How to Think About Risk — Howard Marks

Framing the Subject

The emphasis of this lesson is on how to think about risk, not what to think. Risk, in Marks's view, is the ultimate test of an investor's skill: a return figure by itself says nothing about the quality of the manager's work. The essential follow-up question is always, how much risk was borne to earn that return?

Illustrating Skill Through Manager Profiles

Imagine a market that rises 10% or falls 10%, and consider several managers:

+10 / –10: Simply matches the market. No skill, no value added — an index fund would have done the same.
+20 / –20: No skill or discernment, merely aggressiveness.
+5 / –5: No selection ability, merely defensiveness — something an investor can achieve alone and shouldn't pay much for.
+15 / –10: Market-like losses with superior gains. This is genuine value added, what Marks calls asymmetry.
+10 / –5: Market-like gains in good times with reduced losses in bad times. Marks considers this the most interesting profile, and one that may characterize himself and Oaktree. Keeping up with the market when it does well is "good enough," since that's nearly all the time; the achievement lies in also declining less when the market falters.

What Risk Actually Is

Risk Is Not Volatility

Academics at the University of Chicago in the early 1960s adopted volatility as their measure of risk largely because it is quantifiable and nothing else is. Marks argues volatility may be a symptom or indicator of risk, but it is not risk itself.

Risk Is the Probability of Loss

Real-world investors demand compensation for the possibility of loss, not for volatility. Nobody at Oaktree declines an investment because it might fluctuate; they decline it because the chance of losing money is too high, or they demand a risk premium for that chance.

Risk Cannot Be Quantified

Like anything concerning the future, risk can only be a matter of opinion — it is not quantifiable in advance.
More surprisingly, risk is not quantifiable even after the fact. If you buy something for $1 and sell it for $2, you cannot tell from the outcome whether it was risky. It may have been a safe investment certain to double, or a risky one where you got lucky.

Risk Has Many Forms

Marks's memo Risk Revisited Again catalogues 24–25 varieties of risk. Two deserve special attention:

The risk of missing opportunities — that is, the risk of not taking enough risk.
The risk of being forced out at the bottom. Buying at a high and enduring a decline is survivable if you can hold on, because the next high typically exceeds the last. But selling at the bottom and missing the recovery derails you from investing entirely — Marks calls it the cardinal sin of investing.

The Philosophy of Uncertainty

Peter Bernstein on Walking Into the Unknown

Bernstein observed that risk means "we don't know what's going to happen; we walk every moment into the unknown." There is a range of outcomes; we don't know where within it the result will fall, and often we don't even know what the range is. Risk flows from this ignorance — if we knew the future, there would be no risk.

G.K. Chesterton on the Hidden Inexactitude of Life

Chesterton wrote that the trouble with the world is not that it is unreasonable, nor that it is reasonable, but that it is "nearly reasonable but not quite." Life is a trap for logicians: "its exactitude is obvious, but its inexactitude is hidden; its wildness lies in wait." We understand what is likely and what might plausibly happen instead, but we underappreciate the highly improbable — today's tail events. As Rick Kanine put it: 96% of financial history occurs within two standard deviations, but everything interesting happens outside them.

Four Principles for Thinking About Risk

"More things can happen than will happen" (Elroy Dimson, London Business School). Many outcomes are possible; only one occurs, and we don't know which.
The future is a range of possibilities, not a fixed, predictable outcome — ideally understood as a probability distribution across the likely, less likely, and unlikely-but-possible.
Knowing the probabilities does not tell you what will happen. In backgammon, the odds of each dice combination are exactly known — a seven comes up 6 times in 36 (16.7%), a six 5 times in 36, a two or twelve just once each (~3%) — yet uncertainty remains. Wharton's Chris Geczy expresses it as: "We live in the sample, not the universe." A former football player on Super Bowl morning 2016 captured this perfectly when asked about heavily favored Carolina versus Denver: "Carolina wins eight times out of ten. This could be one of the two." An 80% favorite still means the underdog wins one game in five, which is why the game must be played.
Expected value can be irrelevant. Since only one outcome occurs, the probability-weighted average may not even be among the possibilities — outcomes of 2, 4, 6, and 8, each equally likely, produce an expected value of 5, which cannot happen. Furthermore, a course of action with a higher expected value may embed outcomes you cannot live with, such as a remote risk of ruin, making a lower-expected-value alternative preferable.

