instead. For example, if a fixed 40% stocks / 60% bonds allocation can produce the same risk-adjusted returns as an 80->20 glidepath allocation, that seems preferable to me. Otherwise, our hapless glidepath investor would have to tolerate the much higher level of volatility over the first two-thirds of her portfolio lifetime during which her stock allocation is greater than 40%. Latter day peace-of-mind from the glidepath comes at the cost of a much rougher ride upfront. But since this pathway demonstrably does NOT produce superior risk-adjusted terminal returns to a fixed allocation strategy, the benefit of glidepath is entirely illusory. Frankly, I'd prefer to experience a relatively constant, moderate level of portfolio volatility over time if I know it gets me to the same place as the glidepath. But that's just me.
We don't know where we are, or where we're going -- but we're making good time.
stlutz
Posts: 5585
Joined: Fri Jan 02, 2009 12:08 am

Re: The Glidepath Illusion (it doesn't work)

Post by stlutz »

Browser--Thanks for starting the thread; interesting discussion!

I agree with your conclusions that a fixed "moderate" allocation, say 60/40 makes a lot of sense, although I think my reasoning might be a little different. Taking a bunch of risk when you have no money doesn't move the ball that much--your savings rate makes a much bigger difference when you're age 22. That $2000 you made or lost in your IRA or 401K at that point in time seems like a big deal, but looking back later you realize it's not. So, when you're young there's not a big reason to take a lot of risk.

As you get older and hopefully wealthier, you can benefit from and be hurt by risk. You need those benefits, but you can't afford to lose it all either. So, again a moderate approach makes intuitive sense.

It's really only when you reach that point at which you can say, "I have enough", that moving to a 75% or higher fixed income allocation makes sense.

The problem I do have with the original article is assuming a rather large equity return premium. That was true in the 20th century, but may not be true in this century--we'll see. That's why stocks are risky. If you assume a smaller premium, the 20->80 portfolio option won't look as good.
cjking
Posts: 2039
Joined: Mon Jun 30, 2008 4:30 am

Re: The Glidepath Illusion (it doesn't work)

Post by cjking »

Have not yet read the article, but had this thought...

Isn't a glidepath strategy that slowly but relentlessly move you out of equities regardless of market conditions in one respect similar to a withdrawal strategy that takes a smooth income from a volatile portfolio, in that it imposes a degree of sequence-of-returns risk that could be avoided?

The impact will be moderated by the fact that you are taking a fixed percentage of the portfolio rather than a fixed amount, so you are to some extent varying you withdrawals. However I suspect the degree of variation is to low and the rate of withdrawal from equities to high to confer complete immunity from sequence-of-returns.
cjking
Posts: 2039
Joined: Mon Jun 30, 2008 4:30 am

Re: The Glidepath Illusion (it doesn't work)

Post by cjking »

I've now read the article. I agree with the criticism that essentially what it "proves" is that a higher average equity allocation will cause higher income.

I do believe glidepaths are inferior to alternative strategies, but this article isn't one I'd cite.
YDNAL
Posts: 13774
Joined: Tue Apr 10, 2007 4:04 pm
Location: Biscayne Bay

Re: The Glidepath Illusion (it doesn't work)

Post by YDNAL »

Browser wrote:Interesting article by Rob Arnott concludes you're probably better off maintaining a static allocation to stocks and bonds instead of investing in a Target Date fund.
Young adults should buy stocks; mature adults should favor bonds. Or so we’re taught. It makes intuitive sense. Young people have modest savings and lots of time to recover losses from any bear markets. People approaching retirement have more to lose and less time to recover from bear markets. Typically, they want greater certainty as to how much they can safely spend in retirement and less risk that a decline in the value of their investments will demolish their retirement plans.

This type of logic has spawned a huge retirement planning industry, with a wide array of target-date strategies whose Glidepath mechanisms systematically ramp down portfolio risk as an investor approaches retirement. These products are, for many people, the default option in their 401(k) and other defined contribution pension portfolios. Shockingly, the basic premise upon which these billions are invested is flawed.
Read more.
Prudend Polly, Balanced Burt, Contrary Connie show historical outperformance in Equities which are riskier and provided higher returns. BTW, it cracks me up when someone looks at 1871 data.... did they find the original ledgers carved in stone? The MORE Equities as you accumulated MORE money, the better the outcome based on such historical data. A classic sales pitch from someone selling Equities. All of it is nothing but hogwash.

A glidepath to less risky Assets as you age is a sound strategy IF it conforms to the individual's ability and need for risk. Period, the end.
Landy | Be yourself, everyone else is already taken -- Oscar Wilde
cjking
Posts: 2039
Joined: Mon Jun 30, 2008 4:30 am

Re: The Glidepath Illusion (it doesn't work)

Post by cjking »

I decide to do a similar exercise to the paper, so that I could assess risk-related performance of each strategy, and compare my own suggested strategy.

I used Shiller data for my history, and assumed completely smooth 0% real return on the bonds/cash part. My investor invested $1000/12 per month for 40 years to produce a final balance.

The 4 strategies compared were:
50:50
80->20 glidepath
20->80 glidepath
100 to 0 opportunistic switching

Opportunistic switching is my strategy of switching overnight from a high to a low equity allocation once one has enough. Enough was defined to be $91,712, which was the historical average produced by 50:50 over all rolling periods.

Five statistcs were collected for each strategy
1-3. Final balance: worst case < average < best case
4. Average monthly equity exposure in dollars (a measure of risk taken)
5. Average across periods of ratios of final balance to average monthly equity exposure for each period, a measure of risk-adjusted returns.

