"Alternative" Investments: Balancing Risks and Rewards

An Address to the Endowment Management Seminar
Sponsored by The Common Fund and the Association of Governing Boards
David A. Salem
May 3, 1995

As best I can recall, my instructions for today were as follows: first, as the last outside speaker, I'm supposed to integrate the principles we've been discussing here today into a concrete framework for determining whether and when endowments should broaden their reach beyond domestic stocks and bonds; second, in doing so, I'm supposed to make sure that whatever I say will be thought-provoking if not also novel to people with lots of investment experience but nonetheless intelligible to people with limited investment experience; and third, I'm supposed to do all of this in no more than 30 minutes.

Talk about a no-win situation — why, this may be worse than the World Series of 1945, which some of you may recall was a similar exercise in futility. The wartime draft so weakened major league baseball that year that, when asked just before the start of Game One which team he thought would win the World Series, one baseball writer sneered, "I don't think either team can win it."

As you can see from the handout, my thesis is very simple, namely that the principles that underlie most successful attempts to diversify institutional funds beyond domestic stocks and bonds are in fact the same principles that animate successful investment programs that are confined to these two traditional asset classes. Some people argue otherwise — that to make a lot of money investing in alternative investments, you need to be foolhardy, or lucky, or both, but my own experience suggests otherwise.

Of course you need to take risks, but as we'll see the risks you need to take are at least as manageable as the risks you take investing in domestic stocks and bonds. As for luck, while it always helps to be lucky, you don't need to be lucky to do well investing in alternative investments, unless the type of luck you have in mind is the type of luck referred to in another favorite baseball story of mine — one involving the great pitcher Bob Gibson. When asked one day why Gibson was so successful, one of his teammates had this to say: "Bob Gibson is the luckiest pitcher I've ever seen — he always pitches on days when the other team gets no hits."

Having spent perhaps too much of my allotted time winding up, I'd better starting pitching, starting with Point #1 on your handout. To my way of thinking, perhaps the most important point to keep in mind as you address the all-important issue of asset allocation is that even the longest-term studies of the risk and return characteristics of individual asset classes are extremely sensitive to starting and ending dates. What's worse, studies extolling the virtues of particular investments have a nasty tendency to appear at, or close to, secular peaks in the returns on such investments. I still haven't figured out how the causality flows in such cases — whether the abnormally high returns attract researchers who coincidentally issue their reports just as these returns are about to turn downward, or whether the reports themselves induce so many investors to climb aboard a particular bandwagon that it quickly spins out of control. Whatever the causality, as I say in Point #2 on the handout, the tendency of so-called long-term studies of capital markets to serve as reliable contrary indicators is very striking.

Before I give some examples of this phenomenon, let me hasten to add that I think long-term risk and return studies are indeed useful in at least one important respect: as I say in Point #3, although we don't have nearly as much historical data on so-called alternative assets as we do on publicly traded stocks and bonds, we do have enough data to conclude without hesitation that these alternative asset classes are in fact just as cyclical as domestic stocks and bonds. Admittedly, some so-called alternative assets such as real estate don't display nearly as much short-term volatility as domestic stocks and bonds, but that's because their valuations are often based on appraisals rather than actual transactions, and appraisals tend to smooth out price fluctuations. If these assets were somehow marked to market everyday based on actual transactions, I suspect that they would display just as much volatility as stocks or bonds.

That's the bad news. As you can see in Point #4, the countervailing good news is that the investment cycles for many alternative assets don't necessarily coincide with the price cycles in domestic stocks and bonds. From a fiduciary standpoint, what this means is that by dialing in some exposure to assets that rise and fall at different times than domestic stocks and bonds, you can reduce the short-term volatility of your fund without reducing its expected return. Of course, whether you should in fact seek to minimize your fund's short-term volatility is a different question. My own view is that most institutional investors are overly concerned about short-term volatility. Why do I say this? Because if you analyze their asset mixes using what's known as mean-variance or efficient frontier analysis, you discover that these asset mixes imply an investment time horizon that is typically far shorter than the time horizon which results from even very conservative projections of the expected timing of cash withdrawals from these funds.