The Character of Risk

Risk Is Counterintuitive

In Drachten, Netherlands, traffic lights, signs, and road markings were removed — and accidents and fatalities fell, because drivers responded by driving more carefully.
Avalanche expert Jill Fredston notes that better climbing gear arrives each year, yet fatalities do not decline, because people use the improved equipment to attempt riskier things.

The lesson: risk resides not only in the activity itself but in how participants approach it. In markets, risk is low when investors behave prudently and high when they don't.

Risk Is Perverse

The riskiest thing in the world is the belief that there is no risk; a high level of risk consciousness tends to mitigate risk.
As an asset falls in price, people call it risky — but the lower price actually makes it safer.
As an asset rises, people call it excellent — but the higher price makes it riskier.

This perversity is a primary reason most people fail to understand risk.

Risk Is Hidden and Deceptive

Loss occurs only when risk — the potential for loss — collides with negative events. As Buffett said, "only when the tide goes out do we find out who's been swimming naked." Marks's analogy: a California house may contain a construction flaw that causes no harm for years; only an earthquake tests it and converts latent risk into actual loss. An investment exposed only to rare "improbable disasters" — Nassim Taleb's black swans — can look safe for a long time. The infrequency of loss causes people to underrate the risk they're running.

Risk Is Not a Function of Asset Quality

Contrary to intuition, high quality does not equal safety:

The Nifty Fifty. When Marks began work in September 1969, banks invested in the 50 best, fastest-growing American companies, believing nothing bad could happen to them and no price was too high. An investor who bought that day and held tenaciously for five years lost more than 90% of their money. Roughly half those companies later hit serious fundamental trouble. Perceived quality itself created the overpricing that made them dangerous.
High yield bonds. After leaving equities in 1978, Marks was asked by Citibank to launch its high yield bond activity — investing in the lowest-quality public companies in America — and made money steadily and safely.

The lesson drawn from this juxtaposition: "It's not what you buy, it's what you pay." Investment success comes not from buying good things but from buying things well. No asset is so good it can't become overpriced and dangerous; very few are so bad they can't become cheap enough to be attractive.

The Relationship Between Risk and Return

The Flawed Classic Graph

The familiar Chicago School chart plots return on the vertical axis and risk on the horizontal, with an upward-sloping line implying that riskier assets produce higher returns and that taking more risk is the way to make more money. Marks rejects this: if riskier assets reliably produced higher returns, they wouldn't be riskier.

The correct reading is that investments perceived as riskier must be perceived as offering higher returns in order to attract capital — but they don't have to deliver. Risk arises precisely from the possibility that projected returns won't materialize. The straight line's linearity misleadingly implies dependability.

Marks's Revised Graph

Marks superimposes sideways bell-shaped probability distributions along the same line. As you move from left to right:

The expected return rises, as before.
But the range of possible outcomes widens, and the worst outcomes get worse.

That widening dispersion of bad possibilities is what risk really means.

Handling Risk

The Lottery Ticket Model

Investment outcomes are like pulling a single ticket from a bowl containing the full range of possible outcomes. Superior investors have a better sense of what tickets are in the bowl — what proportion are winners and losers — which lets them judge whether a given lottery is worth entering and how heavily to bet.

Assessment Must Be Subjective

Since risk cannot be measured, gauging it belongs to subject-matter experts using judgment. Marks is openly hostile to false quantification: imprecise qualitative expert opinion about the probability of loss is far more useful than precise but largely irrelevant numbers about past or projected volatility.

The Essence of Risk Management

Bernstein framed it best: because risk exists, things will sometimes differ from what we expect — how well are we prepared to cope when they do? There is no challenge in handling events that unfold as anticipated.

