by Howard Marks
The bottom line is that, looked at prospectively, much of risk is subjective, hidden and unquantifiable.
Where does that leave us? If the risk of loss can’t be measured, quantified or even observed—and if it’s consigned to subjectivity—how can it be dealt with? Skillful investors can get a sense for the risk present in a given situation. They make that judgment primarily based on (a) the stability and dependability of value and (b) the relationship between price and value. Other things will enter into their thinking, but most will be subsumed under these two.
There have been many efforts of late to make risk assessment more scientific. Financial institutions routinely employ quantitative “risk managers” separate from their asset management teams and have adopted computer models such as “value at risk” to measure the risk in a portfolio. But the results produced by these people and their tools will be no better than the inputs they rely on and the judgments they make about how to process the inputs. In my opinion, they’ll never be as good as the best investors’ subjective judgments.
Given the difficulty of quantifying the probability of loss, investors who want some objective measure of risk-adjusted return—and they are many—can only look to the so-called Sharpe ratio. This is the ratio of a portfolio’s excess return (its return above the “riskless rate,” or the rate on short-term Treasury bills) to the standard deviation of the return. This calculation seems serviceable for public market securities that trade and price often; there is some logic, and it truly is the best we have. While it says nothing explicitly about the likelihood of loss, there may be reason to believe that the prices of fundamentally riskier securities fluctuate more than those of safer ones, and thus that the Sharpe ratio has some relevance.
JOEL GREENBLATT: In a similar light, a Sortino ratio looks at only downside volatility rather than both upside and downside volatility. However, neither measure does a good job of measuring risk of future loss.
For private assets lacking market prices—like real estate and whole companies—there’s no alternative to subjective risk adjustment.
A few years ago, while considering the difficulty of measuring risk prospectively, I realized that because of its latent, nonquantitative and subjective nature, the risk of an investment—defined as the likelihood of loss—can’t be measured in retrospect any more than it can a priori.
Let’s say you make an investment that works out as expected. Does that mean it wasn’t risky? Maybe you buy something for $100 and sell it a year later for $200. Was it risky? Who knows? Perhaps it exposed you to great potential uncertainties that didn’t materialize. Thus, its real riskiness might have been high. Or let’s say the investment produces a loss. Does that mean it was risky? Or that it should have been perceived as risky at the time it was analyzed and entered into?
If you think about it, the response to these questions is simple: The fact that something—in this case, loss—happened doesn’t mean it was bound to happen, and the fact that something didn’t happen doesn’t mean it was unlikely.
Fooled by Randomness, by Nassim Nicholas Taleb, is the authority on this subject as far as I’m concerned, and in it he talks about the “alternative histories” that could have unfolded but didn’t. There’s more about this important book in chapter 16, but at the moment I am interested in how the idea of alternative histories relates to risk.
SETH KLARMAN: This is where top-notch management may also make a difference. Astute managers will be aware of the many risks that could threaten their businesses and will take action to mitigate or avoid them. Poor managers will fail to notice risks or fail to act, thereby subjecting their firms to avoidable loss.
In the investing world, one can live for years off one great coup or one extreme but eventually accurate forecast. But what’s proved by one success? When markets are booming, the best results often go to those who take the most risk. Were they smart to anticipate good times and bulk up on beta, or just congenitally aggressive types who were bailed out by events? Most simply put, how often in our business are people right for the wrong reason? These are the people Nassim Nicholas Taleb calls “lucky idiots,” and in the short run it’s certainly hard to tell them from skilled investors.
The point is that even after an investment has been closed out, it’s impossible to tell how much risk it entailed. Certainly the fact that an investment worked doesn’t mean it wasn’t risky, and vice versa. With regard to a successful investment, where do you look to learn whether the favorable outcome was inescapable or just one of a hundred possibilities (many of them unpleasant)? And ditto for a loser: how do we ascertain whether it was a reasonable but ill-fated venture, or just a wild stab that deserved to be punished?
Did the investor do a good job of assessing the risk entailed? That’s another good question that’s hard to answer. Need a model? Think of the weatherman. He says there’s a 70 percent chance of rain tomorrow. It rains; was he right or wrong? Or it doesn’t rain; was he right or wrong? It’s impossible to assess the accuracy of probability estimates other than 0 and 100 except over a very large number of trials.
“RISK,” JANUARY 19, 2006
And that brings me to the quotation from Elroy Dimson that led off this chapter: “Risk means more things can happen than will happen.” Now we move toward the metaphysical aspects of risk.
HOWARD MARKS: Understanding uncertainty: Dimson’s formulation reminds us of a very simple concept: that many things are possible in the future. We can’t know which of the possibilities will occur, and this uncertainty contributes to the challenge of investing. “Single-scenario” investors ignore this fact, oversimplify the task, and need fortuitous outcomes to produce good results.
