An expert suggests a different take on investing: Allocate among, rather than within, asset classes to reduce volatility, increase return, and protect liquidity.
By Adam J. Smith
Prior to 2003, the investment portfolio of Grand Valley State University (GVSU), Allendale, Michigan—which had $60 million in assets as of June 30, 2010—relied on minimal external resources and local community members with some investing background, though not necessarily university endowment experience. The institution had five investment managers, each of whom was using five to nine funds to create portfolios of roughly 70 percent equities and 30 percent bonds. These portfolios were limited by GVSU to using a blend of U.S. and international equities spread among growth/value and large-cap/small-cap mandates. “We were consistently underperforming the average annual return measured by the NACUBO endowment studies. There was a conscious decision that we needed to do something totally different,” says Brian Van Doeselaar, the university's associate controller.
Seven years later, GVSU has revamped its advisory committee, which meets twice annually and now better understands asset allocation decisions and the risk-and-return profile of other commonly used asset classes. The university hired an investment adviser to help evaluate risk and mimic the performance of larger endowments by buying into fund-of-fund investments that it can better afford. Eleven new investment managers specialize in specific asset classes, including marketable and nonmarketable alternatives. As a result, GVSU now consistently outperforms the average returns as measured by the annual NACUBO studies, and institution leaders are better able to explain investment decisions to university stakeholders.
Some endowment managers, like those at GVSU, have begun reevaluating asset allocation practices once accepted as standard. Yet, how can an institution endowment mitigate volatility within its investment portfolio while also maintaining liquidity and without lowering returns? One way is by reexamining asset classes and evaluating each asset class in the context of the entire portfolio. In doing so, college and university endowments can develop a more efficient asset allocation strategy that reduces volatility, increases returns, and protects liquidity.
By Size and Style
Although most institutions agree on the importance of asset allocation, many differ on their implementation. According to the 2009 NACUBO-Commonfund Study of Endowments, the average asset allocation for educational endowments includes equities (both domestic and international), fixed income, alternative strategies, and short-term securities/cash. In addition, many institutions try to increase diversification by allocating to specific size or style mandates within the domestic equity allocation.
However, such a strategy is rarely effective because of relatively high correlations between various size and style segments as well as style drift, which can cause an overlap in holdings between different mandates and higher-than-expected correlations. Others conflate asset classes with investment vehicles, characteristics, or strategies, such as seeking to achieve diversification by treating an investment in hedge funds or other alternative strategies as a separate asset class. Neither method does an adequate job of limiting volatility.
There are two commonly held myths about how higher ed institutions can achieve diversification of their endowment portfolios. Here's an overview of these misconceptions, which might serve as a checklist against which institutions can reevaluate the true level of diversification within their portfolios.
Making alternative investments such as hedge funds. It's important to bear in mind that a hedge fund is an investment vehicle, not an asset class. Because hedge funds can invest in any asset class—including equities, fixed income, real estate, or commodities—saying that an institution's endowment has an allocation to hedge funds is really no different from saying it has an allocation to mutual funds. Therefore, an endowment portfolio that allocates to U.S. equities, international equities in developed markets, and investment-grade fixed income would not necessarily increase its diversification by adding an allocation to hedge funds.
Furthermore, an investment in hedge funds would almost certainly lead to lower liquidity because most hedge funds have lockup periods (often at least one year), offer limited withdrawal periods (e.g., quarterly with a 30-day notice), and have the ability to implement gates, which are restrictions that limit the amount of withdrawals during a redemption period.
Splitting U.S. equity allocations into specific size and style mandates. Smaller institutions might allocate to only two managers: large-cap core and small-cap core. Larger institutions might allocate to several managers: large-cap growth, large-cap value, small-cap growth, and small-cap value. The largest institutions might allocate to a manager within each of the nine size and style categories. Additionally, some institutions regardless of size may allocate to multiple managers within a single size and style category, such as utilizing multiple large-cap core managers. This strategy is based on the premise that correlations between growth and value stocks or large-cap and small-cap stocks are sufficiently low, so that allocating among these types of stocks will reduce the volatility of returns.
A simple correlation analysis proves this premise false because the various size and style segments of the U.S. equity market have relatively high correlations. From July 1, 1995, through July 31, 2010, the average correlation between various segments of the U.S. equity market was 0.70 or higher (see Figure 1). A high correlation coefficient (above 0.70) is indicative of poor diversification benefits in that the monthly returns for the various size and style segments within the U.S. equity market tend to move together. Therefore, splitting the U.S. equity allocation into specific size and style mandates does little to reduce the volatility of returns within the aggregate portfolio. In the end, size and style are merely characteristics used to describe various types of stocks within a single asset class—in this case, U.S. equities—rather than in separate asset classes.
Style drift is another potential problem associated with allocating to specific size and style mandates within the U.S. equity asset class. Style drift occurs when a U.S. equity investment manager hired to manage small-cap value equities begins to invest in stocks outside of its mandate, such as mid-cap or growth stocks. As a result of this style drift, the small-cap value mandate may begin to own some of the same stocks as other mandates in the portfolio, thereby leading to overlap with other U.S. equity investment managers. Overlap often results in different size and style portfolios owning some of the same securities, which can lead to unintended concentration in certain securities and even less diversification than originally intended. Style drift and overlap therefore can lead to even higher-than-expected correlations among U.S. equity managers, further reducing anticipated diversification benefits.
