Excerpted from Oaktree Capital Management's Howard Marks letter to investors,
Risk Revisited Again...
What Risk Really Means
In [my] book, I argued against the purported identity between volatility and risk. Volatility is the academic’s choice for defining and measuring risk. I think this is the case largely because volatility is quantifiable and thus usable in the calculations and models of modern finance theory. In the book I called it “machinable,” and there is no substitute for the purposes of the calculations.
However, while volatility is quantifiable and machinable – and can be an indicator or symptom of riskiness and even a specific form of risk – I think it falls far short as “the” definition of investment risk. In thinking about risk, we want to identify the thing that investors worry about and thus demand compensation for bearing. I don’t think most investors fear volatility. In fact, I’ve never heard anyone say, “The prospective return isn’t high enough to warrant bearing all that volatility.” What they fear is the possibility of permanent loss.
Permanent loss is very different from volatility or fluctuation. A downward fluctuation – which by definition is temporary – doesn’t present a big problem if the investor is able to hold on and come out the other side. A permanent loss – from which there won’t be a rebound – can occur for either of two reasons: (a) an otherwise-temporary dip is locked in when the investor sells during a downswing – whether because of a loss of conviction; requirements stemming from his timeframe; financial exigency; or emotional pressures, or (b) the investment itself is unable to recover for fundamental reasons. We can ride out volatility, but we never get a chance to undo a permanent loss.
Of course, the problem with defining risk as the possibility of permanent loss is that it lacks the very thing volatility offers: quantifiability. The probability of loss is no more measurable than the probability of rain. It can be modeled, and it can be estimated (and by experts pretty well), but it cannot be known.
...
If you accept that the underlying processes affecting economics, business and market psychology are less than 100% dependable, as seems obvious, then it follows that the future isn’t knowable. In that case, risk can be nothing more than the subject of estimation – Keynes’s “intuition or direct judgment” – and certainly not reliably quantified.
* * *
The Unknowable Future
It seems most people in the prediction business think the future is knowable, and all they have to do is be among the ones who know it. Alternatively, they may understand (consciously or unconsciously) that it’s not knowable but believe they have to act as if it is in order to make a living as an economist or investment manager.
On the other hand, I’m solidly convinced the future isn’t knowable. I side with John Kenneth Galbraith who said, “We have two classes of forecasters: Those who don’t know – and those who don’t know they don’t know.”
...
Given the near-infinite number of factors that influence developments, the great deal of randomness present, and the weakness of the linkages, it’s my solid belief that future events cannot be predicted with any consistency. In particular, predictions of important divergences from trends and norms can’t be made with anything approaching the accuracy required for them to be helpful.
* * *
Coping with the Unknowable Future
Here’s the essential conundrum: investing requires us to decide how to position a portfolio for future developments, but the future isn’t knowable.
Taken to slightly greater detail:
- Investing requires the taking of positions that will be affected by future developments.
- The existence of negative possibilities surrounding those future developments presents risk.
- Intelligent investors pursue prospective returns that they think compensate them for bearing the risk of negative future developments.
- But future developments are unpredictable.
How can investors deal with the limitations on their ability to know the future? The answer lies in the fact that not being able to know the future doesn’t mean we can’t deal with it. It’s one thing to know what’s going to happen and something very different to have a feeling for the range of possible outcomes and the likelihood of each one happening. Saying we can’t do the former doesn’t mean we can’t do the latter.
The information we’re able to estimate – the list of events that might happen and how likely each one is – can be used to construct a probability distribution. Key point number one in this memo is that the future should be viewed not as a fixed outcome that’s destined to happen and capable of being predicted, but as a range of possibilities and, hopefully on the basis of insight into their respective likelihoods, as a probability distribution.
There’s little I believe in more than Albert Einstein’s observation: “Not everything that counts can be counted, and not everything that can be counted counts.” I’d rather have an order-of-magnitude approximation of risk from an expert than a precise figure from a highly educated statistician who knows less about the underlying investments. British philosopher and logician Carveth Read put it this way: “It is better to be vaguely right than exactly wrong.”
