Portfolio Expectations: What a Difference a Year Makes

Asset allocation is important to reduce risk.  Approximately 90% of portfolio total return comes from the asset classes you select and the relative weights you give them.

The word “never” is a dangerous term to use, particularly with respect to investing, but we believe it is probably safe to say that an allocated portfolio will never return as much as the best performing asset class, and it will never perform as badly as the worst performing class.

The important assumption is that an allocated portfolio is “expected” over the long-term to produce the blended mean return of the asset classes in the portfolio, and to do so with less volatility and less extreme draw-downs than the more volatile classes in the portfolio.

Unrealistic Expectations:

“Expected” is a tricky operative term.  We always ask new clients to indicate what returns they expect from their portfolio.  The answers are often way out of line with practical potential.  One reason for unrealistic expectations is memory of bubble years, while another may simply be inadequate information. This article is intended to deal with the information aspect of expectations.

Time Frame:

The largest issue with expectations may be the historical time window you use to develop a view.  Some periods create higher expectations than others. We think you need to study the long-term and worst periods, as well as recent periods and current trends to make sure that you understand how things may turn out.

Simple Allocation Example:

Let’s look at one very simple asset allocation (not a recommendation — just an illustration) to see how different time periods support different expectations.

A 60% S&P 500 / 40% Lehman Aggregate Bonds allocation is a classic moderately aggressive balanced portfolio allocation — balance with an equity bias.

That allocation is a popular one, such that Vanguard has offered a balanced fund on that simple allocation formula since 1986 (VBINX).

US Stock/US Bonds as a Personal Risk Tolerance Yardstick:

While we would recommend more than two asset classes in a well constructed portfolio, we think that studying various combinations of S&P 500 (proxy SPY or IVV) and Lehman Aggregate Bonds (proxy AGG or BND) is a useful way to begin thinking about the mix of equity and fixed income assets that is suitable for you — particularly as it relates to risk and your risk tolerance.

You can diversify assets beyond this simple yardstick in a variety of ways with other US equity and fixed income categories, foreign varieties of equity and fixed income, and with other asset types such as real assets, commodities and cash.

In any event, we think being familiar with the characteristics of a spectrum of US equity and fixed income is important as a starting point for developing perspective on return opportunities and risks.

Weekly Short-Term Allocation Results Online:

Each week we update a short-term view of eleven blends (e.g. 90/10 through 10/90) of the S&P 500 and Lehman Aggregate Bonds yardstick on our website.

Very Long-Term 60/40 (81 years):

Vanguard published the long-term returns, best year and worst year, as well as the frequency of down years for several allocation ratios, including 60/40 from 1926 through 2006.

Their study determined the average return to be 8.9% with the worst year down nearly 27% in 1931, the best year up nearly 37% in 1933 (two years later), and down years approximately 1 in 4 years.

click images to enlarge

Short-Term 60/40 (1-wk, 4-wk, 13-wk, 26-wk, 52-wk):

Our short-term monitor of the 60/40 allocation (along with ten other US stock / US bond allocations) reveals more than a 22% decline for a rolling 52-week period.  That, not surprisingly, is near the worst experience in 1931.

If history slightly mimics itself, the long-term picture may hold out hope for a recovery of 30% or more in 2010.  Unfortunately, future history isn’t quite so predictable as past history is certain.

A $1,000,000 portfolio as of 12/31/2007 in a 60/40 allocation would be worth about $780,000 today.  A 30% recovery, would put the portfolio at about $1,014,000.

Long-Term 60/40 (22 years 1986-2007):

Twenty two years isn’t as long as 81 years, but it does span a lot of territory and is more closely aligned with the kind of globalization of trade we experience today than the 81 year history.

Our table below shows the 22 years from 1986 through 2007 to be a kinder time than the 81-year history in terms of frequency of down years (approximately 1 in 5), with a much less severe wost year (-9.16% versus -24.7%), and a higher annualized rate of return (10.17% versus 8.9%).

Sadly, that return was not available for all other starting years.  For example, a 60/40 allocation initiated at the beginning of 2000 returned only 4.03% through 2007.  The good news, and to the point of allocation, is that a 100% S&P 500 allocation made at the same time, returned only 1.68%.

The 100% S&P 500 allocation had only 1 in 5 down years from 1986-2007, but the worst down year was -22.10% as opposed to the -9.16% worst year for the 60/40 allocation.

Stock vs Bonds vs 60/40 Blend 1986-2007

The S&P data is shown in yellow highlight.  The Aggregate Bond and 60/40 data are shown in yellow, pink or green depending on their value.  Yellow indicates the value equals the S&P 500 value. Pink indicates the value is worse than the S&P value, while green indicates the value is better than the S&P value.

But Extend One Year and See Quite a Different Story (1986-2008):

If for the sake of illustration, we treat the current 2008 market results as calendar year 2008, we see even more clearly how important the time window selected is for creating return expectations.

The 22 years from 1986 through 2007 is nearly the same length of time as the 23 years from 1986 through 2008, but what a different story.

