Loading...
Symbols:
Authors:

Portfolio Planning and the Lost Decade

by: Geoff Considine June 13, 2008 | about stocks: DJP / EFA / ICF / IDU / IVV / IWM / SPY / TIP    

Geoff Considine

About this author:
Become a Contributor Submit an Article
  • Font Size:
  • Print

Assessing the Lost Decade

The Wall Street Journal wrote an article back in March in which they noted that the S&P 500 had delivered essentially no return over the past nine years. They dubbed this period “The Lost Decade,” and the article has since received considerable attention in the wake of declining markets and waves of bad news for a range of sectors. The big thing that everyone seems to be debating is whether this is simply a normal part of the economic cycle or if “the wheels are coming off.” To put this decade-long period of low returns on the S&P 500 in historical perspective, I liked the graphical presentation by Bespoke Investment Group.

In this article, they look at rolling ten-year returns for the S&P 500. Their results show that the S&P 500 has most recently delivered a ten-year return comparable to the 1970’s, which was not a good decade for stocks. The historical rolling ten-year periods have exhibited an enormous range of outcomes, and we are currently in a down period. While there is plenty of grim news out there, I wanted to look at this performance in a larger context.

Let’s start by noting that the S&P 500 is a stock index that has its allocations to individual stocks determined by their market capitalizations. There are plenty of good reasons why investors might want to allocate even their large-cap domestic holdings in some other way. Before the meltdown in financials, about 20% of the S&P 500 was allocated to financial firms. The S&P 500 is one way to create an index of the performance of stocks, but it is not the only way. This said, many investors and portfolios managers use the S&P 500 as a benchmark, so we will forge ahead on that basis.

A Statistical Perspective

Instead of looking just at market history, let’s consider Monte Carlo simulations of long-term expected performance for the S&P 500. Quantext Portfolio Planner [QPP] is our portfolio management tool and its uses the Monte Carlo approach. To analyze the relative probabilities of long runs with zero or near-zero total return, I run QPP with all default settings and the standard initialization data of three years (through May 2008). QPP simulates a range of possible future outcomes via Monte Carlo simulation, and thereby allows me to estimate the probability of a decade-long period with zero returns (or worse).

To make this really concrete, let’s make a test case of a 40-year-old who is investing for retirement. QPP provides the following projections for his/her cumulative returns in about nine-year’s time and in about ten-year’s time, assuming investment in IVV or some other low-cost S&P 500 index fund:


100% of Funds Invested in S&P 500 (IVV)

What is most striking when you look at these results is that the possibility of cumulative returns around zero falls between the 5th percentile and 10th percentile---even over a ten-tear period. This means that there is between a 1-in-20 and 1-in-10 chance of having returns at or below zero for a holding period of ten years. This is remarkably consistent with the fact that we have seen six 10-year periods in the last hundred with returns as low as we have seen for the most recent ten-year period (see the Bespoke article linked above).

Theses results highlight the importance of using probabilistic models to plan for the future. The median outcome is for a cumulative return of 115% over the next decade, but there is a 1-in-20 chance of -9% cumulative returns. We experienced negative real returns for the S&P 500 over decade-long periods in the 1970’s and we are experiencing it again now.

The Power of Diversification

A major lesson to be drawn from this sort of analysis is why it is important to diversify effectively across asset classes. Let’s consider these same probability levels for a diversified portfolio. Rather than create a new portfolio, I will simply use one from a recent article.

I decided to use a portfolio of ETFs with 20% bonds (hence the fairly unimaginative name P20):


P20 Portfolio

This portfolio combines allocations to the S&P 500 (IVV) and Russell 2000 (IWM) with healthy allocations to REIT’s (ICF), utilities (IDU), commodities (DJP) and energy stocks (IGE). The bonds are allocated largely to TIPS (TIP). The astute reader will note that this portfolio ended up with zero allocation to emerging markets (EEM)—for details on why, read the article.


Invested in P20 Portfolio

QPP projects that this portfolio will generate considerably higher returns than the S&P 500, with less risk: these are the fruits of diversification. There is still an enormous range in the projected returns, but the bad outcomes (the 5th and 10th percentiles) have improved considerably.