We can’t know what will happen. We can know something about the possible outcomes (and how likely they are). People who have more insight into these things than others are likely to make superior investors. As I said in the last paragraph of The Most Important Thing:
Only investors with unusual insight can regularly divine the probability distribution that governs future events and sense when the potential returns compensate for the risks that lurk in the distribution’s negative left-hand tail.
In other words, in order to achieve superior results, an investor must be able – with some regularity – to find asymmetries: instances when the upside potential exceeds the downside risk. That’s what successful investing is all about.
* * *
Thinking in Terms of Diverse Outcomes
It’s the indeterminate nature of future events that creates investment risk. It goes without saying that if we knew everything that was going to happen, there wouldn’t be any risk.
...
To oversimplify, investors in a given company may have an expectation that if A happens, that’ll make B happen, and if C and D also happen, then the result will be E. Factor A may be the pace at which a new product finds an audience. That will determine factor B, the growth of sales. If A is positive, B should be positive. Then if C (the cost of raw materials) is on target, earnings should grow as expected, and if D (investors’ valuation of the earnings) also meets expectations, the result should be a rising share price, giving us the return we seek (E).
We may have a sense for the probability distributions governing future developments, and thus a feeling for the likely outcome regarding each of developments A through E. The problem is that for each of these, there can be lots of outcomes other than the ones we consider most likely. The possibility of less-good outcomes is the source of risk. That leads me to key point number two, as expressed by Elroy Dimson, a professor at the London Business School: “Risk means more things can happen than will happen.” This brief, pithy sentence contains a great deal of wisdom.
People who rely heavily on forecasts seem to think there’s only one possibility, meaning risk can be eliminated if they just figure out which one it is. The rest of us know many possibilities exist today, and it’s not knowable which of them will occur. Further, things are subject to change, meaning there will be new possibilities tomorrow. This uncertainty as to which of the possibilities will occur is the source of risk in investing.
* * *
Even a Probability Distribution Isn’t Enough
I’ve stressed the importance of viewing the future as a probability distribution rather than a single predetermined outcome. It’s still essential to bear in mind key point number three: Knowing the probabilities doesn’t mean you know what’s going to happen. For example, all good backgammon players know the probabilities governing throws of two dice. They know there are 36 possible outcomes, and that six of them add up to the number seven (1-6, 2-5, 3-4, 4-3, 5-2 and 6-1). Thus the chance of throwing a seven on any toss is 6 in 36, or 16.7%. There’s absolutely no doubt about that. But even though we know the probability of each number, we’re far from knowing what number will come up on a given roll.
Backgammon players are usually quite happy to make a move that will enable them to win unless the opponent rolls twelve, since only one combination of the dice will produce it: 6-6. The probability of rolling twelve is thus only 1 in 36, or less than 3%. But twelve does come up from time to time, and the people it turns into losers end up complaining about having done the “right” thing but lost. As my friend Bruce Newberg says, “There’s a big difference between probability and outcome.” Unlikely things happen – and likely things fail to happen – all the time. Probabilities are likelihoods and very far from certainties.
It’s true with dice, and it’s true in investing . . . and not a bad start toward conveying the essence of risk. Think again about the quote above from Elroy Dimson: “Risk means more things can happen than will happen.” I find it particularly helpful to invert Dimson’s observation for key point number four: Even though many things can happen, only one will.
...
I always say I have no interest in being a skydiver who’s successful 95% of the time.
...
Investment performance (like life in general) is a lot like choosing a lottery winner by pulling one ticket from a bowlful. The process through which the winning ticket is chosen can be influenced by physical processes, and also by randomness. But it never amounts to anything but one ticket picked from among many. Superior investors have a better sense for what’s in the bowl, and thus for whether it’s worth buying a ticket in a lottery. But even they don’t know for sure which one will be chosen. Lesser investors have less of a sense for the probability distribution and for whether the likelihood of winning the prize compensates for the risk that the cost of the ticket will be lost.
* * *
Risk and Return
We hear it all the time: “Riskier investments produce higher returns” and “If you want to make more money, take more risk.”