As most of you know, mean-variance analysis is a technique that helps investors determine which combinations of assets would historically have provided the highest returns for a given amount of volatility, typically measured over one-year time periods. Of course, you can use mean-variance analysis to identify portfolios that provide high returns for a given amount of longer-term variance, but when you lengthen the time horizon beyond the one year employed by most users of mean-variance analysis, you wind up with very concentrated portfolios. The computer basically tilts you in the direction of asset classes that have produced consistently high returns over whatever multi-year time horizon you specify, typically by performing unusually well in the closing years of the time series plugged into the model.

Run a mean-variance model at the end of the 1970s and it will put you heavily into real estate, energy-related investments, and even precious metals; run it in 1983 and it will put you heavily into venture capital and small stocks; run it in 1989 and it will put you heavily into Japanese stocks. Ironically, most fiduciaries that use mean-variance analysis lack the courage to implement the asset mixes recommended by their models, so they end up modifying the results so that they comport more closely with institutional norms. This tendency to monkey around with the inputs so that the output is more comfortable isn't as misguided as it might seem, however, because the results still tend to move the total fund in the direction of what the academics call a more efficient asset mix, by which they mean an asset mix that produces a higher return for whatever level of risk a governing board is comfortable assuming. So I don't want to leave the impression that I think so-called efficient frontier approaches to asset allocation are completely useless. On the other hand, I do think it's important that anyone using such approaches understand their limitations, including one relatively subtle but profoundly important defect that I've identified in Point #5 on your outline.

I suspect that many of you have seen variants of the data that appear on the slide before — statistical analyses which prove that by combining assets whose prices tend to rise and fall at different times you can boost your returns while at the same time reducing your risk. And you can — in theory. In practice, however, few if any investors are both able and willing to do what is needed to actually realize the returns that so-called efficient asset mixes theoretically produce. Why? Because one of the chief reasons so-called efficient asset mixes appear so attractive is because they assume that investors can and will rebalance their assets mixes every year back to the target weight for each asset class.

Unfortunately, this assumption conveniently ignores two ugly facts: first, few investors have the discipline to make the inherently contrarian rebalancing moves that efficient frontier approaches presuppose — to sell Asset A when it's up 50%, for example, and shift the proceeds into Asset B; second, even if investors do have the discipline needed to rebalance their assets in the manner that efficient frontier models presuppose, it can be extremely difficult if not impossible to do so because several of the asset classes that make so-called efficient mixes efficient are inherently illiquid — venture capital being perhaps the best example. The whole phenomenon reminds me of what Mark Twain once said, namely that "Statistics is the art of producing unreliable information from reliable figures."

Moving on, as I say in Point #6, diversifying into alternative investments in an effort to reduce short-term volatility is the wrong reason to extend your endowment's reach beyond domestic stocks and bonds. It's the wrong reason because, as I say in Point #8, careful study of most institutions' liquidity needs indicates that they ought to increase their exposure to what I call non-catastrophic risk, while simultaneously controlling carefully their exposure to catastrophic risk. It's in this context that the alternative assets that I was asked to discuss here today play not just a useful role, but arguably an essential one, for reasons that I'll outline in a minute. Before doing so, however, I need to define what I mean by catastrophic and non-catastrophic risk.

In the context of endowment investing, catastrophic risk is an event which, if not properly hedged against, can permanently impair an institution's capacity to achieve its mission. There are basically two forms of catastrophe that can strike an endowment: deflation, which is an economic environment in which the price of most goods and services actually falls, or the opposite of deflation, which is an environment in which prices spiral upward very rapidly. You don't need to invest in alternative assets of the sort we're discussing today to hedge against deflation — long-term, high quality, non-callable bonds are by far the best form of protection against a major deflation — but hard assets such as real estate, oil and gas reserves, and other commodities do provide a cost-efficient form of protection against high rates of unanticipated inflation, provided that the hedges are acquired at reasonable prices.

As with all forms of insurance, if you pay too much for certain coverage you can wind up with less money than when you started, even if the event that you were trying to hedge against occurs. A good example would be investors who bought into energy-related assets at the peak of energy prices in 1980. The price such investors had to pay to acquire energy-related assets in 1980 was so high that even if inflation had continued spiraling upward — which of course it didn't — the assets would not have provided meaningful protection against further declines in the purchasing power of their stocks and bonds.