Summary

So, let’s put the Lost Decade in perspective. The most recent decade has not been good for the S&P 500. These kinds of decade-long periods have happened before (most recently in the 1970’s). When viewed through QPP, we see that these periods are sufficiently likely that they must be planned for---anything with a 5-10% chance of occurring is worth keeping in mind. This risk can be managed through smarter asset allocation strategies. Investors who have been well-diversified relative to the S&P 500 have fared far better in this decade than those who over-emphasized the S&P 500---and this is likely to be the case in the long-term. This is why many pension funds and other institutional investors have been looking beyond major equity indices for a substantial portion of their assets.

Sadly, a substantial fraction of mutual fund managers are, in fact, ‘closet indexers’ and don’t stray far from this benchmark. This means that many investors have a significant probability of ending up with very low or even negative returns over decades to come. If we add the unfortunate tendency of retail investors to chase hot sectors, it is easy to see a major crisis in the making for current generations with self-directed retirement plans.

Related Articles

This article has 18 comments:

  •  
    Jun 13 12:25 PM
    Very interesting article and right on. Every investor (as opposed to the trader) should heed this insight. I know many pre-retirees NEED higher returns, but the fact is we are in a time of moderate to low average returns from the U. S. stock market. With the outlook for stock market returns in the 5-7% range, it is absolutely essential that a new allocation model be considered for greater return and lower risk. Great article.
    Reply
  •  
    Jun 13 01:47 PM
    Your articles are consistently interesting. I'd love to try QPP, but you are going to have to port it to OpenOffice for us Linux users.
    Reply
  •  
    Jun 13 02:24 PM
    Geoff--Another good article. However, I could not find your rationale for no allocation to EEM, either here or in the earlier articles you cite. I can think of a couple of reasons why you may have excluded EEM, but am interested in your thinking. Thanks.
    Reply
  •  
    Jun 13 04:04 PM
    Acercher:

    If you look at the original article in which I constructed the P20 portfolio, you will see that some of the portfolios did have EEM in them. If often find that EEM can be excluded because of the high correlations between EEM and the S&P500 (73%), EEM and EAFE (86%), and IGE and EEM (77%). From a statistical standpoint, major equity indices plus energy / natural resources indices capture the portfolio benefits of EEM in many portfolios.
    Reply
  •  
    Jun 13 04:05 PM
    Geoff--an interesting article as far as diversification goes. One reminder to ETFers, though, check the portfolio in the ETFs for overlap. IGE and IVV (23% of the portfolio) have 20% overlap in their top 10 holdings. To be fair, P20 doesn't appear to have a great deal of overlap overall.

    In down (and flat) markets, the wonders of diversification are always trotted out. Not that any of us, myself certainly included, can lay claim to the mantle of Warren Buffett, but he had this to say about diversification: Diversification serves as protection against ignorance. If you want to make sure that nothing bad happens to you relative to the market, you should own everything. There is nothing wrong with that. It’s a perfectly sound approach for somebody who doesn’t know how to analyze businesses."

    Remember, the website is Seeking Alpha.
    Reply
  •  
    Jun 13 09:35 PM
    First, you fail to specify what the defualt setting for the average return and volatility on the S&P500 is. The second thing is, I could calculate what that probability might have been with a calculator, no need for fancy-shmancy Monte-Carlo. You only need MC if you are doing a "dynamical" analysis where your decisions in subsequent periods depend on what happened in the meantime. The other way that MC is useful for financial plannning is if you want to consider the effect of regular contributions due to the dollar-averaging. Once again, for the purposes of this calculation of 10 year returns, you dont need MC (that is if you know what a log-normal distribution is)
    Reply
  •  
    Jun 13 10:50 PM
    A bunch of hooey. The positive comments are probably from cohorts. Its always easy to rationalize anything with a proprietary model that nobody can see. The bottom line is that the "lost decade" is real, just like the 70' and yes, let's not forget the 30's from which the Dow (when adjusted for inflation) did not recover until 1981 [nearly 50 years later]. Just maybe all you wishful thinkers will be wishing that the go-go 80's and 90's were just the last hurrah for a dying empire. If you want diversification, try another country
    Reply
  •  
    Jun 14 09:52 AM
    In your prior article you say "the first step is to create a series of possible portfolios". There is no methodology cited...its a subjective list that you then massaged using MC simulation. If one wants diversification without overlap just buy VTU and VEU; you have total diversification but zero alpha.
    Reply
  •  
    Jun 14 11:00 AM
    .My investing record has been as follows 1999:-38.25 2000:+92% 2001 +110.4 % 2002 -9.4% 2003 69.6% 2004 15.55 2005 29.3% 2006 28.4 2007 23% this year up 3.4%