Both of these formulations are terrible. In brief, if riskier investments could be counted on to produce higher returns, they wouldn’t be riskier. Misplaced reliance on the benefits of risk bearing has led investors to some very unpleasant surprises.
This is the essence of investment risk. Riskier investments are ones where the investor is less secure regarding the eventual outcome and faces the possibility of faring worse than those who stick to safer investments, and even of losing money.
* * *
The Many Forms of Risk
The possibility of permanent loss may be the main risk in investing, but it’s not the only risk. I can think of lots of other risks, many of which contribute to – or are components of – that main risk.
In the past, in addition to the risk of permanent loss, I’ve mentioned the risk of falling short. Some investors face return requirements in order to make necessary payouts, as in the case of pension funds, endowments and insurance companies. Others have more basic needs, like generating enough income to live on.
Some investors with needs – particularly those who live on their income, and especially in today’s low-return environment – face a serious conundrum. If they put their money into safe investments, their returns may be inadequate. But if they take on incremental risk in pursuit of a higher return, they face the possibility of a still-lower return, and perhaps of permanent diminution of their capital, rendering their subsequent income lower still. There’s no easy way to resolve this conundrum.
There are actually two possible causes of inadequate returns: (a) targeting a high return and being thwarted by negative events and (b) targeting a low return and achieving it. In other words, investors face not one but two major risks: the risk of losing money and the risk of missing opportunities. Either can be eliminated but not both. And leaning too far in order to avoid one can set you up to be victimized by the other.
Potential opportunity costs – the result of missing opportunities – usually aren’t taken as seriously as real potential losses. But they do deserve attention. Put another way, we have to consider the risk of not taking enough risk.
These days, the fear of losing money seems to have receded (since the crisis is all of six years in the past), and the fear of missing opportunities is riding high, given the paltry returns available on safe, mundane investments. Thus a new risk has arisen: FOMO risk, or the risk that comes from excessive fear of missing out. It’s important to worry about missing opportunities, since people who don’t can invest too conservatively. But when that worry becomes excessive, FOMO can drive an investor to do things he shouldn’t do and often doesn’t understand, just because others are doing them: if he doesn’t jump on the bandwagon, he may be left behind to live with envy.
...
There are many ways for an investment to be unsuccessful. The two main ones are fundamental risk (relating to how a company or asset performs in the real world) and valuation risk (relating to how the market prices that performance). For years investors, fiduciaries and rule-makers acted on the belief that it’s safe to buy high-quality assets and risky to buy low-quality assets. But between 1968 and 1973, many investors in the “Nifty Fifty” (the stocks of the fifty fastest-growing and best companies in America) lost 80-90% of their money. Attitudes have evolved since then, and today there’s less of an assumption that high quality prevents fundamental risk, and much less preoccupation with quality for its own sake.
On the other hand, investors are more sensitive to the pivotal role played by price. At bottom, the riskiest thing is overpaying for an asset (regardless of its quality), and the best way to reduce risk is by paying a price that’s irrationally low (ditto). A low price provides a “margin of safety,” and that’s what risk-controlled investing is all about. Valuation risk should be easily combatted, since it’s largely within the investor’s control. All you have to do is refuse to buy if the price is too high given the fundamentals. “Who wouldn’t do that?” you might ask. Just think about the people who bought into the tech bubble.
Fundamental risk and valuation risk bear on the risk of losing money in an individual security or asset, but that’s far from the whole story. Correlation is the essential additional piece of the puzzle. Correlation is the degree to which an asset’s price will move in sympathy with the movements of others. The higher the correlation among its components, all other things being equal, the less effective diversification a portfolio has, and the more exposed it is to untoward developments.
An asset doesn’t have “a correlation.” Rather, it has a different correlation with every other asset. A bond has a certain correlation with a stock. One stock has a certain correlation with another stock (and a different correlation with a third). Stocks of one type (such as emerging market, high-tech or large-cap) are likely to be highly correlated with others within their category, but they may be either high or low in correlation with those in other categories. Bottom line: it’s hard to estimate the riskiness of a given asset, but many times harder to estimate its correlation with all the other assets in a portfolio, and thus the impact on performance of adding it to the portfolio. This is a real art.
