A red portfolio is a bad place to invent a new strategy

An investor buys a stock, watches it fall, and decides to wait until it returns to the purchase price. Another investor sees the same decline and wants to sell immediately. Both decisions can feel obvious to the person making them. Neither feeling tells us much about what the investment is worth today.

The purchase price has become a personal reference point. It marks the difference between being right and being wrong, between a mistake that is still recoverable and one that must be admitted. Yet the company does not know what either investor paid. Its future cash flows will not change because someone wants to get back to even.

That mismatch is one reason losses deserve more careful discussion than the familiar instruction to stay calm. A loss can change the question a person is trying to answer. Instead of asking where the remaining money belongs, the investor starts asking how to repair the earlier decision.

Daniel Kahneman and Amos Tversky's 1979 paper on prospect theory described choices in which people evaluated gains and losses relative to a reference point. Their account also helped explain why people could prefer certainty when considering gains yet accept a gamble when facing certain losses. This is a description of patterns in choices under risk, with important qualifications. It is not a rule that every losing investor becomes reckless. Prospect theory ↗

The practical concern is the possibility of an unannounced change in objective. Recovering a loss can become more important than evaluating the next decision on its own terms. That can encourage someone to accept an exposure they would have rejected before the loss occurred. It can also encourage avoidance: refusing to examine a position because examining it might require a painful conclusion.

Investment organizations should build their review processes around this problem. If a portfolio falls, the review should identify what changed in the underlying evidence and what changed in the manager's circumstances. Those are separate questions. A weaker business outlook may justify a different decision. So may a genuine need for cash. Embarrassment about the original purchase needs a different kind of attention.

A written investment rationale helps make that distinction possible. Before a position encounters trouble, record the conditions that support it and the evidence that would weaken it. When circumstances change, compare the new information with that record. The purpose is to make revisions explainable. A plan that cannot be revised is brittle; a plan that can be rewritten for any outcome is useless.

Institutions should be especially wary of imposing a recovery deadline without discussing the risks required to meet it. Telling a manager to erase a loss by the end of the quarter may sound like accountability. It also gives the manager an additional target that can conflict with the portfolio's original mandate. If the organization wants a different level of risk, it should authorize that change openly.

There are limits to what a record can do. Nobody can specify every future development. A manager can write an elegant rationale and still misunderstand the investment. Documentation also takes time. For consequential decisions, though, a short record of the actual reasoning is preferable to a detailed explanation constructed after the result is known.

The distinction matters outside investing. A company can continue an unsuccessful project because canceling it would expose the original error. A team can escalate a commitment because retreat would look humiliating. In each case, the decision deserves a review based on the choices available now, including the costs of changing course.

A loss provides information, but its emotional force is not a valuation method. A serious review can conclude that the original position still makes sense. It can also conclude that it should end. What it should never accept as sufficient reasoning is a date, a purchase price, and the wish to make the discomfort disappear.