    My portfolio for most of the decade was MO and RAI which threw off the dividends that enabled me to buy BUD BRK JNJ BNI WMT and others.My book explained the filters and formulas that helped that. Picking the right stocks and staying concentrated is paramount
    Reply
  •  
    Jun 14 09:47 PM
    Guys:

    The point of this article is simply that reasonable expectations for the equity risk premium (which are lower than what we have experienced over recent decades, all in) are consistent with "lost decades" that are not too infrequent in the S&P500. The vitriolic response is wierd--this is fairly basic stats. Now, what is not so basic is the fact that spreading assets in a smarter way that cap weighting makes lots of sense and helps a great deal.

    A number of critics seem not to understand the subtlety of what diversification means. It is not buying everything. My analysis of Berkshire's top holdings--a very concentrated portfolio--shows that the portfolios is actually very well diversified. Diversification is a measurable thing and you can have a concentrated portfolio that is also diversified. Buffett's quote is that WIDE diversification is protection against ignorance--this is a different thing entirely. The very fact that some Seeking Alpha readers find this controversial is argument enough for this kind of article.

    Also, as far as the suggestion that QPP is a proprietary model that noone can see. To the contrary, QPP is extremely well documented, tested, and used by a wide cadre of investors.

    Concentration and diversification are not contradictory terms...
    Reply
  •  
    Jun 15 10:28 AM
    Talk about touchy! I re-read all the comments posted and don't find anyone "vitriolic".... citing Buffett that "WIDE diversification is protection against ignorance" is right on point. However, it still doesn't explain how you allocated funds within the ETF portfolios. Backtest data would have been nice to try to validate your (again) subjective selections. Buffett's stock selection is subjective but obviously very good; asset allocation and diversification is fine but it always comes down to how good are your selections. Knowing how your portfolios would have performed in the past would provide some confidence in your recommendations. My own model portfolio which is not much different than your p40-p70 portfolios has been flat since last October. While better than the U.S. market in that time, not a world beater...I suspect your portfolios, in this time frame, would have produced similar results. BTW, I made a typo in my prior comment....meant VTI and VEU as a way to achieve total (ignorant?) diversification.
    Reply
  •  
    Jun 15 06:35 PM
    Geoff--no vitriol from me. I don't find your approach to diversification at all objectionable. Sure, diversifying beyond the S&P 500 (or any other index) makes sense, but doing so via funds that represent other indexes (either market or sector) means that there are, at the very least, four certainties: 1) the portfolio will never outperform the leading index or sector, 2) the portfolio will never underperform the trailing index or sector, 3) the portfolio will experience less volatility than the most volatile index or sector, and 4) the portfolio will experience more volatility than the least volatile index or sector. Again, there is nothing wrong with damping volatility in this manner, so long as that is the objective within the investing universe the portfolio encompasses. It does not, and cannot, guarantee positive returns over any time period be it short, intermediate or long. Moreover, taking the approach you describe means that it is no longer valid to look at portfolio return against the S&P 500. The returns would need to be weighed against an equal weighted blend of the portfolio constituents to determine the advantage of the weightings found in P20.

