Is there a single best investment allocation strategy for the long-term investor? Some theories favor a one-portfolio-for-all investors approach, emphasizing a best-mix-of-assets program. The more traditional approach, which developed out of mean variance analysis some fifty years ago, tailors an individual's portfolio to his or her age (young investors should take more risk with stocks) and attitudes toward risk (conservative investors should hold more cash).
Research done by Harvard Business School finance professor Luis M. Viceira and his collaborators suggests that long-term portfolio choice is a complex issue, and that some conventional wisdom isn't always wise. For example, there are perfectly good reasons why a conservative long-term investor would hold more stocks and bonds than cash instruments such as certificates of deposit. These investors need to take into account their own spending needs as well as economic factors such as rising interest rates that affect different assets in different ways.
Viceira's research analyzes asset allocation strategies for personal and institutional investors. He teaches investment management and capital markets in Harvard Business School's MBA and doctoral programs. He discusses the findings and implications of his research with HBS Working Knowledge.
Ann Cullen: You point out that historically some assets might look riskier than they actually are and some might look safer than they actually are. Could you elaborate?
Luis M. Viceira: I think that the best example to understand this is to look at U.S. Treasury bonds. Over short horizons, bond prices fluctuate as interest rates change. Increases in interest rates cause bond prices to fall. This loss is largest for bonds of longest maturities. In the postwar period, the short-term price volatility of U.S. Treasury bonds has been about 5.5 percent per year in real terms—that is, after correcting for inflation. This volatility is certainly small when we compare it to the volatility of stock returns, which has been about three times larger, or 17 percent per year. But the average return on long-term bonds in excess of the return on cash has been also small when compared to the average excess return on stocks: It has been about 1.4 percent per year, while for stocks it has been 4.5 times larger, or about 6.3 percent per year. That is, the short-term return per unit of risk on bonds has been significantly smaller that the short-term return per risk on stocks. Thus short-term investors that value assets based on their expected return per unit of short-term volatility may find equities more attractive than bonds.
But investors seeking to finance their long-term spending needs might look at bonds and equities in a very different light: They may value the fact that, regardless of their short-term fluctuations in prices, bonds offer a safe stream of cash flows—their coupons, and their principal at the end of the life of the bond—while cash flows from stocks are not guaranteed. These investors might want to hold bonds in their portfolios for their ability to hedge their spending needs.
Similarly, empirical research by academics and practitioners alike suggest that the annualized volatility of stock returns declines with the investment horizon. This pattern is consistent with stock prices being subject to temporary fluctuations caused by changes in market risk aversion, which adds an additional layer of short-term volatility to stock returns. In recessions, even as long-term fundamentals remain the same, investors may be less willing to hold stocks in their portfolios, causing their prices to fall, while in expansions they may be more willing to hold stocks, causing their prices to rise. Thus stocks appear to be riskier at short horizons than they do at long horizons.
Q: Do you mean that stocks are riskless in the long run?
A: Not at all. I would like to emphasize that my previous statements about long-term stock return risk do not mean, as some have sometimes characterized it, that "stocks are riskless in the long run." This statement lacks empirical or logical support. Stock prices ultimately reflect the economic fundamentals of companies, which in turn are linked to the overall performance of the economy. There is an inherent uncertainty about the future prospects of these companies and the economy which does not go away at any investment horizon. Thus stocks are still risky, even at long horizons.
Q: Your comments suggest that bonds are safe investment for long-term investors, but this seems to contradict the conventional view that "cash is king" for highly conservative investors. What is the problem with this picture?
A: The problem with this picture is that cash (e.g., money market instruments, Treasury bills, and Certificates of Deposit) is riskless only at short horizons. When we buy a Treasury bill or a CD, we know in advance the return we are going to get, and we also know that we will recover our principal intact. But a long horizon investor will typically reinvest the principal every time the money instrument matures, and there is no [guarantee] that rates will stay the same over time. That is, short-dated fixed income instruments are subject to "reinvestment risk" (or, "interest rate risk"). Over long horizons, this risk can be considerable, and may have serious effects on the standard of living of a long-term investor.
