Annual Research Review Part Two: Product and Strategy Implications

In this section of we will review a number of research studies with direct bearing on various products and strategies.

In their paper "Diversification Decisions of Individual Investors and Asset Prices", Goetzmann and Kumar find that "a vast majority of individual investors in our sample are underdiversified…More than 25% of investor portfolios in our sample contain only one stock, more than 50% of them contain fewer than three stocks, and in any given month only 5% to 10% of the portfolios contain more than ten stocks. As a consequence, investor portfolios have extremely high volatility (more than 75% of investor portfolios have higher volatility than the market portfolio) and they exhibit worse risk-return trade-off than randomly constructed portfolios." They also find that "the least diversified group of investors earn 2.4% lower return annually over the 1991-1996 period than the most diversified group on a risk-adjusted basis."

In "Downside Risk and Asset Pricing", Post and van Vliet starts with a simple question: in light of various research findings on the existence of non-market related risk factors (e.g., the size, value, and momentum effects), why do so many investors still put their money in market index funds? Using classical mean/variance optimization, this portfolio is less efficient than portfolios that include the other factors. To answer this question, the authors employ a criteria called second order stochastic dominance, or SSD for short. Without going into the gory statistical details, in determining the optimality of a portfolio, SSD takes into account the impact of skewness and kurtosis in addition to mean and variance. Using this criteria, they solve the mystery: the broad market index turns out to be superior, using SSD, than portfolios using size, value, and momentum that are optimal under the mean/variance criteria.

Past research has shown that individual investors place a heavy emphasis on past performance when deciding how to allocate their savings across different investment products. In "Inferences Regarding Investment Allocation Decisions in the Institutional Plan Sponsor Market" by Heisler, Knittel, Neumann and Stewart the authors find that a similar focus on past performance does not characterize the decisions made by institutional plan sponsors. They note that "the consistency with which investment managers deliver active returns over multiple horizon, without regard to the magnitude of those returns relative to the S&P 500 plays a key role in determining the flow of assets and accounts among investment products." They also find that "the magnitude of [any] one year loss, as well as 3 and 5 year total returns are incremental factors in plan sponsor's allocation decisions."

In the past, we have written about how the careful division of investments between taxable and tax exempt accounts can materially improve long term returns. In "Optimal Asset Location and Allocation with Taxable and Tax Deferred Investing", Dammon, Spatt, and Zhang reach the same conclusion (but note that they only use bonds and equities as asset classes). They conclude that it is better to hold taxable bonds in the tax-deferred account, and equity in the taxable account. However, "it may not be optimal to allocate the entire tax-deferred account to taxable bonds if doing so causes the overall portfolio to be over-weighted in bonds…In this case, investors may hold a mix of stocks and bonds in their tax-deferred account, but only if they hold an all-equity portfolio in their taxable account." Similarly, an investor with a higher allocation to bonds may want to hold some of them (ideally tax-exempt bonds) in their taxable account, assuming their tax-deferred account already holds only bonds.

While we're on the subject of bonds, Ang and Bekaert have produced a very interesting paper on "The Term Structure of Real Rates and Expected Inflation." They find that the unconditional term structure of real interest rates is quite flat, "starting at a rate of about 1.7% and increasing to just over 1.8% at one year, before declining again to 1.7% at the five year maturity." Also, in their paper "Are Treasury Inflation Protected Securities Really Tax Disadvantaged?" Hein and Mercer from the Federal Reserve Bank of Atlanta show that they are not, and in fact on an after-tax basis have outperformed matched maturity conventional (nominal return) Treasury securities.

Moving on to another asset class, Nijman and Swinkels have a very interesting paper titled "Strategic and Tactical Allocation to Commodities for Retirement Savings Schemes." They "find substantial differences in optimal strategic [commodities] allocations for pension plans with nominal and inflation-indexed liabilities. In the latter, commodities reduce the risk on the funding ratio by more than 30 percent."

In "Private Equity Performance", Kaplan and Schoar provide more information on the risks and returns of investing in venture capital and buyout funds. They being by noting the self-selection bias in the Venture Economics data series they use: roughly fifty percent of the funds reported raised do not provide performance data. However, using the performance data for the remaining funds that do report it, they find that "on average, LBO fund returns net of fees are lower than those on the S&P 500; VC fund returns are lower than the S&P 500 on an equal weighted basis, but higher than the S&P 500 on a capital weighted basis." They also "document substantial persistence in fund performance in the private equity industry, for both LBO and VC funds." They attribute this to these funds proprietary access to a flow of new investment opportunities, as well as to differences in their ability to add value to their respective investments. The authors also caution that "funds raised in boom times (and partnerships that are started during booms) are less likely to raise follow-on funds, indicating that these funds likely perform poorly."

Two other papers look at the choice between different types of hedge funds: "Fund of Hedge Funds Portfolio Selection" by Davies, Kat and Lu, and "Portfolios With Hedge Funds" by Chen, Feldman, and Goda. Both papers reach the same conclusion we did in our analysis of this asset class: Equity Market Neutral and Global Macro funds seem to provide the most benefit to portfolios that include a wide range of other asset classes.

Moving on to the analysis of specific investment products, in "Predictable Investment Horizons and Wealth Transfers Among Mutual Fund Shareholders", Woodrow Johnson makes three important points. First, the liquidity costs associated with the presence of both short and long term investors in a mutual fund can be expensive to the latter. He provides an estimate for one fund in which it amounts to .51% (51 basis points) per year in foregone returns. Second, there are factors that can be used to identify in advance mutual fund shareholders' likely holding period. Third, this argues strongly for either different funds for each group, or for the imposition of additional fees which will avoid the transfer of wealth caused by this liquidity effect from long term to short term holders.

Last but certainly not least, Elton, Gruber and Blake have written a very important paper on "The Adequacy of Investment Choices Offered by 401K Plans." After examining over 400 plans, they conclude that "for 62% of the plans, the types of choices offered to plan participants are inadequate, and that over a twenty year period this makes a difference in terminal wealth of over 300%." Stunning. The authors also make a very strong case for offering plan participants index funds that track a wider range of asset classes. As they note, "investors in 401K plans are sacrificing significant return [they estimate 3.16% per year] because plan administrators are offering an incomplete set of investment alternatives." The authors also find that "funds included in the plans are riskier than the general population of funds in the same category." Specifically, "plan administrators offer plan participants mutual funds with less variance than randomly selected funds, but funds that are more highly correlated." Moreover, when the authors "examine one category of investment choices, S&P 500 index funds, they find that the index funds chosen by 401K plan administrators are on average inferior to the S&P 500 index funds selected by the aggregate of all investors." In sum, this paper further reinforces our long-held opinion that in the world of defined contribution pension plans there is a great need for more "prudent experts."

| Global Asset Class Returns | Annual Research Review Part Two: Product and Strategy Implications | Equity Market Valuation Update | Model Portfolio Update | Annual Research Review Part One: Has the Death of Efficient Markets Theory Killed Indexing Too? | This Month's Letter to the Editor: Changes to one's portfolio - over time or all at once? |