A 60/40 allocation initiated in 1986 and held for 23 years through today produced a 8.53% return (remarkably similar to the 8.9% of the 81 years from 1926-2006).  It experienced 5 down years out of 23 years overall (~ 1 in 4.5 years).  The worst year was -21.82%, much closer to the 81 year worst of -24.7% than the -9.16% for the 22 years from 1986-2007.

What a difference one year makes in setting up expectations!

Stock vs Bonds vs 60/40 Blend 1986-2008 (11/10/08)

Reality Check on Our Prior Advice:

About two years ago (01/02/07) Seeking Alpha published one of our blog posts titled, “Realistic Expectations for Returns Going Forward“.  We were probably a bit too optimistic in the upper end of a possible range, but we were definitely cautioning against too much optimism in the face of a bubbly emerging market.

A central theme was the difficulty of consistently beating benchmarks in successive years without some down years, and the limits of performance by even the recognized best institutional managers.

Back then we said;

The reality of plausible long-term returns, consistent with likely investor behavior in the face of down periods, is not as great as many would think. Most investors can’t take the pain of staying in during thick and thin. Consequently, they get out too late in down legs, and stay timid too long and miss important parts of up legs. The result is performance below their benchmark.

There are the shooting stars each year that double or more, and if we could pick those correctly and never make bad choices, we could become rich beyond imagination. The problem is that it doesn’t happen.

Look at the most celebrated current mutual fund manager Bill Miller. He exceeded his 15-year (1991-2005) S&P 500 benchmark average 12.7 % with his 15-year Legg Mason Value Trust [LMVTX] portfolio average of 16.4%. Very impressive, but almost nobody has done that with such consistency. Others have done better in terms of total return, but not necessarily by exceeding the benchmark in each discreet year.

Looking more broadly …”World’s Best Money Managers” … for large-cap US equity (roughly the same category as the S&P 500), as published by Nelson’s (part of Thomson Financial). These results include institutional portfolios, such as pension funds and separate accounts, as well as mutual funds.

… the average 10-year (1996-2005) performance of the top 40 managers was 14.95% versus Bill Miller’s 10-year 17.57% average. There were 3 of the top 40 managers who did better than Miller with returns of 18.05%, 22.67% and 26.97% (Capital Management Associates, Warwick Capital Management and Hillman Capital Management, in that order).

Certainly, managers mandated to work in a particular region, style or sector could have long excess return streaks, but their mandate won’t let them get out of the way when the cycle turns against them. And, apparently managers with broad mandates have not been able to do too much more than about 15% on average over the last 10 or so years according to Thomson Financial.

In summary, the US market produced a recent 10-year return in the vicinity of 10%, and the best money managers with broad mandates have produced returns in the range of 13% to 18% over the same recent 10-year period.

We would conclude that unless you are a major risk taker, a minor genius, and incredibly consistent, you should plan on a range of 9% to 15% with a tendency toward 10% for your own investments.

Harvard, Yale and Princeton Endowments:

Top university endowments have billions in assets, cadre’s of full-time analysts, pick of the cream of the crop in institutional advisors, access to asset classes not available to any but the very-rich and institutions — far more resources and options than almost anyone reading this article will ever have.

How have they done? What can their results tell us about realistic expectations?

The headline story for this week’s Barron’s, “Crash Course” was about how even the best endowments have been kicked in the stomach by the current market.  This year will go a long way to bring their long-term results back toward “normal”.

The article pointed to 10-year trailing returns (as of June 30) for Harvard, Yale and Princeton as 13.8%, 16.3%, 14.2%.  They said the average endowment has returns of about 6% for the 10-year period versus 3% for the S&P 500.

Barron’s speculates that, since June 30, endowment returns may have been in the range of minus 25%.  If that were so, the 10-year returns of the Ivies, might be something like 11+% to 13.5+% as of 12/31/2008.

Conclusion:

Diversify asset classes in any long-term portfolio design.

Look at multiple short-term, long-term and very long-term periods before setting up expectations.

Be aware of current trends and things coming down the track when setting up short-term expectations.

Be realistic about potential returns, and whether you can endure the volatility and maximum draw-down years associated with the return you wish to achieve.

If you expect long-term returns in the 10+% range, be prepared for a volatile ride in riskier assets.

If you are thinking about bigger numbers, such as 12% to 15% annualized over long periods, you probably can’t get it done with just US stocks and bonds, if you can do it at all.

Comments

[Larry House:]

Interesting piece. I think most of us have to lower expectations when it comes to annualized return. I know I have, and I am increasing my safety requirement (if there is such a thing as "safety" in this investment climate). I know many investors just dismiss bonds as a terrible investment, but if stocks are likely to be flat for who knows how long, bond yields and relative safety seem like a good place to be.

2008 Nov 11 04:18 PM

[dshort:]

Thanks, Richard. This article is another reason why you're in my watchlist.

The 81-year time frame really resonates with those of us who study market history. And when analyzing long-term performance, it's also important to factor in the impact of inflation, as these two charts illustrate:

dshort.com/charts/SP-C...
dshort.com/charts/SP-C...

Thanks again!

2008 Nov 11 06:30 PM

[QVM Group:]

'70s stagflation survivor.

Thanks, and very helpful charts on your site. I recommend other readers view them.

2008 Nov 11 10:23 PM