* * *
To move to the biggest of big pictures, I want to make a few over-arching comments about risk.
The first is that risk is counterintuitive.
- The riskiest thing in the world is the widespread belief that there’s no risk.
- Fear that the market is risky (and the prudent investor behavior that results) can render it quite safe.
- As an asset declines in price, making people view it as riskier, it becomes less risky (all else being equal).
- As an asset appreciates, causing people to think more highly of it, it becomes riskier.
- Holding only “safe” assets of one type can render a portfolio under-diversified and make it vulnerable to a single shock.
- Adding a few “risky” assets to a portfolio of safe assets can make it safer by increasing its diversification. Pointing this out was one of Professor William Sharpe’s great contributions.
The second is that risk aversion is the thing that keeps markets safe and sane.
- When investors are risk-conscious, they will demand generous risk premiums to compensate them for bearing risk. Thus the risk/return line will have a steep slope (the unit increase in prospective return per unit increase in perceived risk will be large) and the market should reward risk-bearing as theory asserts.
- But when people forget to be risk-conscious and fail to require compensation for bearing risk, they’ll make risky investments even if risk premiums are skimpy. The slope of the line will be gradual, and risk taking is likely to eventually be penalized, not rewarded.
- When risk aversion is running high, investors will perform extensive due diligence, make conservative assumptions, apply skepticism and deny capital to risky schemes.
- But when risk tolerance is widespread instead, these things will fall by the wayside and deals will be done that set the scene for subsequent losses.
Simply put, risk is low when risk aversion and risk consciousness are high, and high when they’re low.
The third is that risk is often hidden and thus deceptive. Loss occurs when risk – the possibility of loss – collides with negative events. Thus the riskiness of an investment becomes apparent only when it is tested in a negative environment. It can be risky but not show losses as long as the environment remains salutary. The fact that an investment is susceptible to a serious negative development that will occur only infrequently – what I call “the improbable disaster” – can make it appear safer than it really is. Thus after several years of a benign environment, a risky investment can easily pass for safe. That’s why Warren Buffett famously said, “. . . you only find out who’s swimming naked when the tide goes out.”
Assembling a portfolio that incorporates risk control as well as the potential for gains is a great accomplishment. But it’s a hidden accomplishment most of the time, since risk only turns into loss occasionally . . . when the tide goes out.
The fourth is that risk is multi-faceted and hard to deal with. In this memo I’ve mentioned 24 different forms of risk: the risk of losing money, the risk of falling short, the risk of missing opportunities, FOMO risk, credit risk, illiquidity risk, concentration risk, leverage risk, funding risk, manager risk, over-diversification risk, risk associated with volatility, basis risk, model risk, black swan risk, career risk, headline risk, event risk, fundamental risk, valuation risk, correlation risk, interest rate risk, purchasing power risk, and upside risk. And I’m sure I’ve omitted some. Many times these risks are overlapping, contrasting and hard to manage simultaneously. For example:
- Efforts to reduce the risk of losing money invariably increase the risk of missing out.
- Efforts to reduce fundamental risk by buying higher-quality assets often increase valuation risk, given that higher-quality assets often sell at elevated valuation metrics.
At bottom, it’s the inability to arrive at a single formula that simultaneously minimizes all the risks that makes investing the fascinating and challenging pursuit it is.
The fifth is that the task of managing risk shouldn’t be left to designated risk managers. I’m convinced outsiders to the fundamental investment process can’t know enough about the subject assets to make appropriate decisions regarding each one. All they can do is apply statistical models and norms. But those models may be the wrong ones for the underlying assets – or just plain faulty – and there’s little evidence that they add value. In particular, risk managers can try to estimate correlation and tell you how things will behave when combined in a portfolio. But they can fail to adequately anticipate the “fault lines” that run through portfolios. And anyway, as the old saying goes, “in times of crisis all correlations go to one” and everything collapses in unison.