Results (note balance figures are give as min < average < max)
50:50: balance 52,653 < 91,712 < 136,601, exposure 18,144, balance/exposure 5.1
80->20: balance 60,280 < 80,601 <104,954, exposure 14,186, balance/exposure 5.7
20->80: balance 45,692 < 105,019 < 180,842, exposure 22,918, balance/exposure 4.6
100 switch 0: balance 91,714 < 93,170 < 99,024, exposure 15,596, balance/exposure 7.2

So 20->80 glidepath recommended in the article actually has worst risk-adjusted performance, 80->20 beats 50:50, contrary to my expectation, but my opportunistic switching strategy demolishes the rest.

One caveat with regard to opportunistic switching: it is sensitive to target, setting the target too high could result in a much worse worst-case scenario. I doubled the target, and the worst final balance fell from 91,801 to 68,831, and this also lowered the average balance/exposure ratio to 4.7.

(Edited to correct some figures after finding an immaterial glitch in spreadsheet.)
Last edited by cjking on Wed Sep 26, 2012 9:00 am, edited 1 time in total.
User avatar
Joe S.
Posts: 479
Joined: Sat May 05, 2007 12:11 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Joe S. »

arnott wrote:Consider an investor, Prudent Polly, who plans to save for retirement by investing in a standard Glidepath portfolio.
It's interesting to look at all the subtle ways that Arnott stacks the deck to make the glidepath look bad. For instance he lies and says he is presenting a "standard Glidepath portfolio." However the glidepath he chooses is one that has 80% stocks at age 22 and 20% stocks at age 63. Very few glidepaths have this steep an angle. In fact, I'd challenge anyone to find a published glidepath that recommends ~80% stocks at age 22 and ~20% stocks at age 63. He creates a strawman strategy then beats it up.

(Of course the point that the glidepath strategy has a net lower exposure to stocks is another effect, but others have discussed this.)
barnaclebob
Posts: 5588
Joined: Thu Aug 09, 2012 10:54 am

Re: The Glidepath Illusion (it doesn't work)

Post by barnaclebob »

This article is worthless. Of course a portfolio heavier on stocks will have done better on average, that's not the point of a TR fund. You increase bond percentage to reduce risk at the expense of lower returns in most cases.

I have seen another article that talked about how most NFL teams would be better off to always go for a 2 point conversion, same principle.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: The Glidepath Illusion (it doesn't work)

Post by Rodc »

cjking wrote:I decide to do a similar exercise to the paper, so that I could assess risk-related performance of each strategy, and compare my own suggested strategy.

I used Shiller data for my history, and assumed completely smooth 0% real return on the bonds/cash part. My investor invested $1000/12 per month for 40 years to produce a final balance.

The 4 strategies compared were:
50:50
80->20 glidepath
20->80 glidepath
100 to 0 opportunistic switching

Opportunistic switching is my strategy of switching overnight from a high to a low equity allocation once one has enough. Enough was defined to be $91,712, which was the historical average produced by 50:50 over all rolling periods.

Five statistcs were collected for each strategy
1-3. Final balance: worst case < average < best case
4. Average monthly equity exposure in dollars (a measure of risk taken)
5. Average across periods of ratios of final balance to average monthly equity exposure for each period, a measure of risk-adjusted returns.

Results (note balance figures are give as min < average < max)
50:50: balance 52,653 < 91,712 < 136,601, exposure 18,144, balance/exposure 5.1
80->20: balance 60,280 < 80,601 <104,954, exposure 14,186, balance/exposure 5.7
20->80: balance 45,692 < 105,019 < 180,842, exposure 22,918, balance/exposure 4.6
100 switch 0: balance 91,801 < 95,223 < 102,650, exposure 15,967, balance/exposure 7.1

So 20->80 glidepath recommended in the article actually has worst risk-adjusted performance, 80->20 beats 50:50, contrary to my expectation, but my opportunistic switching strategy demolishes the rest.

One caveat with regard to opportunistic switching: it is sensitive to target, setting the target too high could result in a much worse worst-case scenario. I doubled the target, and the worst final balance fell from 91,801 to 68,831, and this also lowered the average balance/exposure ratio to 4.7.
Interesting. I have wondered about such a strategy. Seems very prudent and real world with the caveat that as you note, one has to pick a modest target and not get greedy.
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.
ks289
Posts: 655
Joined: Sun Mar 11, 2012 12:42 pm

Re: The Glidepath Illusion (it doesn't work)

Post by ks289 »

barnaclebob wrote:This article is worthless. Of course a portfolio heavier on stocks will have done better on average, that's not the point of a TR fund. You increase bond percentage to reduce risk at the expense of lower returns in most cases.

I have seen another article that talked about how most NFL teams would be better off to always go for a 2 point conversion, same principle.
I like these types of exercises to question the risk/benefit of different approaches to retirement planning (and really any important decision). Granted, projecting the future using historical data has its limitations.

If using a TR approach reduces volatility and avoids low probability/disastrous outcomes but is likely to SIGNIFICANTLY underperform a fixed allocation or other approach, I am willing to at least think about and perhaps adjust (if this can be done) my risk tolerance/preference. This may not work for me (or most), but I like to at least ponder it.

