Recognizing, Avoiding, and Controlling Risk
Three distinct skills, easy to conflate — and risk recognition is hardest exactly when calm markets make it feel least necessary.
Marks separates risk management into three distinct skills that are easy to conflate but genuinely different: risk recognition — being aware that meaningful risk exists in a given situation, especially when everything currently looks calm; risk avoidance — choosing not to take on risks that aren't adequately compensated; and risk control — limiting the damage from risks that are taken on anyway, through sizing and diversification.
He's specific that risk recognition is hardest exactly when it matters most — during extended calm, low-volatility periods, often late in a bull market per the pendulum metaphor covered earlier, investors as a group tend to systematically underestimate how much risk is actually present, precisely because recent experience hasn't punished that risk-taking yet.
Risk control, the third skill, acknowledges that risk avoidance alone isn't always possible or even desirable — some risk-taking is necessary to earn a return at all — which is why sizing positions appropriately and diversifying across genuinely independent risks is treated as a distinct, necessary skill rather than a fallback for investors who failed at avoidance.
These three skills build directly on the previous chapter's definition of risk as permanent-loss probability rather than volatility. Recognition asks whether that probability is meaningfully elevated in a given situation; avoidance asks whether the potential return adequately compensates for whatever probability exists; and control asks how to limit the damage on the specific occasions the probability actually materializes despite the first two skills having been applied correctly.
| Skill | What it actually involves |
|---|---|
| Recognition | Being aware meaningful risk exists, especially when things currently look calm |
| Avoidance | Choosing not to take on risks that aren't adequately compensated |
| Control | Limiting damage from risks taken on anyway — sizing, diversification |
During an extended stretch of low volatility and rising prices, the recent, lived experience of taking on more risk has simply been rewarded, repeatedly — which can make risk feel, to investors as a group, like it's genuinely diminished, even when the actual underlying risk may in fact be building rather than fading.
This connects directly to the pendulum metaphor from earlier in this course: a long calm stretch is, in Marks' framing, often exactly the period when the pendulum is swinging furthest toward greed, since sustained good outcomes are what erode caution in the first place. The paradox worth internalizing is that recent calm is weak evidence of low risk and can, in the specific conditions Marks describes, actually be a marker of risk quietly accumulating rather than receding.
A stretched valuation, or excessive leverage quietly building up in the system, can coexist with historically low measured volatility for a long stretch — recognizing the risk in exactly this kind of environment, when recent evidence seems to argue against it, is the hardest and most valuable version of this skill.
Even a genuinely well-analyzed, high-conviction position can turn out badly for reasons no analysis could have anticipated — which is why appropriate position sizing and real diversification across independent risks remain necessary even when an investor is highly confident in a specific view, rather than being treated as an admission of doubt about that view.
The word "independent" is doing real work in this discipline, and it's a common point of failure — an investor can hold what looks, on paper, like a diversified set of positions that all share a hidden common exposure, such that a single adverse event affects all of them simultaneously. Genuine diversification requires actively checking for these shared, sometimes non-obvious dependencies, not simply counting the number of different position names in a portfolio.
Imagine an investor evaluating an attractive-looking opportunity during a period of unusually calm, rising markets. Recognition means actively questioning whether the calm itself is masking building risk, rather than taking it at face value. Avoidance means concluding, after that scrutiny, that the compensation on offer doesn't adequately reward the risk actually present, and declining the specific opportunity. Control means that for the portion of the portfolio still exposed to the broader market's own risk regardless of that individual decision, position sizes and genuine diversification limit how much damage a systemic downturn could still do. All three skills operate on the same underlying situation, at three different stages of the decision.
- Risk recognition, avoidance, and control are three distinct skills, easily conflated but genuinely different in practice.
- Risk recognition is hardest during calm, low-volatility periods, precisely because recent experience seems to argue against caution — and calm itself can be a marker of the pendulum swinging toward greed.
- Avoidance alone isn't always possible or desirable, since earning any return requires taking on some risk.
- Genuine diversification requires checking for hidden, shared dependencies between positions — not simply counting the number of different names held.
- Position sizing and real diversification remain necessary even for high-conviction bets — they're not an admission of doubt, but a distinct, separate discipline.