Reading a PM’s Book: Exposure Dissection and a Pre-Committed
Drawdown Protocol for Discretionary Risk Oversight

In one line: Portfolios quietly hold bets nobody chose, so I hunt them nightly and cut risk by rules written beforehand.

Abstract. A portfolio is a set of claims: the manager attributes its returns to idiosyncratic judgment, while the risk seat must establish what the book actually owns. I describe two disciplines, both run nightly by this system on its own strategies. The first is an exposure dissection separating intended positioning from unintended exposure across six ordered layers (market, styles, industry and sector, macro sensitivities, crowding and liquidity, and the residual idiosyncratic share), decomposed by risk contribution rather than dollar weight and audited for factors concealed inside the residual. The second is a pre-committed drawdown protocol that fixes diagnosis, escalation ladder, and de-grossing order before the loss, so that a failing book is governed by rules set on a calm day. Both disciplines are applied to the system itself, which retired its founder’s namesake model under its own triggers.

1. Introduction

A portfolio is a set of claims. The manager claims the returns come from his ideas; the risk seat’s task is to establish what the book actually owns. These notes distill fifteen years of practice on the risk side of that conversation. They cover two disciplines, both practiced nightly by this system on its own strategies: a dissection of intended versus unintended exposure, and a protocol for the days on which the manager is losing money.

2. Exposure dissection: intended versus unintended

The organizing question is not whether the book is risky; risk is the mandate. The question is which exposures were chosen and which arrived uninvited. A fundamental stock picker is paid for idiosyncratic judgment; everything else in the book is either deliberate ballast or a stowaway. The dissection proceeds in six layers, cheapest explanation first.

2.1 The six layers

1. Market. Net and gross beta. The oldest stowaway is a “hedged” book whose longs are high-beta and shorts low-beta, net long the market while its net dollar exposure reads zero.
2. Styles. Size, value, momentum, quality, low-volatility, short-term reversal. The book’s returns are regressed on the factor set and the loadings checked against what the manager believes he runs.
3. Industry and sector. Concentrations, and the peer-relative view. A “diversified” book of one sector’s suppliers is a single bet expressed through many names.
4. Macro sensitivities. Rates, oil, the dollar, whatever the book’s names transmit. Duration is not only a bond concept.
5. Crowding and liquidity. Ownership overlap with other funds; days-to-liquidate at a reasonable participation rate. The position that cannot be exited determines the worst week.
6. The residual idiosyncratic share. After the preceding layers, what fraction of the book’s risk is genuinely idiosyncratic. A stock picker whose variance is mostly factor is a closet factor fund at stock-picker fees. The decomposition is run rolling rather than quarterly, because weights drift and books wander into factor bets that no one chose. Passivity is not factor-neutrality.

2.2 Two adjustments

Two adjustments come from this system’s own workshop. The first is the blame counter: dollar weights misstate where risk lives, so the book is decomposed by risk contribution, whose pieces sum exactly and whose ranking rarely matches the position sheet. The second follows from random-matrix work: even the residuals can conceal a common factor in a concentrated regime, so idiosyncratic is treated as a label to verify rather than a property to assume.

3. The drawdown protocol

The worst time to design a policy is inside the loss. The protocol exists so that the decisions are made on a calm day and merely executed on the worst one.

1. Diagnose before acting. Decompose the loss. Is it factor or idiosyncratic? A few names or the whole book? Inside the risk budget or breaching it? Is the manager losing on his thesis, or on exposures he did not know he held? The two have opposite remedies, and acting before answering treats both wrongly.
2. Run the pre-committed ladder. Early-warning indicator → the manager adjusts; first threshold → escalation; second threshold → mandatory de-risking. Mechanical de-grossing at drawdown levels is the exercise of nerve set in advance, not its failure. Triggers set at promotion are never renegotiated mid-drawdown; this system applies the same rule to its own strategies.
3. Cut in the right order. The de-grossing sequence is fixed in advance (Table 1): gross exposure is reduced first, unintended factor exposures next, concentrated and illiquid positions early, and the highest-conviction idiosyncratic names last.
4. Separate loss from error. Good process loses money on bad days; that is what risk means. The review asks whether the process was followed, not whether the outcome was pleasant. Repeated breaches, however, are structure rather than weather, and structure escalates.
5. Talk to the PM. Thesis intact, or broken by new information? The risk seat’s deepest task is keeping a skilled manager solvent long enough for the skill to compound, and distinguishing the defense of a position from the defense of a person.