Conversely, as I say in Point #11, catastrophe hedges are sometimes so mispriced that the "insurance" they provide has a so-called negative cost — meaning that you actually get paid to become insured. Again, the best example of this phenomenon comes from the early 1980s, when the classic deflation hedges — long-term, high quality, non-callable bonds — were so undervalued that people with courage enough to buy them were going to make a killing, even if we didn't have a deflation. A more recent example would be natural gas properties, which can provide good inflation-hedging protection but which were so undervalued in 1992 that the inflation protection they provided had a zero or arguably even negative cost.

To its great credit, The Common Fund had the courage to introduce its energy program during a period of significant weakness in energy prices, and although it's still far too early to draw any conclusions about the long-term success of this program I think it's increasingly clear that the timing of its launch was pretty good.

The same thing can be said with even greater force about The Common Fund's first major foray into foreign stock markets in the mid-1980s. Although most of the member schools that responded to The Common Fund's call to invest a portion of their endowments overseas did so to enhance their returns, not to hedge against catastrophic declines in domestic stock or bond prices, the timing of this call was superb. Coming as it did in the mid-1980s, The Common Fund's introduction of an international equity vehicle occurred at a time when foreign currencies were so cheap in relation to the U.S. dollar that U.S.-based endowments were going to do very nicely even if foreign stock markets hadn't appreciated much in local currency terms. In other words, far from being a risky or speculative move, investing in foreign stocks in the mid-80s was in my judgment a very prudent thing to do: it was like a coin flip where the payoff is, "Heads, you win ... tails, you don't lose."

Those are pretty good odds — in fact, they're precisely the odds that you should insist upon when you invest in alternative assets primarily to enhance your endowment's returns, as opposed to hedging against disasters such as deflation or hyperinflation. Now this is all very fine and well, you might say, but how does one go about finding investments that provide such favorable odds — where the upside is considerable if things go right, but the downside is very limited if they don't? Here's my answer: you find them by seeking out investments that entail a reasonably high level of what I call non-catastrophic risk. As you can see in the outline, far from avoiding this form of risk, I believe that long-term investors should willingly assume it. They should assume it because if they do so in a reasonably intelligent manner they'll be amply rewarded for taking such risks, by which I mean they'll help their institutions earn higher returns without exposing them to additional catastrophic risk. I don't have time to discuss multiple examples of how one can profit from willingly assuming these kinds of risk, but let me describe one type of alternative investment that, at least until recently, neatly illustrated each kind of risk and how they're interrelated.

The type of investment I have in mind are portfolios of busted loans purchased from banks that want to clean up their balance sheets. I won't take a poll, but I suspect no one in this room has ever bought a busted bank loan for their own account, at least not from their friendly neighborhood stock broker. The reason we haven't is because we can't, or at least we couldn't until quite recently, when the ever-inventive mutual fund industry began making packages of busted bank loans available to even individual investors. The reason you couldn't invest in busted bank loans through conventional channels until quite recently is because, unlike trading in conventional stocks or bonds, it's very difficult to standardize such transactions. Putting the point somewhat differently, at the time the first wave of institutional money started flowing into busted bank loans a few years ago, the so-called structural risks inherent in such investments was extremely high. In fact, it was so high that many institutional investors were then, and will likely remain, uninterested in owning busted bank loans at any price, no matter how high the expected returns are net of the concededly high costs of acquiring these assets.

Are busted bank loans inherently imprudent investments? It seems to me that the only logical answer is: it depends ... it depends on the price at which they can be acquired. If the price is low enough, they're anything but imprudent; if the price is high enough, of course they're imprudent, just as it now appears it was imprudent to invest a lot of money in IBM at the zenith of its popularity on Wall Street just a few years ago. Of course, now that much of the structural risk inherent in busted bank loans has been eliminated by enterprising investment bankers and mutual fund sponsors, the price one has to pay to acquire such loans has gone up, and hence the returns one might expect to earn have gone down. There's still lots of money to be made investing in busted bank loans, but as is true of every so-called alternative investment, as busted bank loans make their way from the fringes of institutional investing into the mainstream, investors intent on earning truly superior returns have to be increasingly selective.

