Virtually no investor argues against holding for the long term. The principle appears in nearly every investment document, every personal finance seminar, and every reassuring statement issued after a market decline. Its correctness enjoys ample historical support, and its formulation is simple enough to require no explanation at all. Yet holding-period data consistently shows something rather different. Most investors keep a given position for a small fraction of the horizon they claim to be working with. The gap between the stated horizon and the realised one is wide enough to require explanation.
Attributing that gap to a lack of discipline explains very little in practice. It restates the outcome and identifies nothing that produced it. A more specific explanation lies in the construction of the principle itself, which has never actually been defined. Nobody has specified whether long term means three years, seven years or twenty. Nobody has specified which circumstances would justify selling earlier than that. A principle without boundaries cannot be violated, because any sale whatever can be said not to breach it.
A second structural problem is that the principle is declared in calm conditions and tested only under stress. At the moment of declaration, the future decline is entirely abstract. At the moment of testing, the present decline is concrete and arrives with a reason attached that sounds thoroughly sound. People under pressure rarely overturn their principles outright, and what happens instead is reinterpretation. This time is different, the change is structural, and long-term holding always assumed the fundamentals were intact.
In hindsight this reinterpretation looks like an excuse and is usually described as one. In the moment it is entirely persuasive, because the information it draws upon is real information. That is precisely what makes it so difficult to guard against in advance. Any principle expressed in loose language will, under pressure, be reread in whatever direction the pressure happens to point. The person doing the rereading will almost never notice that it is happening at all. Reinterpretation feels, from the inside, exactly like thinking clearly under difficult conditions.
The vagueness is not accidental in every case. A principle stated loosely enough can be endorsed by everybody, including parties whose interests are served by activity rather than by patience. It reassures during declines, demands nothing during advances, and can never be shown to have failed. Precision would remove all three of those advantages at once. None of this makes the principle wrong, and the historical evidence behind it remains strong. It does mean that anyone wanting the principle to do real work must supply the precision themselves.
What makes a principle executable is a boundary written down in advance. Resolve has surprisingly little to do with it. Which conditions constitute a legitimate reason to sell, and which conditions do not. Which developments fall inside expectations, and which fall outside them entirely. This work has to be completed during calm periods, because its only function is to substitute for judgement during periods when judgement is compromised. Once it exists on paper, a person may still sell, but they will at least know whether they are executing the plan or rewriting it.
A principle without boundaries is, strictly speaking, a preference. Preferences are honoured when matters go well and quietly abandoned when they do not. The periods during which matters go badly are the only periods in which a principle has any value whatever. The difficulty of long-term holding therefore lies not in the length of time involved. It lies in the fact that most people never turn it into something capable of being broken. Something incapable of being broken is also incapable of being kept, so it never gets tested at all.