Perhaps you recall the opening sentence of this chapter: Investing consists of exactly one thing: dealing with the future. Yet clearly it’s impossible to “know” anything about the future. If we’re farsighted we can have an idea of the range of future outcomes and their relative likelihood of occurring—that is, we can fashion a rough probability distribution. (On the other hand, if we’re not, we won’t know these things and it’ll be pure guesswork.)
HOWARD MARKS: Understanding uncertainty: The possibility of a variety of outcomes means we mustn’t think of the future in terms of a single result but rather as a range of possibilities. The best we can do is fashion a probability distribution that summarizes the possibilities and describes their relative likelihood. We must think about the full range, not just the ones that are most likely to materialize. Some of the greatest losses arise when investors ignore the improbable possibilities.
If we have a sense for the future, we’ll be able to say which outcome is most likely, what other outcomes also have a good chance of occurring, how broad the range of possible outcomes is and thus what the “expected result” is. The expected result is calculated by weighting each outcome by its probability of occurring; it’s a figure that says a lot—but not everything—about the likely future.
But even when we know the shape of the probability distribution, which outcome is most likely and what the expected result is—and even if our expectations are reasonably correct—we know about only likelihoods or tendencies. I’ve spent hours playing gin and backgammon with my good friend Bruce Newberg. Our time spent with cards and dice, where the odds are absolutely knowable, demonstrates the significant role played by randomness, and thus the vagary of probabilities. Bruce has put it admirably into words: “There’s a big difference between probability and outcome. Probable things fail to happen—and improbable things happen—all the time.” That’s one of the most important things you can know about investment risk.
JOEL GREENBLATT: When thinking about a portfolio of investments, one thing to keep in mind is that the correlation of these improbable occurrences can affect many of your investments at the same time.
While on the subject of probability distributions, I want to take a moment to make special mention of the normal distribution. Obviously investors are required to make judg
ments about future events. To do that, we settle on a central value around which we think events are likely to cluster. This may be the mean or expected value (the outcome that on average is expected to occur), the median (the outcome with half the possibilities above and half below) or the mode (the single most likely outcome). But to cope with the future it’s not sufficient to have a central expectation; we have to have a sense for the other possible outcomes and their likelihood. We need a distribution that describes all of the possibilities.
Most phenomena that cluster around a central value—for example, the heights of people—form the familiar bell-shaped curve, with the probability of a given observation peaking at the center and trailing off toward the extremes, or “tails.” There may be more men of five feet ten inches than any other height, somewhat fewer of five feet nine inches or five feet eleven inches, a lot less of five feet three inches or six feet five inches, and almost none of four feet eight inches or seven feet. Rather than enumerating the probability of each observation individually, standard distributions provide a convenient way to summarize the probabilities, such that a few statistics can tell you everything you have to know about the shape of things to come.
The most common bell-shaped distribution is called the “normal” distribution. However, people often use the terms bell-shaped and normal interchangeably, and they’re not the same. The former is a general type of distribution, while the latter is a specific bell-shaped distribution with very definite statistical properties. Failure to distinguish between the two doubtless made an important contribution to the recent credit crisis.
In the years leading up to the crisis, financial engineers, or “quants,” played a big part in creating and evaluating financial products such as derivatives and structured entities. In many cases they made the assumption that future events would be normally distributed. But the normal distribution assumes events in the distant tails will happen extremely infrequently, while the distribution of financial developments—shaped by humans, with their tendency to go to emotion-driven extremes of behavior—should probably be seen as having “fatter” tails. Thus, when widespread mortgage defaults began to occur, events thought to be unlikely befell mortgage-related vehicles on a regular basis. Investors in vehicles that had been constructed on the basis of normal distributions, without much allowance for “tail events” (some might borrow Nassim Nicholas Taleb’s term “black swans”), often saw the wheels come off.
Now that investing has become so reliant on higher math, we have to be on the lookout for occasions when people wrongly apply simplifying assumptions to a complex world. Quantification often lends excessive authority to statements that should be taken with a grain of salt. That creates significant potential for trouble.
Here’s the key to understanding risk: it’s largely a matter of opinion. It’s hard to be definitive about risk, even after the fact. You can see that one investor lost less than another in bad times and conclude that that investor took less risk. Or you can note that one investment declined more than another in a given environment and thus say it was riskier. Are these statements necessarily accurate?
For the most part, I think it’s fair to say that investment performance is what happens when a set of developments—geopolitical, macro-economic, company-level, technical and psychological—collide with an extant portfolio. Many futures are possible, to paraphrase Dimson, but only one future occurs. The future you get may be beneficial to your portfolio or harmful, and that may be attributable to your foresight, prudence or luck.
PAUL JOHNSON: Fully understanding this insight will be a major step toward understanding investment performance and has important philosophical ramifications for investing.
The performance of your portfolio under the one scenario that unfolds says nothing about how it would have fared under the many “alternative histories” that were possible.