Among Asset Classes
Rather than trying to diversify within an asset class, college and university endowments should instead consider diversifying among them. For example, an institution could reduce its allocation to U.S. equities and diversify into other equity or equity-like asset classes. Other equity asset classes include international stocks in developed and emerging markets and real estate investment trusts (REITs). Equity-like asset classes include commodities and high-yield bonds. Although high-yield bonds are fixed income securities, their standard deviation tends to be more like equities than fixed income. Therefore, investors should consider including high-yield bonds as part of their allocation to equity-like volatility.
U.S. equities, as measured by the Russell 3000 Index, had an average correlation of 0.60 with the five other equity or equity-like asset classes during the period July 1, 1995, to July 31, 2010 (see Figure 2). Additionally, the other five equity or equity-like asset classes offer average correlations that are lower than that of the various size and style segments of the U.S. equity market. Therefore, shifting some of a portfolio's U.S. equity assets into other equity or equity-like asset classes could provide better diversification benefits, which may lead to lower volatility of returns for the portfolio.
A simple risk-and-return analysis of hypothetical portfolios constructed using broad-based indexes that represent liquid asset classes illustrates how diversification among asset classes can increase returns while lowering volatility. This analysis is based on a 60 percent allocation in asset classes with equity-like volatility and 40 percent allocation in those with fixed income-like volatility. For the period Oct. 1, 1997, to July 31, 2010, a basic portfolio that is comprised of 50 percent U.S. equities, 10 percent international equities in developed markets, and 40 percent investment-grade fixed income (Mix 1) would have generated an average annualized return of 4.9 percent with an annualized monthly standard deviation of 9.7 percent (see Figure 3).
For Mix 2, the U.S. equity allocation is reduced by 5 percentage points to 45 percent and a 5 percent allocation to commodities is added to the portfolio, which increases the return to 5.0 percent and reduces the standard deviation to 9.3 percent. Each mix (3 to 5) adds a 5 percent allocation to another asset class so that Mix 6 results in a portfolio with a 30 percent allocation to U.S. equity, a 35 percent allocation to investment-grade fixed income, a 10 percent allocation to international equities in developed markets, and a 5 percent allocation each to commodities, Treasury Inflation-Protected Securities (TIPS), REITs, international equities in emerging markets, and high-yield bonds. As a result, Mix 6 is the most efficient portfolio in the analysis in that it has the highest return and the lowest standard deviation of any portfolio.
Using an investment strategy based on allocations among asset classes may carry additional advantages. By conducting an asset allocation study in relation to an institution's liabilities (i.e., spending policy), investment stewards go beyond simply creating a less-volatile portfolio. They also create a base from which they can evaluate investment managers and better explain their decisions.
Like GVSU, Concord University, Athens, West Virginia—which had $21 million in assets as of June 30, 2010—recently changed its asset allocation strategies. The institution has switched from five asset classes to nine, adding commodities, REITs, and emerging markets to its portfolio after evaluating the risk and return of each class and the correlation of each with the five asset classes already included in the university's portfolio. Concord's investment policy has evolved from a vague directive of “manage the portfolio conservatively” to a structured investment policy statement that measures each investment manager's results against expected benchmarks for each asset class.
According to Greg Allen, immediate past chair of the Concord Foundation board of directors, these changes have allowed the university to better understand its endowment performance and communicate results to alumni and other interested groups. When an alumnus called earlier this year with concerns about recent investment decisions, Allen was able to tell him not only that the foundation was being professionally managed and watched closely, but also what each manager was doing and how each performed against a benchmark. “That soothed him,” Allen said. “If it had been three years ago, I couldn't have told him the return of our entire trust.”
Another benefit is the potential for greater liquidity. New investment vehicles such as exchange-traded funds and institutional mutual funds allow even small endowments to access emerging market stocks, high-yield bonds, and other asset classes. If it so desires, an institution that holds a mutual fund or an exchange-traded fund can liquidate its entire position and access all of its cash in one day for mutual funds and within three days for exchange-traded funds. As a result, within certain asset classes, institutional mutual funds and exchange-traded funds not only provide better access, but can also provide options that are more cost-effective and more liquid than a separate account.
Another benefit is the potential for greater liquidity. New investment vehicles such as exchange-traded funds and institutional mutual funds allow even small endowments to access emerging market stocks, high-yield bonds, and other asset classes.
Even with these advantages, the choice to expand out of traditional equity classes into equity-like classes can create a lack of understanding from auditors and others who pay attention to endowment performance. Van Doeselaar, who interacts regularly with GVSU auditors, says he is constantly challenged on the endowment's use of alternative strategies. Even so, “We're looking for results,” he says. “We're not going to back away because the auditors have issues [with these asset allocation strategies].”
Investment committees also may spend considerably more time vetting alternative investment options because of their lack of transparency. “We do find that we spend more time now doing due diligence with these types of investments,” says Brian Copeland, GVSU assistant vice president for business and finance. He adds that attending seminars and webinars and being able to show auditors a well-planned investment policy statement has gone a long way toward helping auditors understand decisions.
Newly added asset classes may also generate some initial trepidation among boards and finance committees. In today's environment, the words “emerging markets” bring the question “Why would you add it?” says Allen. “Yet, when added to a well-diversified portfolio, the discussion changes. By adding asset classes, we hope we can reduce our risk and possibly increase our return.”
Risk Worth Taking
Some level of risk is inherent with any investment strategy. Yet, institution endowments diversified among asset classes rather than by size and style segments of the U.S. equity market or investment vehicles such as hedge funds have an opportunity to achieve better returns through lower correlations and less volatility, while also maintaining liquidity and transparency. And, endowment managers could be in a much better position to communicate their asset allocation decisions to all institutional constituents and stakeholders.
ADAM J. SMITH is investment strategist for Lancaster Pollard Investment Advisory Group, Columbus, Ohio.