Autopsies Without Blame: Building a Post-Mortem Discipline for Retired Strategies

Somewhere in nearly every institutional portfolio sits a strategy that was terminated last quarter. The proceeds were reallocated, the manager was paid, the custodial accounts were closed. The file went dark. And in the overwhelming majority of cases, no one sat down to ask the only question that gives a closure meaning beyond its arithmetic: what did we learn?

This is not negligence in the ordinary sense. The professionals involved acted in good faith, followed their mandates, and moved on to the next problem. The failure is structural, not personal. American institutional investing — endowments, foundations, corporate plans, insurance general accounts — has built elaborate machinery for capital deployment and almost none for capital reflection. We audit what we hold. We almost never audit what we released.

The Asymmetry of Institutional Memory

Consider what happens when a strategy is launched. There is an investment memo, a diligence binder, a signed investment policy statement amendment, a board presentation. The decision is documented at its moment of maximum optimism, when every argument for proceeding has been sharpened for scrutiny.

Now consider what happens when that same strategy is terminated. A brief memo, perhaps. A line item in the meeting minutes. Occasionally nothing at all beyond an email chain. The asymmetry is striking: we preserve the reasoning behind our convictions but discard the reasoning behind our reversals. Yet the reversal is almost always the more instructive document.

The problem compounds over time. Staff turnover in the institutional space is notoriously fluid — CIOs move between funds, consultants rotate accounts, analysts take their context to buy-side seats elsewhere. Within three to five years, the institutional memory of why a strategy died is typically gone. What remains is folklore: a simplified, often inaccurate narrative that hardens into orthodoxy. "We tried that in 2019 and it didn't work" becomes a permanent veto, unmoored from the actual conditions of the original decision.

Why Hindsight Goes Unharvested

Three forces conspire against the post-mortem.

First, incentive misalignment. Nobody's compensation or reputation is advanced by revisiting a closure. The team that recommended the strategy has moved on emotionally; the team that terminated it wants to look forward. A post-mortem offers career downside and no career upside — a formula that guarantees voluntary silence.

Second, the sunk-cost psychology of closure itself. Terminating a strategy is often emotionally taxing, particularly when it was championed by a respected colleague or a longtime manager. The institutional instinct after a hard decision is relief, not interrogation. Scheduling a dissection of the decision for six weeks later feels like reopening a wound.

Third, and most subtle: hindsight bias makes post-mortems feel pointless to the untrained. Once the outcome is known, it seems obvious in retrospect. "Of course the strategy underperformed — rates rose." This illusion of inevitability is precisely why the exercise matters. The goal is not to confirm what happened but to recover what was knowable at the time, and to distinguish between a good decision with a bad outcome and a bad decision that got lucky.

A Framework for the Autopsy

We propose a simple discipline: every strategy termination triggers a structured post-mortem within ninety days, conducted by someone who was not the decision's principal advocate. The protocol has four components.

Reconstruct the thesis. Pull the original memo. Restate, in the decision-makers' own words, what the strategy was supposed to do, under what market conditions, over what horizon. This is the control document. Without it, the review drifts toward judging outcomes against expectations that were never actually held.

Separate process from outcome. Grade the decision, not the result. A well-reasoned allocation that failed due to unforeseeable regime change deserves a different institutional response than a poorly reasoned allocation that happened to work. Conflating the two is how institutions learn to chase results rather than refine judgment.

Interrogate the exit, not just the entry. Most post-mortems fixate on why the strategy failed. The more valuable questions concern the exit itself: Did we act on the thesis breaking, or on pain? Would a committee with no prior involvement have exited at the same point? Was the termination a decision or a capitulation?

Write the doctrine. The output should not be a report that files itself away. It should be a short, standing document — one page — that articulates the transferable lesson and is attached to the policy framework where future decisions of similar character will be made. "Before allocating to semi-liquid credit, read the 2022 post-mortem" is a governance mechanism, not a suggestion.

The Compounding Case

The compounding logic here is straightforward. An institution that terminates one strategy per year and conducts disciplined post-mortems accumulates, over a decade, a proprietary body of decision research that no consultant can replicate and no benchmark can capture. It becomes the rare organization that knows its own failure modes by name.

There is also a cultural dividend. Teams that know closures will be examined — fairly, blamelessly, and with an eye toward process rather than persons — make more honest recommendations at inception. The knowledge that an autopsy awaits tends to improve the quality of the living.

The machinery of American institutional investing is superb at putting capital to work. It is nearly silent on what to do when the work ends. Closing that gap requires no new technology, no consultant, and no budget line — only the governance will to treat every dead strategy as tuition already paid, and to collect the education it purchased.