--
Now if your favorite NFL team designed a special play or acquired a monster player (how about Walter Payton going over the pile) who had a 50+% success rate on 2-point conversions (as opposed to the current 40-something%), you'd have to think about adjusting strategy in favor of this "riskier" approach!
cjking
Posts: 2039
Joined: Mon Jun 30, 2008 4:30 am

Re: The Glidepath Illusion (it doesn't work)

Post by cjking »

Rodc wrote: Seems very prudent and real world with the caveat that as you note, one has to pick a modest target and not get greedy.
Just to flesh out the sensitivy of the strategy to over-ambition, a target of 120,000 (compared to original $91,712) actually gets a higher worst case scenario of 93,896, but one just a little higher at 128,000 gets you the 68,831. The latter figure equates to the worst balance a fixed 100:0 strategy would have delivered.

Edited to add another interesting statistic: it seems 106,000 is the highest target where every single rolling period hits the target.
Last edited by cjking on Wed Sep 26, 2012 9:46 am, edited 1 time in total.
trico
Posts: 684
Joined: Wed Nov 12, 2008 6:04 pm

Re: The Glidepath Illusion (it doesn't work)

Post by trico »

E=MC2 divided by the opposite of target date retirement 2010 x pie squared = $210000.00 at retirement. This is why I buy EIA annuities.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Browser »

Jim Otar conducted a study in which he compared the results of a fixed 65/35 stock-bond allocation to two glidepath strategies. Strategy #1 is an age-in-bonds strategy that starts with 70% in stocks at age 30 and systematically reduces that to 35% by age 65. Strategy #2 starts with 85% in stocks at age 30 and reduces that to 25% by age 65. Looking at the median values for all 35-year periods from 1900, there was only a small difference in median terminal portfolio values between these three strategies.

During bull market periods, the fixed AA resulted in a 23% greater mdian terminal portfolio value than the 85->25 glidepath portfolio. There was little difference in median terminal portfolio value between the two glidepath strategies. The SD of the fixed AA portfolio was slightly higher, but Otar calls this "good volatility" because it is associated with higher portfolio returns.

What about unlucky periods (the lowest decile of 35-year returns since 1900)? Here the fixed AA produced a 7% higher median terminal portfolio value with a lower SD than the 85->25 glidepath portfolio. There was little difference between the two glidepath strategies.

Based on these data, Otar does not recommend the glidepath portfolio strategy to his clients. Rather, he prefers a fixed AA based on the client's risk tolerance and regular review and appropriate adjustments based on changes in the client's situation.

http://retirementoptimizer.com/
We don't know where we are, or where we're going -- but we're making good time.
Tigermoose
Posts: 823
Joined: Sat Jul 21, 2012 5:02 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Tigermoose »

cjking,

With your opportunistic switching strategy, what do you do to mitigate the risk of high inflation eating away what you thought was enough in fixed income during retirement?
Institutions matter
yobria
Posts: 5978
Joined: Mon Feb 19, 2007 10:58 pm
Location: SF CA USA

Re: The Glidepath Illusion (it doesn't work)

Post by yobria »

Mr. Arnott is confused. Holding return constant, an optimal glide path distributes risk (major stock crash) evenly over your investing life. This is the expected case going forward, regardless of any historical data.

Plus it's just common sense - if you "front load the risk" (assume more earlier, less later vs static), you have more time to recover if the risk shows up.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Browser »

Also consider the discussion of glidepath by Wade Pfau:
But when you allow the stock allocation to change over time, what I find in these simulations is that retirees are better off by starting with a lower stock allocation but then actually increasing it over time, rather than decreasing it. This provides protection when wealth is the highest, and then allow for more aggressiveness later on in retirement. Bad retirements generally happen because of poor results just after retirement, and that is what the reverse glidepath helps to protect against.
Also an article by Blanchett in the December 2007 Journal of Financial Planning in which he concludes that a fixed AA is superior to a range of glidepath strategies.

Where exactly is the data to support the superiority of the glidepath strategy in producing higher returns or better risk-adjusted returns than simple fixed AA? I'm all ears.
Image
We don't know where we are, or where we're going -- but we're making good time.
grayfox
Posts: 5569
Joined: Sat Sep 15, 2007 4:30 am

Re: The Glidepath Illusion (it doesn't work)

Post by grayfox »

nisiprius wrote: Once you have decided that you are going to jump Snake Canyon, you are safer on a rocket-powered Evel Knievel Skycycle, which might make it, then on a Ducati, which can't possibly. The Ducati has a higher change of crashing into the bottom of Snake Canyon. That doesn't mean a Skycycle is safer than a Ducati. It only means a Skycycle is safer than a Ducati if you are a daredevil to begin with. You baked in an implied high risk tolerance when you decided to jump Snake Canyon.
Is it a choice whether one needs to jump Snake Canyon or not?

Suppose someone age 55 is working, earns 100K per year and is living on 75K per year. Typical middle class income and lifestyle.
They are saving 25K per year for retirement. 25% is a good savings rate.
So far they have accumulated 1 million dollars. That's a good sum, probably more than most have.
Suddenly they get laid off. They can't find another job. This happens to plenty of people.

So they have 1 million dollars and spend 75K per year. 7.5% withdrawal rate required
With some major belt tightening they get their spending down to 50K per year. You can only cut so much, unless you move to Slab City
Their starting withdrawal rate has to be 5%.
You say they are trying to jump Snake Canyon and need a rocket-propelled portfolio, i.e. a lot of stocks. Treasury bonds have negative real yields out to 20-years

They didn't choose 5% withdrawal rate. Circumstances chose it.