Investors seeking to finance their long-term spending needs might look at bonds and equities in a very different light.
Perhaps the best way to see the relevance of reinvestment risk for investors is to look at the experience of many retirees in the U.S. in the last few years. About a year ago, the Wall Street Journal ran a story about how Fed rate cuts had forced many retirees in the U.S. to "pinch pennies" (WSJ, July 7, 2003). Many people who retired in the '80s and '90s followed the strategy of investing and reinvesting their savings in CDs and money market instruments, with the purpose of preserving the capital, and living off the interest on those deposits. This seemed a sound, safe strategy. Indeed, in the '80s and early '90s, when interest rates were reasonably high, this strategy worked rather well. But as interest rates, particularly short-term interest rates, fell dramatically all the way to 1 percent in June of 2003, this strategy backfired. The WSJ article told how many of these retirees were forced to either spend their principal to keep up their standard of living, or to dramatically cut their spending, or both.
A long-term view of investing does take into consideration reinvestment risk. This changes dramatically our perspective on what constitutes the "riskless asset" for a long-term investor. If an investor is interested in financing her spending needs with certainty over a long period of time, the safe strategy is to invest in long-term coupon bonds (or annuities), not cash.
There is catch with this, though. Standard annuities and Treasury bonds pay fixed coupons, and inflation can seriously erode the purchasing power of those coupons over time. Thus, on an inflation-adjusted basis, the riskless asset is not a nominal annuity or Treasury bond, but rather inflation-protected annuities and Treasury bonds (also known as TIPS). The demand for TIPS has grown considerably since they were created in 1997 when current Harvard University President Summers was Secretary of the Treasury. This demand has peaked recently, as inflationary pressures are building in the economy. Not surprisingly, institutional investors with long-term spending needs that grow with inflation (such as university endowments and foundations) are among the largest holders of these securities.
Q: Traditional academic models of asset allocation are often at odds with conventional wisdom on investing. According to your research, some of this conventional wisdom might actually make sense for long-term investors. How does static portfolio asset allocation contradict conventional wisdom, and how does the long-term view change this?
A: Traditional asset allocation models—I am thinking here of mean-variance models, which are very popular in the asset management industry—typically recommend that the mix of risky assets (e.g., the ratio of stocks to bonds) should be the same for all investors, regardless of how aggressive or conservative they are, and regardless of their investment horizon. Conservative investors should simply hold more cash than aggressive investors. For example, it might recommend that an aggressive portfolio of 60 percent equities, 30 percent bonds and 10 percent cash (a bond-stock ratio of 1:2), and a conservative portfolio of 20 percent equities, 10 percent bonds and 70 percent cash (a bond-stock ratio of still 1:2).
These recommendations are at odds with conventional wisdom, which has is that the mix of risky assets in a portfolio should depend on an investor's risk aversion and on the investor's horizon. Conventional wisdom might recommend an aggressive portfolio of 60 percent equities, 30 percent bonds and 10 percent cash (a bond-stock ratio of 1:2), and a conservative portfolio of 20 percent equities, 60 percent bonds and 20 percent cash (a bond-stock ratio of 3:1).
The recommendations of traditional asset allocation models derive mostly from the fact that these models usually adopt a short-term view of risk and return. This view is rather innocuous when annualized risk does not change across investment horizons, but otherwise it can lead to unsuitable investment advice for long-term investors. I have already argued that the recommendation that conservative investors should simply hold more cash might not be appropriate for long-term conservative investors, because reinvestment risk makes cash risky at long horizons. Thus a long-term perspective on conservative investing leads to recommend bonds, not cash, for long horizon investors. Similarly, under the view that the annualized volatility of stock returns declines with the investment horizon, aggressive long-term investors should hold on average more equities in their portfolios than aggressive short-term investors. These recommendations are closer to conventional wisdom on investing.
