Risk, Volatility and Drawdowns

Volatility: the wobble, quantified

When a factsheet says a world equity fund's volatility is ~15%, it means yearly returns have typically landed within ±15 percentage points of their average, and within ±30 in the rough years. Daily translation: moves of ±1% are unremarkable, ±2–3% days happen many times a year, and a handful of ±4–5% days arrive in stormy seasons. Bonds run at a quarter to a half of that; money markets at approximately zero.

Volatility is the statistic; two of its properties matter practically. It clusters: calm begets calm, storm begets storm, so turbulent weeks come in runs rather than singly, which is exactly when checking daily does its damage (Chapter 61). And it is symmetric in measurement but not in feeling: the +3% days finance the returns, the −3% days write the headlines, and a portfolio you only hear from on bad days will always sound worse than it is.

Drawdowns: the experience itself

Volatility is weather; a drawdown, the percentage fall from the previous peak, is the storm you personally stand in. The historical record for broad global/US equities, the honest table every investor should see before their first one (Dimson–Marsh–Staunton Global Investment Returns Yearbook and MSCI index data):

• Corrections (−10% to −20%): roughly every 1–2 years. Duration: weeks to months. Routine.
• Bear markets (−20% or worse): roughly every 5–7 years; average depth around −33%; typical peak-to-recovery spans of 2–4 years.
• The generational ones: 1973–74 (~−48% real, worse), 2000–02 (~−49% US, tech-heavy indices far worse), 2008–09 (~−55% global), 2020 (−34% in five weeks, recovered within months), 2022 (equities −25% with long bonds down double digits, the year the classic cushion failed).

Two features of this record matter more than the depths. Time underwater exceeds time falling: the 2000 and 2008 episodes each kept investors below their prior peak for years, and the psychological test is less the crash than the third year of flat. And recovery has been complete, so far, for broad diversified indices, which is precisely the property single stocks lack (Chapter 18): the index replaces its casualties; your concentrated position doesn't.

Risk vs volatility: the distinction that changes decisions

For a long-horizon investor, the deepest risk is not the wobble; it is permanent loss and permanent shortfall. Sort the threats accordingly:

• Volatility of a diversified portfolio: temporary by the historical record; converts into return over time; the thing you're paid for (Chapter 18).
• Permanent loss: concentrated positions going to zero, leveraged products decaying, selling a diversified portfolio at the bottom, fraud. Note that the last of the diversified investor's permanent losses is self-inflicted: the market offered a fluctuation, and the sale converted it.
• Shortfall risk: the quiet one, arriving via decades in cash "waiting for safety" (Chapter 17) or allocations too timid for the goal. Being unable to retire is also a risk outcome; it just lacks a scary chart.

This sorting is why the book's risk tools are what they are: diversification against permanent single-name loss, allocation sized to hold through fluctuations, and written pre-commitment against the self-inflicted category.

The deliverable: your worst case, in euros, in advance

Chapter 19 introduced the method; this chapter arms it with the record above. Take your intended mix, apply its historical worst drawdown to your projected portfolio at a meaningful size, and write the resulting euro figure into your one-page policy with the sentence that does the work: "When this happens, it is the plan happening, not the plan failing."

Concretely, at a €200,000 portfolio: 100/0 → worst case around −€110,000 on the 2008 template, potentially years underwater; 80/20 → roughly −€90,000; 60/40 → roughly −€65,000 (with 2022's asterisk that the bond cushion is usual, not guaranteed). Choose the largest number that reads, from your chair, as endurable weather. If none does, the answer is a calmer rung, not a cleverer fund.

A worked example

March 2020, compressed for study. Global equities fell 34% in five weeks; a €150,000 80/20 portfolio touched roughly €112,000.

• Investor A had the written page from Chapter 19: worst case −€45,000, sentence included. Reality (−€38,000) arrived inside the pre-approved envelope; she rebalanced on schedule, buying equities at the lows because the policy said so, and her only journal entry reads "as specified."
• Investor B held the identical portfolio with no page. Each red day was new information; at −30% the position "clearly wasn't working," he sold near the bottom, and re-entered eight months later, ~25% higher. Identical assets, identical storm; the difference in outcome, roughly a fifth of his portfolio, was manufactured entirely between his ears, in advance, by the absence of one page.

The market recovered in months, unusually fast; the lesson doesn't depend on that. It depends on the fact that A's decision was made in calm 2019 and B's in panicked March, and decisions inherit the quality of the moment that makes them.