With negative real yields on safe Treasuries, say -1.0% the max withdrawal rate is probably about 1.5%. Retirees are being forced into risky assets.
The choice is take risk or live on $15,000 per year, which is $1,250 per month. The choice between Iraq and a hard place.
rkhusky
Posts: 17768
Joined: Thu Aug 18, 2011 8:09 pm

Re: The Glidepath Illusion (it doesn't work)

Post by rkhusky »

Browser wrote:Also consider the discussion of glidepath by Wade Pfau:
But when you allow the stock allocation to change over time, what I find in these simulations is that retirees are better off by starting with a lower stock allocation but then actually increasing it over time, rather than decreasing it. This provides protection when wealth is the highest, and then allow for more aggressiveness later on in retirement. Bad retirements generally happen because of poor results just after retirement, and that is what the reverse glidepath helps to protect against.
Also an article by Blanchett in the December 2007 Journal of Financial Planning in which he concludes that a fixed AA is superior to a range of glidepath strategies.

Where exactly is the data to support the superiority of the glidepath strategy in producing higher returns or better risk-adjusted returns than simple fixed AA? I'm all ears.
Image
The Pfau quote is talking about retirement, where it seems perfectly reasonable to have a fixed AA or even a modest inverse glide path.
As I approach retirement and leaving my relatively secure job, I am more interested in minimizing the worst-case scenarios, than in maximizing risk-adjusted returns.
yobria
Posts: 5978
Joined: Mon Feb 19, 2007 10:58 pm
Location: SF CA USA

Re: The Glidepath Illusion (it doesn't work)

Post by yobria »

Browser wrote:Where exactly is the data to support the superiority of the glidepath strategy in producing higher returns or better risk-adjusted returns than simple fixed AA? I'm all ears.
Image
The data isn't anywhere - we're looking forward, so it doesn't exist yet. Anyway I can prove anything I want by tweaking parameters in a historical simulation. "The data" shows that increasing your stock allocation to 100% at retirement is usually a winning move. That wouldn't lead me to mindlessly switch to that allocation.

The key here is deciding what we care about - what problem we're trying to solve. What I care about is a major decline in stock prices when I'm near or in retirement, when I don't have the human capital to recover. And a glide path mitigates this problem - holding expected final value constant, I take more risk when young, less when old relative to a lazy portfolio. That's a mathematical certainty on an expected basis, which is the only basis we have going forward.

Like the article, I'm talking mainly about the accumulation phase here.
rkhusky
Posts: 17768
Joined: Thu Aug 18, 2011 8:09 pm

Re: The Glidepath Illusion (it doesn't work)

Post by rkhusky »

grayfox wrote:
nisiprius wrote: Once you have decided that you are going to jump Snake Canyon, you are safer on a rocket-powered Evel Knievel Skycycle, which might make it, then on a Ducati, which can't possibly. The Ducati has a higher change of crashing into the bottom of Snake Canyon. That doesn't mean a Skycycle is safer than a Ducati. It only means a Skycycle is safer than a Ducati if you are a daredevil to begin with. You baked in an implied high risk tolerance when you decided to jump Snake Canyon.
Is it a choice whether one needs to jump Snake Canyon or not?

Suppose someone age 55 is working, earns 100K per year and is living on 75K per year. Typical middle class income and lifestyle.
They are saving 25K per year for retirement. 25% is a good savings rate.
So far they have accumulated 1 million dollars. That's a good sum, probably more than most have.
Suddenly they get laid off. They can't find another job. This happens to plenty of people.

So they have 1 million dollars and spend 75K per year. 7.5% withdrawal rate required
With some major belt tightening they get their spending down to 50K per year. You can only cut so much, unless you move to Slab City
Their starting withdrawal rate has to be 5%.
You say they are trying to jump Snake Canyon and need a rocket-propelled portfolio, i.e. a lot of stocks. Treasury bonds have negative real yields out to 20-years

They didn't choose 5% withdrawal rate. Circumstances chose it.

With negative real yields on safe Treasuries, say -1.0% the max withdrawal rate is probably about 1.5%. Retirees are being forced into risky assets.
The choice is take risk or live on $15,000 per year, which is $1,250 per month. The choice between Iraq and a hard place.
I would say they look for another job, even if it's only for $20K/yr, and reduce their spending to $40K/yr to help them limp along until they can get SS.
YDNAL
Posts: 13774
Joined: Tue Apr 10, 2007 4:04 pm
Location: Biscayne Bay

Re: The Glidepath Illusion (it doesn't work)

Post by YDNAL »

Browser wrote:Also consider the discussion of glidepath by Wade Pfau:
But when you allow the stock allocation to change over time, what I find in these simulations is that retirees are better off by starting with a lower stock allocation but then actually increasing it over time, rather than decreasing it.
Retirees don't typically adjust their AA and neither do TR Funds in retirement - like Vanguard when it gets to 30/70 a la TR Income VTINX. You could start retirement with "a lower allocation" to riskier Assets to avoid getting slammed with a 50% drop in the first few years of a long retirement. Ability and Need for risk... that's all!
Suppose someone age 55 is working, earns 100K per year and is living on 75K per year. Typical middle class income and lifestyle.
They are saving 25K per year for retirement. 25% is a good savings rate.
I'm sure Uncle Sam wouldn't like a $100K earner to pay $0 in taxes to be able to spend $75K and save $25K.
Landy | Be yourself, everyone else is already taken -- Oscar Wilde
pkcrafter
Posts: 15461
Joined: Sun Mar 04, 2007 11:19 am
Location: CA
Contact:

Re: The Glidepath Illusion (it doesn't work)

Post by pkcrafter »

jcKing wrote:
Results (note balance figures are give as min < average < max)
50:50: balance 52,653 < 91,712 < 136,601, exposure 18,144, balance/exposure 5.1
80->20: balance 60,280 < 80,601 <104,954, exposure 14,186, balance/exposure 5.7
20->80: balance 45,692 < 105,019 < 180,842, exposure 22,918, balance/exposure 4.6
100 switch 0: balance 91,801 < 95,223 < 102,650, exposure 15,967, balance/exposure 7.1
John, I'm not sure how you get 91,801 as a minimum balance on the 100/0 portfolio. The minimum should not be much different that the 80/20, and a lower minimum would be expected since the dispersion would even greater. In Arnott's data, the 80/20 portfolio has the lowest minimum and minimums increase with bond exposure. Your data shows just the opposite and I am guessing that using 0 for fixed income would seem to favor a portfolio with higher equities as your calculations show. Certainly a 20/80 portfolio is punished with 0 return from non-equity.