Q: There's speculation the Fed will continue to raise rates in the year ahead. What are the implications of interest rate hikes on your theories on long-term investing?
A: I have already argued that long-term bonds are the relevant asset for long-term investors. Thus these investors should care about long-term interest rates. The Fed has control over short-term interest rates, but it is less clear whether it can influence long-term interest rates. In fact, the spread between long-term and short-term interest rates (the slope of the yield curve) is currently quite large by historical standards. Some observes attribute this to the fact that the Fed policy of lowering short-term interest rates over the last few years has not fully translated into dramatically lower long-term interest rates. The spread of long-term interest rates over short-term interest rates is notoriously difficult to predict. Thus even as the Fed raises short-term interest rates over the next few months, we might not see long-term interest rates increasing in the same proportion. In fact, over the last couple of weeks we have seen a rally in the bond market, as previous increases in long-term rates have reversed on news that the economy is not growing as fast as we thought it would be, and that inflationary pressures might not be as large as we thought they were. Long-term rates depend mostly on expectations of long-term economic growth and inflation. The Fed can affect long-term rates only to the extent that it can influence those expectations.
Long-term conservative investors should not probably be in the business of speculating on the direction of long-term rates, which are very difficult to predict anyway, and instead focus on hedging their long-term spending needs.
Q: With the lowering of the bar to entry in hedge funds, many more people and institutions are investing in these funds. Given your research, do you think this is a good thing?
A: The investment management industry is going through a process of deep restructuring. Traditionally, indexing and active investment were mixed together in products such as mutual funds. The massive growth of cost-efficient indexing, derivative products and technological progress have made it possible now to separate active and passive money management. One way to look at hedge funds is as the pure active side of the money management industry. Their growth would then be explained as the other side of the growth in indexing: All part of the same process of separation of active and passive investment. From this perspective, hedge fund investing is all about "alpha," e.g., investment skill, which is not necessarily linked to "beta," or risk exposure. Through the use of derivative products such as swaps it is now possible for many investors to invest in "alpha" without compromising their risk exposures. My research is more oriented to determining the right risk exposures of long-term investors. Once they have decided upon them, they can then look for skill elsewhere.
There is no evidence that demographic changes have influence on stock returns and on liquid assets in general.
Q: Stock prices and other investment information are readily available today for free on the Internet. Do you think the pervasiveness of financial information on the Web is getting investors more interested in short-term investing?
A: I think more information is better than less information. I also think that it is wonderful that I can monitor my investments as often as I like, and that I can perform all kinds of financial transactions from my home computer at a fraction of the cost I would have had to pay twenty years ago. In this sense, these are good times for investors. Whether cheaper costs and easier access has made some people more prone to invest more frequently than they should is an open question. Even if this is the case, I think that for society this is a small price to pay for the enormous advantages that they have brought to all of us. The pension fund is also undergoing a deep structural change, with the responsibility of funding retirement increasingly in the hands of individuals. Thus we all have a need to improve our financial education, and fortunately it is now easier than before to do so.
Q: There's been much talk in the press that the first big wave of aging baby boomers will soon be retiring. In view of your theories, how will this effect the investment environment?
A: There is no evidence that demographic changes have influence on stock returns and on liquid assets in general. As long as the long-term prospects of the U.S. economy don't change as baby boomers age and retire, we might well find that younger investors and investors from all over the world are willing to buy the equities that baby boomers sell to finance their retirement. A different story could play if, for example, the need to finance their retirement leads to changes in taxes that reduce the expected after-tax returns on stocks or other assets. This would evidently have a negative effect on the prices of those assets.
The return on residential real estate might be affected by the aging of baby boomers though, if this leads to shifts in the demographic structure of different areas of the country. Housing prices are affected by local factors. Thus we might have declines in real estate values in places where the local population is aging and there is no younger population to replace them.