Why Drawdowns Are Portfolio Events, Not Market Events
The market supplies the decline; the portfolio supplies the loss. What a drawdown costs you is decided by your structure — leverage, obligations, liquidity, behaviour — not by the index chart.
Investors study drawdown history the way meteorologists study storms — how deep past declines went, how long recovery took — as if the market's statistics were a forecast of their own experience. The premise is wrong. Two investors can hold the same assets through the same decline and emerge with opposite outcomes: one fully recovered, the other permanently impaired. Nothing about the market differed between them. Everything about the portfolio did — which means the drawdown, as an event that costs money, happened at the portfolio level, not the market level.
The framing we use: the market supplies the decline; the portfolio supplies the loss. A falling price is, in itself, a change in a quote — a reversible mark-to-market event on claims you still own. It becomes a permanent loss only at a moment of conversion: the moment shares are surrendered at depressed prices, whether because leverage demanded it, an obligation fell due, a fund's structure imposed it, or fear chose it. Those four converters — leverage, obligations, liquidity mismatch, and behaviour — are all properties of the portfolio and its owner. The market controls the timing and depth of declines, which is why forecasting them has such a poor record. The portfolio controls the conversion terms, which is why engineering them has such a good one. Risk management, on this framing, is not the art of predicting the storm; it is the removal of the mechanisms that convert weather into wreckage.
The arithmetic that makes conversion expensive
Before the mechanism, the arithmetic — because the cost of converting at the bottom is not intuitive, it is geometric. Percentage losses and gains are asymmetric: a portfolio that falls by half must double to recover, and the required recovery grows faster than the loss that caused it. This asymmetry is harmless to an investor who neither adds nor removes capital — for them the arithmetic runs in reverse on the way back up, and the round trip nets to the market's round trip. The asymmetry becomes destructive the moment capital exits during the trough. Every unit withdrawn at depressed prices is a unit that misses the recovery leg, permanently. This is the sequence-of-returns problem: for a portfolio being drawn upon, the ordering of returns matters as much as their average, because withdrawals interact with the path. Two portfolios with identical average returns over a long horizon can end in wealth positions that differ by multiples, purely on whether the deep decline arrived early or late in the withdrawal schedule. The market's return sequence is not something anyone chooses. The withdrawal schedule's collision with it is.
Three layers of why
The illustration: one decline, two populations
The 2008–2009 episode, read as history, is the cleanest natural experiment because the market's own story had a definitive ending: broad equity indexes fell catastrophically and subsequently recovered to new highs within a few years. For a hypothetical holder with no leverage, no forced obligations, and no capitulation, the entire episode — for all its violence — netted out to a temporary mark. Yet the episode permanently impaired an enormous population of investors, and the mechanisms are instructive because none of them was 'the market'. Leveraged holders were carried out at the lows by margin calls they could not meet. Investors whose incomes and businesses were hit by the same recession were forced to liquidate portfolios to fund life, at the worst prices of the cycle. Holders of vehicles with liquidity mismatches met gates and suspensions, then sold what remained liquid at distressed prices. And a large cohort simply sold near the bottom — not forced by anything but the overwhelming plausibility of the pessimistic case at the moment it was most fully priced in.
The 2020 decline compressed the same lesson into weeks: an extremely violent fall followed by an unusually fast recovery, which meant the entire cost of the episode was borne almost exclusively by those who converted — the leveraged, the obligated, and the frightened — while the unforced holder's experience rounded to volatility. Two very different market events; the identical portfolio-level sorting. The market chose the depth and the timetable. The portfolio's structure chose who paid.
The strongest case against this framework
The serious objection is that this framework smuggles in an assumption of recovery, and recovery is not a law. Individual companies go to zero and stay there. Entire national markets have spent decades below prior peaks — Japanese equities after 1989 being the canonical case. An investor with a genuinely short horizon experiences mark-to-market as real risk regardless of any eventual recovery they will not be present for. On this view, 'a decline only costs you if you sell' is survivorship-flavoured comfort: sometimes the decline is the permanent event, and refusing to sell simply rides the impairment all the way down.
We accept every example and note that each one argues for the framework, not against it. Single names going to zero is the case for breadth: what has historically recovered from deep declines is the diversified claim on aggregate corporate earnings, not any individual ticket in it — so breadth is precisely the portfolio decision that makes 'temporary' a defensible working assumption. A decades-long national stagnation is the case against geographic concentration, and it is worth noting that the canonical example was also a case of extreme valuation concentration that a globally diversified holder wore as a heavy but survivable position — the lesson lands on portfolio construction again. And the short-horizon objection is the strongest of the three, which is why its answer is the framework's central rule rather than an exception to it: capital with a short horizon is an obligation, and obligations are exactly what should never be funded with long-duration risk assets in the first place. In every case, the objection identifies a real permanent-loss channel — and in every case the channel runs through a portfolio decision: concentration, geography, or liability matching. That is the claim.
What this changes about how you prepare
The practical output is an inversion of where preparation effort goes. Most drawdown preparation is spent on the market side — estimating probabilities, watching indicators, considering hedges — where the portfolio owner has no control and forecasting has a poor record. This framework moves the effort to the conversion side, where control is total. Sterling's Embedded Intelligence runs drawdown analysis in exactly these terms: not 'how far could this portfolio fall' as a headline number, but an audit of converters — what leverage exists and on what terms it can be called; which obligations fall due inside the horizon and what asset funds each one; which holdings have liquidity that degrades in stress; and what the owner's own behavioural record in past declines suggests about the fourth converter. Sterling's summary metric is deliberately unglamorous: how many years of obligations can this portfolio meet without selling a risk asset at all? That number — not the projected percentage decline — is what determines whether the next drawdown will be an event in your portfolio or merely an event in the market.
How to apply this framework
- Audit the converters, not the forecasts. List every mechanism that could force a sale at depressed prices — callable leverage, obligations due inside the horizon, liquidity mismatches, and your own behavioural record. Each one removed in calm markets is a permanent-loss channel closed before it can open.
- Denominate your drawdown budget in obligations covered, not percentage decline. The question 'can I tolerate a forty percent fall?' is unanswerable in the abstract. The question 'how many years of spending can I meet without selling a risk asset?' has a number, and that number is the portfolio's true drawdown capacity.
- Match every liability to an asset that cannot be marked down when it is due. Short-horizon capital funded with equities is a conversion event waiting for its trigger. Fund known obligations from cash and short duration, and the remaining portfolio earns the right to treat market declines as repricings rather than emergencies.
- Write the behaviour policy before the decline, because tolerance is state-dependent. Decide in calm what you will do — and specifically what you will not do — at defined decline thresholds, and treat the document as binding. The bottom will arrive with a persuasive narrative attached; the policy exists because you already knew that.