Arnott also notes that the 20/80 inverse portfolio does have higher variability of returns, but the variability is mostly all on the upside.

Paul
When times are good, investors tend to forget about risk and focus on opportunity. When times are bad, investors tend to forget about opportunity and focus on risk.
cjking
Posts: 2039
Joined: Mon Jun 30, 2008 4:30 am

Re: The Glidepath Illusion (it doesn't work)

Post by cjking »

Tigermoose wrote:cjking,

With your opportunistic switching strategy, what do you do to mitigate the risk of high inflation eating away what you thought was enough in fixed income during retirement?
The premise of the original article was that an annuity is bought on the 65th birthday, so what happens after that is outside the scope of the exercise.

However, let's say you don't buy an annuity, and plan for the standard scenario where you spend your portfolio over 30 years, and if still alive at the end of the 30th year of retirement, presumably shoot yourself in the head. The assumed return for safe assets in my simulation was 0% real, i.e. your money is keeping up with inflation, but doing no better. In real life this should be achievable, there will be times like the present when maybe it isn't, but on average over a long period I would guess you could in fact get up to 1% more than inflation.
garlandwhizzer
Posts: 3567
Joined: Fri Aug 06, 2010 3:42 pm

Re: The Glidepath Illusion (it doesn't work)

Post by garlandwhizzer »

I agree with the tone of Arnott's article and even though I'm 65 I still have 70% invested in equities. I do not personally accept the glide path concept at all. But it is not a question of knowledge or intelligence or research but rather one of temperament. If one cannot emotionally tolerate the inevitable ups and downs, the volatility and the bumps in the road of equities, glide paths and bonds are appropriate. ASSUMING OF COURSE THAT ONE HAS ENOUGH MONEY TO BE 100% CERTAIN THAT THEY WON'T RUN OUT, EVEN WITH THE LIKELY PROSPECT THAT TREASURIES AND TIPS REAL RETURNS GOING FORWARD WILL BE AT OR NEAR ZERO FOR A DECADE OR MORE, AND ADD TO THAT THE UNCERTAINTY OF HOW LONG ONE IS GOING TO LIVE AND HOW MUCH MEDICAL CARE AND LONG TERM CARE IS GOING TO COST. If one is certain of all those things, no problem with the glide path. If not, there is substantial risk, far in excess of the volatility of stocks in my opinion. Personally I don't put much stock in glide path models based on the past because at no time in the past history of the US have bond returns been as low as they are now. Asset returns going forward is not just related to asset mix but, significantly, what price you pay for those assets. The current E/P ratio of Vanguard's funds are listed below.

TSM 6
TBM 59
ITT 161

Would you rather pay $161 for a dollar of earnings or $6? Or would you rather buy the TIPS fund and guarantee yourself a loss of 1% a year in real dollars? The arithmetic to me at least looks compelling for having significant exposure to equities whatever your age if your temperament will allow you to sleep at night.

Garland Whizzer
User avatar
Epsilon Delta
Posts: 8090
Joined: Thu Apr 28, 2011 7:00 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Epsilon Delta »

Browser wrote: Both the 50/50 and 80->20 portfolios have the same average 50/50 stock-bond allocation over time; they differ in the temporal distribution of that allocation.
The 80->20 only has a 50/50 average if you do a time average. If you weigh by dollars invested it is considerably less than 50% stocks, perhaps as little as 35% stocks. This is not because the money is growing, it is because the investor is continually putting in new money, which is invested more conservatively due to the glide path.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Browser »

yobria wrote:Mr. Arnott is confused. Holding return constant, an optimal glide path distributes risk (major stock crash) evenly over your investing life. This is the expected case going forward, regardless of any historical data.

Plus it's just common sense - if you "front load the risk" (assume more earlier, less later vs static), you have more time to recover if the risk shows up.
One concern I have is that the glidepath method actually systematically reduces your ability to recover financially from a crash. This is because you are running down the growth horsepower of your portfolio over time by progesssively diluting the equity allocation. It's a sort of inverse-rebalancing method in which I reallocate from stocks to bonds even during bear market periods in which my portfolio allocation is tilting more and more toward bonds anyway because of declining equity values. If I want to maintain an even risk profile over time, doesn't that imply that I should have a constant equity allocation and rebalance to maintain that allocation? The only reason I can see for reducing equity exposure is to reduce portfolio volatility, not to improve returns or risk-adjusted returns (there's no evidence from either historical data or Monte Carlo simulations that it does - which is the best we can do without a Ouija Board). At some point it becomes very important to reduce portfolio volatility; e.g., close to retirement when expected returns are less important than hanging on to what we've accumulated. But that doesn't require adopting a mechanical, inefficient risk-reward strategy throughout my entire investment horizon.
We don't know where we are, or where we're going -- but we're making good time.
User avatar
ObliviousInvestor
Posts: 4212
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: The Glidepath Illusion (it doesn't work)