Again, we don't have time to discuss specifically how each of the forms of non-catastrophic risk I've identified in the handout manifests itself in each of the asset classes that I foolishly agreed to discuss here today, but suffice it to say that the more prevalent these risks are in a given investment opportunity, the more attractive that opportunity is apt to be. As you can see, structural risk is arguably the most manageable form of risk that investors get paid to take, because you can usually reduce it, if not eliminate it altogether, with hard work.

As you can also see, there are at least three other forms of non-catastrophic risk that, on average and over time, investors are amply paid to bear. Interestingly, all of these risks are manifest to a greater or lesser extent in the non-traditional or alternative investment that arguably commands the most attention from institutional investors today: so-called emerging markets. Indeed, as was true of busted bank loans a few years ago, or venture capital in the 1970s, emerging markets display all of the forms of non-catastrophic risk identified in the outline: the bookkeeping can be a nightmare; information on individuals companies in lesser developed or developing countries is often imperfect at best; there are often huge imbalances between buyers and sellers, especially when the local investors who still dominate many of these markets are panicked by a political crisis; and the so-called reputational risk to U.S.-based fiduciaries of investing in these markets remains relatively high, although it's falling at a rapid enough rate that I am encouraging endowment trustees intent on wading into these waters for the first time to be extremely cautious.

Of the various forms of non-catastrophic risk identified in the outline, the most interesting by far in terms of investment opportunity is in fact reputational risk, by which I mean the fear of potential embarrassment to someone acting in a fiduciary capacity if an investment that they've approved performs poorly, especially immediately after it is acquired. As is almost always the case, when the reputational or career risk of investing in something is extraordinarily high, as it was for example with respect to natural gas stocks at the beginning of 1992 or with respect to gold stocks at the beginning of last year, so are the expected returns. The converse is also true, of course: when the reputational risk of investing in something is extraordinarily low, as it was for investors in IBM a few years ago, or for investors in Philip Morris as recently as two years ago, the future returns on that asset get pushed way, way down — often into negative territory.

Unfortunately for those of us who make a living managing money, even if an investor is able and willing to make investments entailing a high degree of reputational risk, the ever-increasing efficiency of U.S. stock and bond markets is making it harder and harder to outperform market averages. While fat pitches of the sort that natural gas stocks represented two years ago or that gold stocks represented at the beginning of last year occasionally come across the plate, with so many smart people and fast computers sifting through the data looking for pricing inefficiencies, fat pitches are increasingly rare in the domestic stock and bond markets, which is why I say in Point #17 that trustees of institutions seeking to earn abnormally high returns have no choice but to assume ever-increasing amounts of structural, or informational, or liquidity risk, or make bigger bets on those very infrequent occasions when the domestic stock or bond markets provide a chance to make lots of money by assuming a lot of reputational risk.

Regardless of which strategy your own institution pursues, I would encourage you to think very carefully about the risks inherent in every investment that you've made or are considering making, not to make sure that your downside is limited but rather to make sure that your upside is adequate. I know that sounds bass-ackwards — that the normal aim of due diligence work is to determine whether the risks inherent in an investment are sufficiently low, not sufficiently high — but with so much money flowing into so-called alternative investments, I personally think that a contrarian approach to risk management makes a lot of sense. If a so-called alternative investment doesn't display a sufficient measure of each of the risks we've discussed, it's probably not going to deliver the kinds of returns that caused you to consider alternative investments in the first place. On the other hand, if it is sufficiently risky, it's entirely possible that years from now people will accuse you of possessing the same type of luck that the great pitcher Bob Gibson possessed on all those days when the batters he faced got no hits.

I'm almost out of time, but before I hush up I want to take the opportunity of appearing before this very distinguished audience to share with you a concern I have about the institutions that you represent. I realize that I'm not going to make any friends in doing so, but I am so intrigued and, I should emphasize, depressed by what I'm about to describe that I feel compelled to share it with you.