    Berkshire, or more specifically Buffett, doesn't look at risk premiums or benchmarks. Nor, I suspect, does Buffett care about diversification as it relates to returns during any time period. He simply seeks to buy good companies at reasonable prices. The reasonable price approach addresses the risk you are trying to reduce via diversification because any index/sector approach is not a business oriented approach to investing. From a business perspective, there is no inherent margin of safety in buying an index or sector fund. It is comparing apples to oranges. Whatever the measure of diversification within Berkshire's portfolio turns out to be, Buffett would likely consider the measure a curiosity at best.
    Reply
  •  
    Jun 16 11:47 AM
    User8470: I am sorry if my comments were too strident. For details on the specific portfolio construction, you have to go beyond the limits of this article or the original that proposed the portfolios. There is always a measure of subjectivity and I do not hold out these ETF portfolios as the best that is possible. I started with a range of broad asset classes and iterated (by hand) towards a quasi-efficient frontier using QPP. This is only quasi-efficient because I also assume that most investors will balk at going beyond certain limits in a range of asset classes. I also did not 'fine tune' the allocations too much--that gets pointless.

    CM in MA: Useful comments and I agree with points (1)-(4)--if investors could grasp these points, they would be far better off. We see in investor behavior (like performance chasing) that maony (if not most) investors don't understand these themes.

    As far as comparing P20 to an equal-weighted portfolio of these assets--that is an interesting idea but may not be a great benchmark--depends on the risk, etc. Interesting idea, though.

    I like your points on Buffett, too. The very high diversification benefits we find when we analyze Berkshire is fascinating. I can't explain this apparent paradox in his stated style and the fact that his portfolio looks like a textbook application of portfolio theory to me.

    Reply
  •  
    Jun 16 01:48 PM
    Geoff--yeah, don't know if the equal weighted portfolio would be a good benchmark or not. Seems to me, though, the only other choice would be to lock in the original P20 allocation as the benchmark for subsequent changes. My rationale for that is the need to isolate the effects of future reallocation decisions. After all, once the portfolio strays from equal weighting, it becomes an exercise in index/sector timing and rotation.

    On a related topic, are you familiar with the government's Thrift Savings Plan? www.tsp.gov The reason I ask is that it is potentially a giant laboratory to study investor behavior vis-a-vis asset allocation decisions and results. As of Nov 2007, there were 3.8 million participants with $231.5 billion in assets. The TSP provides three asset class choices: equities (large-cap, small/mid cap, international), bonds, cash. There are some recently introduced lifecycle funds consisting of these choices. Awhile back, there was some discussion of including other asset classes, but that has not happened yet. For all intent and purposes, reallocation is cost-free, so that removes a factor from the decision process. It is interesting that limits were recently established for interfund transfers--folks were rebalancing/reallocati... too frequently and driving up plan costs for other participants. So, even with just five choices, it would seem there were some people attempting to time the swings in the markets rather than using some basic asset allocation to smooth the long-term ride. Anyway, I couldn't say whether the government has ever studied the results plan participants have achieved, or whether they'd even be interested in doing so. However, it strikes me as a well-controlled experiment in investor behavior.
    Reply
  •  
    Jun 16 01:50 PM
    above should have read: "folks were rebalancing/reallocati... too frequently"
    Reply
  •  
    Jun 16 01:50 PM
    ok, I give up: I typed "reallocating&quo... correctly twice and both times it changed to reallocati...
    Reply
  •  
    Jun 16 04:24 PM
    CM in MA:

    The kind of study you are referring to has been done by DALBAR and is repeated every year (Google the term QAIB DALBAR):

    www.thorntonwealth.com...

    The statistics show that retail investors do chase performance and that this exacts an enormous cost. This is also seen in Morningstar's investor return stats. As I pointed out in some recent articles, the very first things for retail investors to do are (1) invest in a well-diversified low-cost portfolio, (2) avoid chasing performance.

    The DALBAR studies are shocking...and they find the same stuff every year.

    BTW, I was in the TSP when I was at NASA....
    Reply
  •  
    Jun 16 05:37 PM
    Geoff--thanks for the link to the DALBAR study. I gave it a quick look, but will read it in detail.

    I was in the TSP while in the military. I think it is a great program. Keeps things simple. Many 401k/403b plans could (and do) do worse than use the TSP as a model.
    Reply
More Comments on IVV »
More by Geoff Considine

Investment Resources
Money Management 

Research 

Seeking Alpha provides these ads as a service to its leading contributors
Articles on related themes
Asset Allocation with ETFs 

Portfolio & Risk Management 

Closed-End Funds 

Commodity ETFs