“Value at Risk” was supposed to tell the banks how much they could lose on a very bad day. During the crisis, however, VaR was often shown to have understated the risk, since the assumptions hadn’t been harsh enough. Given the fact that risk managers are required at banks and de rigueur elsewhere, I think more money was spent on risk management in the early 2000s than in the rest of history combined . . . and yet we experienced the worst financial crisis in 80 years. Investors can calculate risk metrics like VaR and Sharpe ratios (we use them at Oaktree; they’re the best tools we have), but they shouldn’t put too much faith in them. The bottom line for me is that risk management should be the responsibility of every participant in the investment process, applying experience, judgment and knowledge of the underlying investments.
The sixth is that while risk should be dealt with constantly, investors are often tempted to do so only sporadically. Since risk only turns into loss when bad things happen, this can cause investors to apply risk control only when the future seems ominous. At other times they may opt to pile on risk in the expectation that good things lie ahead. But since we can’t predict the future, we never really know when risk control will be needed. Risk control is unnecessary in times when losses don’t occur, but that doesn’t mean it’s wrong to have it. The best analogy is to fire insurance: do you consider it a mistake to have paid the premium in a year in which your house didn’t burn down?
Taken together these six observations convince me that Charlie Munger’s trenchant comment on investing in general – “It’s not supposed to be easy. Anyone who finds it easy is stupid.” – is profoundly applicable to risk management. Effective risk management requires deep insight and a deft touch. It has to be based on a superior understanding of the probability distributions that will govern future events. Those who would achieve it have to have a good sense for what the crucial moving parts are, what will influence them, what outcomes are possible, and how likely each one is. Following on with Charlie’s idea, thinking risk control is easy is perhaps the greatest trap in investing, since excessive confidence that they have risk under control can make investors do very risky things.
Thus the key prerequisites for risk control also include humility, lack of hubris, and knowing what you don’t know. No one ever got into trouble for confessing a lack of prescience, being highly risk-conscious, and even investing scared. Risk control may restrain results during a rebound from crisis conditions or extreme under-valuations, when those who take the most risk generally make the most money. But it will also extend an investment career and increase the likelihood of long-term success. That’s why Oaktree was built on the belief that risk control is “the most important thing.”
Lastly while dealing in generalities, I want to point out that whereas risk control is indispensable, risk avoidance isn’t an appropriate goal. The reason is simple: risk avoidance usually goes hand-in-hand with return avoidance. While you shouldn’t expect to make money just for bearing risk, you also shouldn’t expect to make money without bearing risk.
* * *
At present I consider risk control more important than usual. To put it briefly:
- Today’s ultra-low interest rates have brought the prospective returns on money market instruments, Treasurys and high grade bonds to nearly zero.
- This has caused money to flood into riskier assets in search of higher returns.
- This, in turn, has caused some investors to drop their usual caution and engage in aggressive tactics.
- And this, finally, has caused standards in the capital markets to deteriorate, making it easy for issuers to place risky securities and – consequently – hard for investors to buy safe ones.
Warren Buffett put it best, and I regularly return to his statement on the subject:
. . . the less prudence with which others conduct their affairs, the greater the prudence with which we should conduct our own affairs.
While investor behavior hasn’t sunk to the depths seen just before the crisis (and, in my opinion, that contributed greatly to it), in many ways it has entered the zone of imprudence. To borrow a metaphor from Chuck Prince, Citigroup’s CEO from 2003 to 2007, anyone who’s totally unwilling to dance to today’s fast-paced music can find it challenging to put money to work.
It’s the job of investors to strike a proper balance between offense and defense, and between worrying about losing money and worrying about missing opportunity. Today I feel it’s important to pay more attention to loss prevention than to the pursuit of gain. For the last four years Oaktree’s mantra has been “move forward, but with caution.” At this time, in reiterating that mantra, I would increase the emphasis on those last three words: “but with caution.”
Economic and company fundamentals in the U.S. are fine today, and asset prices – while full – don’t seem to be at bubble levels. But when undemanding capital markets and a low level of risk aversion combine to encourage investors to engage in risky practices, something usually goes wrong eventually. Although I have no idea what could make the day of reckoning come sooner rather than later, I don’t think it’s too early to take today’s carefree market conditions into consideration. What I do know is that those conditions are creating a degree of risk for which there is no commensurate risk premium. We have to behave accordingly.