Post by ObliviousInvestor »

Browser wrote:If I want to maintain an even risk profile over time, doesn't that imply that I should have a constant equity allocation and rebalance to maintain that allocation?
Yes. Many people, however, (for various reasons) do not want to maintain an even risk profile over time.
Mike Piper | Roth is a name, not an acronym. If you type ROTH, you're just yelling about retirement accounts.
cjking
Posts: 2039
Joined: Mon Jun 30, 2008 4:30 am

Re: The Glidepath Illusion (it doesn't work)

Post by cjking »

pkcrafter wrote:jcKing wrote:
Results (note balance figures are give as min < average < max)
50:50: balance 52,653 < 91,712 < 136,601, exposure 18,144, balance/exposure 5.1
80->20: balance 60,280 < 80,601 <104,954, exposure 14,186, balance/exposure 5.7
20->80: balance 45,692 < 105,019 < 180,842, exposure 22,918, balance/exposure 4.6
100 switch 0: balance 91,801 < 95,223 < 102,650, exposure 15,967, balance/exposure 7.1
John, I'm not sure how you get 91,801 as a minimum balance on the 100/0 portfolio. The minimum should not be much different that the 80/20, and a lower minimum would be expected since the dispersion would even greater. In Arnott's data, the 80/20 portfolio has the lowest minimum and minimums increase with bond exposure. Your data shows just the opposite and I am guessing that using 0 for fixed income would seem to favor a portfolio with higher equities as your calculations show. Certainly a 20/80 portfolio is punished with 0 return from non-equity.

Arnott also notes that the 20/80 inverse portfolio does have higher variability of returns, but the variability is mostly all on the upside.

Paul
Think you've misunderstood the strategies, have another read of the original post. Only 50:50 is a conventional Boglehead fixed allocation portfolio, there is no 80:20 or 100:0 portfolio in the mix. The "80->20" and "100 switch 0" strategies are not different versions of the same thing, with different asset allocations, they are utterly different strategies. I won't explain again, because this is in the original post. (But tell me if it's still not clear and I will try and expand on what I said there.)

See also my posts to Rodc, in fact the "100 switch 0" strategy" would historically have achieved whatever target you aimed for, up to 106,000, in every single rolling period, i.e. the minimum always >= the target. The target in the original simulation you quote was 91,712, the minimum at 91,801 beats that, by design.

You might be right that 0% assumed return is slightly conservative, anything between 0% and say 1% would arguably be a reasonable assumption for very safe assets. (Remember I am assuming 0% volatility, so we are really talking cash rather than bonds here.) Using 0% instead of say 1% would slightly favour high equity allocations, if one strategy had a higher allocation than others, but actually the strategy with the highest allocation does worst. The best strategy has the second-lowest average allocation at 15,967 and the second-best strategy has the lowest average equity equite exposure of 14,186. If you really believe a difference of 11%-13% (depending which way round you measure it) in equity exposure significantly handicaps the second-best strategy, let me know and tomorrow I will re-run with safe assets set to a 1% return, and post the figures.
Tigermoose
Posts: 823
Joined: Sat Jul 21, 2012 5:02 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Tigermoose »

cjking wrote:
Tigermoose wrote:cjking,

With your opportunistic switching strategy, what do you do to mitigate the risk of high inflation eating away what you thought was enough in fixed income during retirement?
The premise of the original article was that an annuity is bought on the 65th birthday, so what happens after that is outside the scope of the exercise.

However, let's say you don't buy an annuity, and plan for the standard scenario where you spend your portfolio over 30 years, and if still alive at the end of the 30th year of retirement, presumably shoot yourself in the head. The assumed return for safe assets in my simulation was 0% real, i.e. your money is keeping up with inflation, but doing no better. In real life this should be achievable, there will be times like the present when maybe it isn't, but on average over a long period I would guess you could in fact get up to 1% more than inflation.

Is the amount needed to presume a 0% real return the low or high amount in your calculations? I am referring to the following statement.

"One caveat with regard to opportunistic switching: it is sensitive to target, setting the target too high could result in a much worse worst-case scenario. I doubled the target, and the worst final balance fell from 91,801 to 68,831, and this also lowered the average balance/exposure ratio to 4.7."
Institutions matter
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Browser »

ObliviousInvestor wrote:
Browser wrote:If I want to maintain an even risk profile over time, doesn't that imply that I should have a constant equity allocation and rebalance to maintain that allocation?
Yes. Many people, however, (for various reasons) do not want to maintain an even risk profile over time.
I thought that's what yobria was throwing out as the reason to use glidepath - he states:
an optimal glide path distributes risk (major stock crash) evenly over your investing life
.
I'm just pointing out that a fixed AA is more suitable for achieving that objective if that's what you're trying to do. Whether it's a worthy objective is an individual decision. I think it improves risk-adjusted returns, but it's pretty hard to overcome the intuitive appeal of the glidepath illusion.
We don't know where we are, or where we're going -- but we're making good time.
555
Posts: 4955
Joined: Thu Dec 24, 2009 6:21 am

Re: The Glidepath Illusion (it doesn't work)

Post by 555 »

Epsilon Delta wrote:
Browser wrote: Both the 50/50 and 80->20 portfolios have the same average 50/50 stock-bond allocation over time; they differ in the temporal distribution of that allocation.
The 80->20 only has a 50/50 average if you do a time average. If you weigh by dollars invested it is considerably less than 50% stocks, perhaps as little as 35% stocks. This is not because the money is growing, it is because the investor is continually putting in new money, which is invested more conservatively due to the glide path.
Exactly. I think Browser was taken in by this sleight of hand. I wonder what the SEC would think.
rkhusky
Posts: 17768
Joined: Thu Aug 18, 2011 8:09 pm