• A portfolio can be set up to withstand 99 percent of all scenarios but succumb because it’s the remaining 1 percent that materializes. Based on the outcome, it may seem to have been risky, whereas the investor might have been quite cautious.
• Another portfolio may be structured so that it will do very well in half the scenarios and very poorly in the other half. But if the desired environment materializes and it prospers, onlookers can conclude that it was a low-risk portfolio.
• The success of a third portfolio can be entirely contingent on one oddball development, but if it occurs, wild aggression can be mistaken for conservatism and foresight.
Return alone—and especially return over short periods of time—says very little about the quality of investment decisions. Return has to be evaluated relative to the amount of risk taken to achieve it. And yet, risk cannot be measured. Certainly it cannot be gauged on the basis of what “everybody” says at a moment in time. Risk can be judged only by sophisticated, experienced second-level thinkers.
Here’s my wrap-up on understanding risk:
Investment risk is largely invisible before the fact—except perhaps to people with unusual insight—and even after an investment has been exited. For this reason, many of the great financial disasters we’ve seen have been failures to foresee and manage risk. There are several reasons for this.
• Risk exists only in the future, and it’s impossible to know for sure what the future holds. … No ambiguity is evident when we view the past. Only the things that happened, happened. But that definiteness doesn’t mean the process that creates outcomes is clear-cut and dependable. Many things could have happened in each case in the past, and the fact that only one did happen understates the variability that existed.
• Decisions whether or not to bear risk are made in contemplation of normal patterns recurring, and they do most of the time. But once in a while, something very different happens. … Occasionally, the improbable does occur.
• Projections tend to cluster around historic norms and call for only small changes. … The point is, people usually expect the future to be like the past and underestimate the potential for change.
• We hear a lot about “worst-case” projections, but they often turn out not to be negative enough. I tell my father’s story of the gambler who lost regularly. One day he heard about a race with only one horse in it, so he bet the rent money. Halfway around the track, the horse jumped over the fence and ran away. Invariably things can get worse than people expect. Maybe “worst-case” means “the worst we’ve seen in the past.” But that doesn’t mean things can’t be worse in the future. In 2007, many people’s worst-case assumptions were exceeded.
• Risk shows up lumpily. If we say “2 percent of mortgages default” each year, and even if that’s true when we look at a multiyear average, an unusual spate of defaults can occur at a point in time, sinking a structured finance vehicle. It’s invariably the case that some investors—especially those who employ high leverage—will fail to survive at those intervals.
• People overestimate their ability to gauge risk and understand mechanisms they’ve never before seen in operation. In theory, one thing that distinguishes humans from other species is that we can figure out that something’s dangerous without experiencing it. We don’t have to burn ourselves to know we shouldn’t sit on a hot stove. But in bullish times, people tend not to perform this function. Rather than recognize risk ahead, they tend to overestimate their ability to understand how new financial inventions will work.
• Finally and importantly, most people view risk taking primarily as a way to make money. Bearing higher risk generally produces higher returns. The market has to set things up to look like that’ll be the case; if it didn’t, people wouldn’t make risky investments. But it can’t always work that way, or else risky investments wouldn’t be risky. And when risk bearing doesn’t work, it really doesn’t work, and people are reminded what risk’s all about.
“NO DIFFERENT THIS TIME,” DECEMBER 17, 2007
HOWARD MARKS: Understanding uncertainty: Investing requires us to make deci
sions about the future. Usually we do so by assuming it will bear a resemblance to the past. But that’s far from saying outcomes will be distributed the same as always. Unusual and unlikely things can happen, and outcomes can occur in runs (and go to extremes) that are hard to predict. Underestimating uncertainty and its consequences is a big contributor to investor difficulty.
6
The Most Important Thing Is … Recognizing Risk
My belief is that because the system is now more stable, we’ll make it less stable through more leverage, more risk taking.
MYRON SCHOLES
The received wisdom is that risk increases in the recessions and falls in booms. In contrast, it may be more helpful to think of risk as increasing during upswings, as financial imbalances build up, and materializing in recessions.
ANDREW CROCKETT
No matter how good fundamentals may be, humans exercising their greed and propensity to err have the ability to screw things up.
Great investing requires both generating returns and controlling risk. And recognizing risk is an absolute prerequisite for controlling it.
Hopefully I’ve made clear what I think risk is (and isn’t). Risk means uncertainty about which outcome will occur and about the possibility of loss when the unfavorable ones do.
PAUL JOHNSON: This is the clearest and most accurate definition of risk I have heard. Marks nails it with this statement.
The next important step is to describe the process through which risk can be recognized for what it is.
Recognizing risk often starts with understanding when investors are paying it too little heed, being too optimistic and paying too much for a given asset as a result. High risk, in other words, comes primarily with high prices. Whether it be an individual security or other asset that is overrated and thus overpriced, or an entire market that’s been borne aloft by bullish sentiment and thus is sky-high, participating when prices are high rather than shying away is the main source of risk.