Marks refines the definition: while Bernstein's uncertainty encompasses good surprises as well as bad, risk should emphasize the bad. Risk is the possibility that, from the range of uncertain outcomes, an unfavorable one materializes. It can mean permanent loss of capital when bad things happen, or missing gains when good things happen — and these must be balanced. If an investment has a one-third chance of being down in a year, many would refuse it; but what about the two-thirds chance it rises? Real-world decisions cannot be made in one dimension.

Risk Control Must Be Continuous: Soccer, Not American Football

Marks bridles at the phrase "risk-on / risk-off market." Because we never know when bad things will strike, risk control must be applied constantly, not sporadically.

American football is therefore a poor model for markets: possession alternates, offense and defense take turns on the field (four downs to gain ten yards), personnel are swapped during stoppages, and each unit knows exactly when its turn has come. Real investing works the opposite way. As in the game the rest of the world calls football — soccer — the same players play the whole match, nobody announces when to attack or defend, and there are almost no stoppages in which to change tactics or personnel.

This matters because one of the investor's central decisions is when to be on offense, when to be on defense, and how much to allocate to each — yet no one signals the right moment, and no one pauses the game to give you time to think.

The Automobile Insurance Model of Risk Control

Marks's preferred analogy for risk management is car insurance. Nobody reaches the end of a year without an accident and regrets having been insured. We value insurance for the safety it provides, irrespective of whether a loss actually occurs in any given period. Risk control is worth paying for even in the years when it appears, in hindsight, to have been unnecessary.

Intelligent Bearing of Risk: The Life Insurance Analogy

In a 1981 cable television interview, Marks was asked how he could invest in high yield bonds knowing some issuers would go bankrupt. His answer: the most conservative companies in America are the life insurance companies — how can they insure people's lives knowing everyone will eventually die?

The resolution lies in four disciplines that life insurers apply, and that Oaktree applies to credit:

They take a risk they are aware of. An insurer is not shocked when a policyholder dies; mortality is expected, not a surprise.
They take a risk they can analyze. When Marks bought his first policy as a young man, the company sent a doctor to his home to assess his health.
They take a risk that can be diversified. No insurer covers only smokers, only skydivers, only residents of the San Andreas fault, or only the old or only the young. They build a mixed, diversified portfolio.
They take a risk they are well paid to bear. Using actuarial assumptions, they estimate expected payouts, allow a margin for uncertainty, and charge a premium accordingly.

Oaktree mirrors each step with credit risk: it is knowingly assumed, analyzed, diversified across large numbers of holdings that respond to different factors, and compensated through the risk premium or yield premium paid for bearing default risk.

Risk Control as the Hallmark of Superior Portfolios

Skilled investors assemble portfolios that will produce good returns if things go as expected and resist declines if they don't. This asymmetry is, in Marks's view, the cornerstone of superior investing.

Crucially, building risk control into a portfolio alongside upside potential is a hidden accomplishment. Because risk only turns into actual loss occasionally — when the tide goes out — the protection is invisible most of the time. Nevertheless, the prudent investor, and ideally his or her clients, knows the risk is being controlled even during the long stretches when that control never has to prove itself.

Manage Risk — Don't Avoid It

Risk control is indispensable; risk avoidance is not an appropriate goal.
Will Rogers: "You've got to go out on a limb sometimes, because that's where the fruit is."
Marks's experience observing others confirms that risk avoidance equates to return avoidance.
The intelligent bearing of risk should allow investors to earn good returns while keeping risk under control.

Concluding Principles

You shouldn't expect to make money without bearing risk.
You shouldn't expect to make money merely for bearing risk.
Risk is best handled through accurate subjective judgments made by experienced, expert investors who emphasize risk consciousness — not through false precision.
The great challenge of investing is to limit uncertainty while still preserving substantial potential for gain.

What Makes Outstanding Investors Outstanding

Exceptional investors are exceptional for a simple reason: they possess a superior sense for the probability distribution governing future events — the full set of "tickets in the bowl" — and for whether the potential return adequately compensates for the risks lurking in the distribution's unattractive left-hand tail.

That insight is precisely what enables the asymmetry that defines great investing: participating strongly in the gains when markets rise, while avoiding many of the losses when they fall.