Re: The Glidepath Illusion (it doesn't work)

Post by rkhusky »

Browser wrote:
ObliviousInvestor wrote:
Browser wrote:If I want to maintain an even risk profile over time, doesn't that imply that I should have a constant equity allocation and rebalance to maintain that allocation?
Yes. Many people, however, (for various reasons) do not want to maintain an even risk profile over time.
I thought that's what yobria was throwing out as the reason to use glidepath - he states:
an optimal glide path distributes risk (major stock crash) evenly over your investing life
.
I'm just pointing out that a fixed AA is more suitable for achieving that objective if that's what you're trying to do. Whether it's a worthy objective is an individual decision. I think it improves risk-adjusted returns, but it's pretty hard to overcome the intuitive appeal of the glidepath illusion.
I would think that a major stock crash would have more of a negative effect when you are closer to retirement, i.e. you may have to start withdrawing money, locking in losses, before the market recovers. Therefore, having less in equities, the closer one is to retirement, evens out the risk of running out of money in retirement. Also, as one gets older, it is more likely that health issues could force you out of employment and age can make it harder to get a new job should one be laid off.
yobria
Posts: 5978
Joined: Mon Feb 19, 2007 10:58 pm
Location: SF CA USA

Re: The Glidepath Illusion (it doesn't work)

Post by yobria »

Browser wrote:One concern I have is that the glidepath method actually systematically reduces your ability to recover financially from a crash. This is because you are running down the growth horsepower of your portfolio over time by progesssively diluting the equity allocation. It's a sort of inverse-rebalancing method in which I reallocate from stocks to bonds even during bear market periods in which my portfolio allocation is tilting more and more toward bonds anyway because of declining equity values. If I want to maintain an even risk profile over time, doesn't that imply that I should have a constant equity allocation and rebalance to maintain that allocation? The only reason I can see for reducing equity exposure is to reduce portfolio volatility, not to improve returns or risk-adjusted returns (there's no evidence from either historical data or Monte Carlo simulations that it does - which is the best we can do without a Ouija Board). At some point it becomes very important to reduce portfolio volatility; e.g., close to retirement when expected returns are less important than hanging on to what we've accumulated. But that doesn't require adopting a mechanical, inefficient risk-reward strategy throughout my entire investment horizon.
Here's how I think about it, very simple, no Monte Carlo simulations needed. I want to hold my risk profile constant as well. I define that as a holding the damage done by a 50% decline (no recovery) in stock prices at any time during my working life constant.

If you hold a lazy portfolio, you'll see that the damage done by such a decline increases as you approach retirement. A stock crash when you're 25 doesn't hurt at all, one when you're 65 is severely damaging. An optimal glide path model holds this risk constant instead of backloading it (when you have less time and money to recover). You can use Excel to build a model factoring in your unique variables to estimate the optimal expected path.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Browser »

555 wrote:
Epsilon Delta wrote:
Browser wrote: Both the 50/50 and 80->20 portfolios have the same average 50/50 stock-bond allocation over time; they differ in the temporal distribution of that allocation.
The 80->20 only has a 50/50 average if you do a time average. If you weigh by dollars invested it is considerably less than 50% stocks, perhaps as little as 35% stocks. This is not because the money is growing, it is because the investor is continually putting in new money, which is invested more conservatively due to the glide path.
Exactly. I think Browser was taken in by this sleight of hand. I wonder what the SEC would think.
Er.. you forgot about rebalancing. Quoting Arnott from the lined article:
Note, if we systematically replace equities with bonds every year so that we are a 50/50 investor at the midpoint of our career, our returns will fall into the same return distribution, over time, whichever path we pursue. Our average allocation will be 50/50 in all three cases!
We don't know where we are, or where we're going -- but we're making good time.
User avatar
Epsilon Delta
Posts: 8090
Joined: Thu Apr 28, 2011 7:00 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Epsilon Delta »

Browser wrote:
555 wrote:
Epsilon Delta wrote:
Browser wrote: Both the 50/50 and 80->20 portfolios have the same average 50/50 stock-bond allocation over time; they differ in the temporal distribution of that allocation.
The 80->20 only has a 50/50 average if you do a time average. If you weigh by dollars invested it is considerably less than 50% stocks, perhaps as little as 35% stocks. This is not because the money is growing, it is because the investor is continually putting in new money, which is invested more conservatively due to the glide path.
Exactly. I think Browser was taken in by this sleight of hand. I wonder what the SEC would think.
Er.. you forgot about rebalancing. Quoting Arnott from the lined article:
Note, if we systematically replace equities with bonds every year so that we are a 50/50 investor at the midpoint of our career, our returns will fall into the same return distribution, over time, whichever path we pursue. Our average allocation will be 50/50 in all three cases!
No, I did not forget about the re-balancing. Mr. Arnott's statement is roughly true if we consider a lump sum investment at the beginning of the glide path. However in this case of yearly contributions only 1/41 of the sum gets this treatment, the next years 1/41 has a glide path from 78.5->20 so is invested 49.25/50.75 on average. Averaging this for all 41 years gets 35% average in stocks.
555
Posts: 4955
Joined: Thu Dec 24, 2009 6:21 am

Re: The Glidepath Illusion (it doesn't work)

Post by 555 »

