When to Hold a Compounding Business and When to Sell
The hardest question in long-term investing is not what to buy. It is knowing when the reason to keep holding has quietly changed.
I wrote earlier this year about the mistake I keep repeating: selling great businesses once they reach my estimate of fair value. The pattern is simple to describe and hard to break. You buy at a discount, the stock appreciates, and something trained into you over years of looking for cheap stocks starts to feel uncomfortable. The position no longer looks cheap. So you sell.
And then you watch the business compound for another five years without you.
That piece identified the mistake and traced its psychological roots. This one is about what to actually do instead, a complete framework for the hold vs. sell decision that goes beyond the answer I always knew was true but never found fully operational: just hold great businesses for a long time.
That answer is correct in theory and nearly useless in practice. Because the real difficulty is not holding. It is knowing when something has genuinely changed, when continuing to hold is no longer discipline but denial. When the business you own has stopped being the compounder you thought it was, and the thesis you are defending exists more in your memory than in the current evidence.
That distinction is what separates long-term investors who build durable wealth from those who ride positions up and back down again, telling themselves the whole time that they are being patient.
The framework has three parts. The first is a classification step, because the hold vs. sell logic is fundamentally different depending on what kind of position you actually own. The second is a set of five questions I now run through before acting on any sell impulse, questions designed to surface whether the impulse is grounded in something that has genuinely changed or driven by the psychological patterns that consistently produce premature exits. The third is the harder half: the specific conditions that actually justify selling a compounding business, even when everything still looks fine on the surface.
Part One: Know What You Own Before You Decide What to Do With It
The most common source of hold vs. sell confusion is a failure to be explicit, at the time of purchase, about which kind of investment you are making. Most investors treat this as obvious. It is not. I have confused the two categories more times than I would like to admit, and the confusion is expensive in both directions.
Mean-reversion positions are businesses you own primarily because they are cheap relative to their normalized earning power. The thesis is that the market has temporarily mispriced the business, sentiment is too negative, a one-time event has depressed reported earnings, a cyclical trough has made the multiple look high on depressed numbers, and that when conditions normalize, the stock will re-rate toward a reasonable multiple of sustainable earnings. The opportunity is the gap between current price and fair value. Once that gap closes, the thesis is complete.
For these positions, selling near fair value is not a failure of nerve. It is the correct execution of the strategy. You captured the mean reversion. The work is done. The mistake is staying after the thesis has been delivered, holding because you have grown attached to the business, or because you have started telling yourself a different story about why the position is actually a long-term compounder. That story may even be true. But if it was not the thesis that justified buying, it should not be the thesis that justifies holding at a price that no longer offers a discount.
Compounding positions are different in kind, not just in degree. These are businesses with genuinely high returns on invested capital and a meaningful runway of reinvestment opportunities ahead of them, businesses whose intrinsic value grows materially each year regardless of what the stock price does. The thesis is not that the market will recognize a discount. The thesis is that this business will be worth materially more in five and ten years than it is today, and that the passage of time is itself a source of return rather than a cost to be managed.
For these positions, selling when price reaches your initial estimate of fair value is often the worst thing you can do, not because the stock will necessarily keep rising in the short term, but because your initial intrinsic value estimate was a point-in-time calculation that did not account for the value the business would continue creating while you held it. The intrinsic value you estimated at purchase is the floor of the opportunity, not the ceiling.
The failure mode for mean-reversion plays is holding too long. The failure mode for compounding positions is selling too soon. These are opposite errors that produce opposite outcomes, and the difference between them starts with being honest, at the outset, about which thesis you are actually making.
Before any hold vs. sell decision, I write down the original thesis in one or two sentences, not the narrative surrounding the business, but the specific mechanism by which I expected to earn a return. If the honest answer is “I bought it because it was cheap relative to what the business normally earns,” I have a mean-reversion play with a clear exit condition. If the honest answer is “I bought it because this business will compound intrinsic value at high rates for a long time and I wanted to own that compounding,” I have a different kind of position with entirely different exit criteria.
Part Two: The Five Questions Before Any Sell Decision
Regardless of which category the position falls into, I now run through five questions before acting on any sell impulse. The purpose is not to talk myself into holding. It is to force the distinction between an impulse grounded in genuine business change and an impulse driven by the psychological patterns, anchoring, loss aversion, the desire for closure, that reliably produce mistakes in both directions.
Question 1: Has the business changed, or has the price changed?
This is the most important question and the one most frequently skipped. When a position rises significantly from your purchase price, it starts to feel expensive. The feeling is real. But the question is whether the business has become expensive relative to its current intrinsic value, or whether intrinsic value has simply grown alongside the price.
A position that has doubled from your purchase price is not expensive if owner earnings per share have also roughly doubled over that period. The multiple you are paying has not changed. The business has grown. Conversely, a position that has risen 30% while owner earnings per share have stagnated is now genuinely more expensive in economic terms, even though the absolute gain is smaller.
This question forces a valuation update, not just an observation of price change. If you have not revised your estimate of intrinsic value since purchase, using the latest earnings, the latest competitive assessment, the latest view on reinvestment opportunities, you do not have enough information to make a sound decision. The price change is data. The business change, or the absence of it, is the answer.
Question 2: Is the return on invested capital still high and stable?
ROIC is the single most important metric for a compounding position because it is the actual mechanism of compounding. A business earning 20% on its invested capital and reinvesting those earnings at similar rates doubles intrinsic value roughly every four years. A business where ROIC has quietly fallen from 20% to 11% is no longer compounding at the rate the original thesis assumed, and the gap between expectation and reality may not yet be reflected in the stock price.
Look at the ROIC trend across a full cycle, not just the most recent year. Ask whether any decline is cyclical or structural. Cyclical ROIC compression, from a period of heavy capital investment ahead of an expansion, from an industry-wide pricing environment that is temporary, from a demand disruption that does not affect the underlying economics, does not change the long-term thesis if the competitive position is intact. Structural ROIC compression, from a genuine erosion of the advantage that generated the returns in the first place, from customers finding credible alternatives, from a cost base that has grown faster than pricing power will allow revenue to absorb, changes the thesis entirely. No amount of patience reverses structural deterioration.
The distinction requires judgment, not a formula. But the judgment is possible, and it is worth making carefully before you act.
Part One defines what you own. Questions 1 and 2 tell you whether the core of the thesis is intact. But the three questions that follow go further, into the territory where most hold vs. sell mistakes actually live: the endowment effect that makes you value what you own more than what you could own, the motivated reasoning that lets you defend a thesis the evidence no longer supports, and the opportunity cost you are paying every day you hold a position that is merely good when something better is available. Those questions are below. So is the harder half of the framework: the four specific conditions that actually justify selling a compounder, each with the exact signals I watch for and the reasoning that distinguishes real deterioration from the noise that surrounds any long-term position.
