Intuitively, the pressure to sell ought to scale with the loss, so that deeper falls become correspondingly harder to hold through. Several of history's fastest and most severe declines suggest something different. Where a fall completed itself within days and was followed by a recovery, most investors held on. Afterwards they frequently described the episode as one they had not felt especially strongly at the time. It was the shallower declines, the ones dragging on for months, that left the heaviest record of realised losses.
The explanation may lie in the different ways people handle events and conditions. An event has a beginning and an end, and people carry a script for events. They endure it, they wait, and in due course it concludes. A condition has neither an ending nor a script attached to it. A slow decline with no clear cause and no visible sign of finishing cannot be classified as an event. None of the existing responses can therefore be applied to it.
What gets consumed during such a period is a stock of belief that the situation is temporary. Every day without a recovery draws that stock down a little further. The rate of depletion has little to do with the size of the decline and a great deal to do with its duration. When the stock reaches zero the sale occurs, and what triggers it is rarely any fresh piece of bad news. It is simply one more day on which nothing at all changed.
This accounts for a pattern often observed and seldom explained. The peak in capitulation frequently arrives after a decline has already begun to level off, rather than at its most violent point. At the most violent point, attention remains fixed upon an event that appears to have a shape. Once the decline levels off the sense of event recedes, and what remains is a condition. The condition is the thing that actually wears people down. A condition has no shape to it, and shapeless things are very hard to wait out.
Duration also interacts with how often a position is examined. The same decline, reviewed once a quarter, presents itself as a single observation. Reviewed daily, it presents itself as sixty consecutive confirmations that nothing has improved. The underlying facts are identical in both cases, but the number of times the belief is asked to hold is not. Much of what gets described as a difference of temperament between investors turns out to be something simpler. It is a difference in how often they look.
If tolerance is principally a function of time rather than of magnitude, the unit used in preparation ought to change accordingly. Most risk assessment asks about magnitude, enquiring how large a decline could be absorbed. Answers to that question tend towards optimism, because the respondent is imagining a discrete event. Asking in units of time produces a more honest answer. How long could an existing plan be left unchanged while sitting inside a decline with no visible end? That version of the question tends to produce a smaller and considerably more accurate number.
Reframing the question will not make declines comfortable, and it will not shorten the waiting by a single day. Its only function is to move a person's estimate of themselves out of a dimension where overestimation comes easily. Investment plans usually fail at the point where the investor misjudged themselves. Markets are misjudged far less often than the record afterwards suggests. Correcting the second kind of error requires nothing from the market and everything from the person holding the plan. Almost all of that work has to be completed before it is needed.