Browser wrote:
555 wrote:
Epsilon Delta wrote:
Browser wrote: Both the 50/50 and 80->20 portfolios have the same average 50/50 stock-bond allocation over time; they differ in the temporal distribution of that allocation.
The 80->20 only has a 50/50 average if you do a time average. If you weigh by dollars invested it is considerably less than 50% stocks, perhaps as little as 35% stocks. This is not because the money is growing, it is because the investor is continually putting in new money, which is invested more conservatively due to the glide path.
Exactly. I think Browser was taken in by this sleight of hand. I wonder what the SEC would think.
Er.. you forgot about rebalancing. Quoting Arnott from the lined article:
Note, if we systematically replace equities with bonds every year so that we are a 50/50 investor at the midpoint of our career, our returns will fall into the same return distribution, over time, whichever path we pursue. Our average allocation will be 50/50 in all three cases!
Bingo! If the SEC investigates Ray Lucia then they should investigate Rob Arnott.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: The Glidepath Illusion (it doesn't work)

Post by Browser »

Here's how I think about it, very simple, no Monte Carlo simulations needed. I want to hold my risk profile constant as well. I define that as a holding the damage done by a 50% decline (no recovery) in stock prices at any time during my working life constant.
You are defining risk as downside volatility. But downside volatility can only be relatively constant over time with a static AA. Glidepath is supposed to systematically reduce downside portfolio volatility (defined as percentage change), not maintain it constant. :confused

Now maybe you're talking about dollar loss instead of percentage loss - the idea being that a portfolio worth $100K and 100% invested in stocks could result in a $50K dollar loss if there is a 50% decline in stocks. But if the portfolio has grown to $500K, then the stock allocation would need to be reduced to just 20% in order that a 50% stock market decline would result in the same $50K loss. (Actually, this is a simplification, since it depends on what the bond portion of the portfolio does as well. With 80% in bonds, you would only need a 12.5% increase in bond values to completely offset the dollar loss from equities). In any event, do you know of any glidepath funds that have the objective of producing a constant dollar loss risk? I don't. Here's what Vanguard says:
Vanguard Target Retirement Funds offer a diversified portfolio within a single fund that adjusts its underlying asset mix over time. The funds provide broad diversification while incrementally decreasing exposure to equities and increasing exposure to bonds as each fund’s target retirement date approaches.
I don't see anything about trying to even out out dollar loss risk, do you? You might assume that's what a glidepath fund will do, but that's a pretty naive assumption IMO. I think people make a lot of unfounded assumptions about these things. Maybe you can roll your own to try to accomplish this, but you'll probably have to put on your Monte Carlo hat in order to determine the "optimal glidepath" for actively managing your portfolio. I'd be interested in seeing the strategy you come up with.
We don't know where we are, or where we're going -- but we're making good time.
Easy Rhino
Posts: 3278
Joined: Sun Aug 05, 2007 11:13 am
Location: San Diego

Re: The Glidepath Illusion (it doesn't work)

Post by Easy Rhino »

I dunno, there's something weird, but intriguing, about Arnott's numbers. The inverse glidepath scenario has a higher minimum and 10 percentile return than the other strategies. If we worry about downside risk, this would seem to be better. But that seems so horribly counterintuitive... maybe we should ask Arnott for his numbers? maybe he incorporated a smooth inflation rate rather than year-by-year inflation or something? Maybe the price for an inflation adjusted SPIA varies depending on the infation at the time of purpose? I dunno, something just seems strange.

One gripe I have bout glidepaths is they seem start going conservative too late, and they do it too steeply (right around retirement). As noted, Arnott's glidepath is very atypical, it looks like it starts off 80/20, gets 50/50 after 20 years, and ends up at 20/80 at 40 years.
User avatar
nisiprius
Advisory Board
Posts: 52219
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: The Glidepath Illusion (it doesn't work)

Post by nisiprius »

Browser, It has nothing to do with returns. It has nothing to do with risk-adjusted returns. It has everything to do with risk tolerance. An investor with higher risk tolerance should have a higher stock allocation. An investor whose risk tolerance does not decline with age could well consider a stationary allocation. An investor whose risk tolerance does decline with age should reduce their stock allocation with age. Because not achieving some optimum portfolio according to some metric is an inconsequential mistake, but exceeding your risk tolerance is a big mistake.

All the economist-type papers I've seen do factor risk tolerance into their equations in an explicit and serious way. Arnott doesn't even pay lip service to it.

Now, there do seem to be people who believe that risk tolerance doesn't exist. Or that one can voluntarily and quickly change one's risk tolerance by a simple act of will. Or that there is such a thing as an objectively, measurably "right" or "wrong" risk tolerance. I am not among them. As evidence of the serious nature of exceeding one's risk tolerance, I could point to several example, but let me choose Ben Stein's despairing and panicky columns from 2008:
I decided to [file] my stock transaction receipts… once upon a time I liked to review them. Now, they're a catastrophe. My losses are staggering even in the most plain vanilla index funds. In the emerging markets and developed markets, investments that had once provided immense gains, the losses are worse. Even in my beloved RQI, the high-income, leveraged REIT index fund, the losses are beyond belief.

Instead of just jumping off my balcony, which wouldn't get me more than a broken leg, I am going to try to make some sense of what has happened…. I am more than the mere composite of the stocks and bonds I own. I hate myself for being so dependent on how much money I have for my self image. As for retirement, well, I get sick and bored if I am not on the road most of the time anyway. The reason I am not suicidal right now is that I have a wife who would be fine with it if we had to live a more modest life style.
I am not sure what his stock allocation was during that time, but I would suggest that if you have Ben Stein's risk tolerance, you should not have the stock allocation Ben Stein had in 2008.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
Post Reply