As many of you know, one of the most significant developments now occurring in the money management business is the increasing popularity of so-called long/short or market neutral strategies. As per usual, The Common Fund was an early proponent of such strategies, which are premised on the unarguable fact that the structural and behavioral impediments to short selling create an advantage for active managers who are able to both purchase undervalued issues and sell short overvalued ones.

Now, what does the increasing popularity of market neutral or short selling techniques have to do with the outlook for higher education, beyond the fact that some colleges can and likely will make a lot of money using such techniques, especially if they delegate responsibility for choosing money managers to The Common Fund? I think there is indeed a connection between short selling techniques and the outlook for higher education, or at least I thought so when this idea first popped into my mind, which it did when I and my own board were interviewing long/short or market neutral managers for the common fund for foundations that I head. As I sat there listening to one especially able manager describe what he looks for when picking stocks to sell short, it dawned on me that higher education as it is currently configured is theoretically one of the greatest short selling opportunities of all time. Why is higher education an ideal short sale? Because when you examine it from an investor's perspective, which is to say in an utterly objective if not ruthless manner, it displays all three earmarks of a classic short sale: first, there's a large and growing gap between the true costs of the product, taking into account all relevant subsidies, and the price that the ultimate consumer of the product is able to pay for it; second, as an industry, higher education is on the receiving end of one of the most harmful regulatory changes that has ever hit any industry, namely the forced elimination of mandatory retirement for tenured faculty; third, and most importantly, there is, or I should say there soon will be, an incredible amount of excess capacity in the industry.

Why do I say this? Because, as most truly objective observers would agree, technology has finally evolved to the point where the age-old limits on the number of students that a professor can effectively instruct have been removed, and there's absolutely no reason why we as a society should continue to employ thousands of Shakespeare scholars or thousands of professors in other disciplines when only a fraction of them are truly superior teachers, and only an even smaller number publish anything of lasting merit during their careers. As most of you in this room know only too well, top professors are already like pro athletes in the sense that they continually sell their talents to the highest bidder. This problem, if it can be called that, will only become more acute as it becomes more economic for students to assemble their own custom-designed degree programs from courses offered around the country, all from the comfort of their own homes.

Admittedly, we will still need physical campuses in order to transmit certain knowledge — laboratories for the physical sciences, studios for the visual and performing arts — but there's no physical impediment to transmitting other forms of knowledge via digital networks, and as much as we would all prefer to maintain the status quo of ivy-clad buildings and tailgate weekends, I doubt that we as a society will be able to afford it much longer, if indeed we're not fooling ourselves in thinking we can afford it today. As someone who benefited greatly from a four-year residential college experience, I would be the last one to argue that the modem-based education I've just described will prove superior to the experience most of us enjoyed between the ages of 18 and 22, but try as I might I don't see how we as a society will generate the wealth to pay for such an experience for more than a small fraction of those seeking a college education.

The problem for an investor who shares this rather pessimistic view of higher education, of course, is that there are no obvious assets to sell short that are likely to plummet in price as the excess capacity I've just described is wrung out of the system. The only opportunity that comes to mind is to sell short, or at least not own, assets that are likely to decline in value as a result of the same technological changes that'll make it possible for most students to get a college education without spending much if any time on a college campus. What assets fit the bill? The assets that come immediately to mind are high rise office buildings in major urban centers, which were sensible places to conduct business in the days of carbon paper and hierarchical organizational structures but which are increasingly obsolete in a world of digital networks and decentralized decision-making. A lot of folks have been gleefully snapping up high rise office buildings at what think they are attractive prices, but I fear that what appears to have been the end of a secular decline in the price of such assets was merely a cyclical low, to be followed by further price declines as the so-called information superhighway begins carrying more and more passengers.

In closing, let me emphasize that I have not said the things I've just said about higher education because I don't value higher education or don't care about it. Rather, I've said them because I care about higher education very deeply; at the same time, I'm powerfully impressed — perhaps overly so — by the fact that colleges and universities are more insulated than are most for-profit enterprises from the disciplinary power of the capital markets, and this makes it much harder for even well-intended administrators and trustees to make the truly hard choices that need to be made to ensure their institutions' long-term viability.

I'll stop right there. Thank you very much for your attention.

Return to top of page