The Half-Life of Moats: Concentration, Conviction, and Decay

I love, and hate, airports. In an airport you can almost feel the slow churn of human misery as people are ordered, flowed, and entrained to particular rhythms of being. And at the same time, once you stop and realise this, you can watch how it unfolds, and make conscious steps to anticipate how difference is produced. So it was poetry, almost, that it was in an airport recently where I thought about how this is happening in markets.

There is a particular sound a market makes when it stops thinking. It is really the sound of agreement. Everyone owns the same ten companies, recites the same theses, and points to the same handful of founders as proof that the future has already been decided. We have learned to call this conviction, and to treat it as a virtue. I am increasingly coming to think it is the most fragile thing in the system in this first third of the 21st century.

Conviction, in its healthy form, implies a willingness to act under uncertainty. You weigh the odds, you concentrate where you have an edge, and you accept that you might be wrong. That last clause is doing all the work, because the moment it falls away, conviction stops being a probabilistic posture and becomes something else: a quasi-religious creed. And a market organised around a creed is no longer pricing the future. It is, writ like the golden calf-turned-bull itself, worshipping a version of it.

This is where I think we actually are. Capital has concentrated into a small set of obvious winners on the belief that their dominance is not a bet but a fact. The largest companies are priced as though the future is known and they already own it. The most celebrated founders are treated less as operators who made good calls and more as prophets whose continued success is doctrinal. Monopoly, concentration, and rent-seeking are the visible structures, but the thing underneath them is religious. We have decided what is true, and we have arranged our money to confirm it.

I want to be precise, because the easy version of this argument is wrong, and I have no interest in the easy version. The problem is not bigness. Scale is often the correct answer. I believe in natural monopolies, and I have written elsewhere about the difference between a secret of depth and a secret of opacity, and that distinction is the whole game here too. A secret of depth is a quite real, hard-won advantage that cannot be easily undone: decades of engineering, a dataset that deepens with every observation, the kind of position that earns its rents. I have no quarrel with that. A secret of opacity is something else: an arrangement that depends not on depth but on the absence of the right lens, on no one yet being able to see clearly enough to compete. The two look identical on a balance sheet. They are not identical at all.

The error of the conviction regime is that it has stopped telling them apart. It prices every moat as though it were a moat of depth, durable for decades, when a great many of the positions commanding the highest premiums are moats of opacity whose half-life is collapsing. The secret that took twenty years to leak now leaks in twenty months. And yet the market is still paying twenty-year prices for it. That is the mispricing. The creed is paying for a permanence that the world has quietly stopped supplying.

This is also why the standard defence fails. If conviction is risky, the textbook says, diversify. But diversification has been hollowed out by the very regime it was meant to hedge against. When everyone is convinced of the same future, diversification collapses into diversified conviction. You can hold forty positions and still own a single idea, because the forty are all expressions of the same consensus about who wins. Spreading your capital across forty instances of the same belief is not a hedge. It is the belief, wearing a costume. The diversified portfolio of the conviction era is monocultured MANGOs with good manners.

So the sequence we inherited is exhausted. Diversification came first, as a response to not knowing. Conviction came next, as a response to thinking we knew. And the thing we are missing is what comes after conviction, once we admit that conviction itself has become the risk. I have written about it before, and I lean in further here, to call that thing obviation.

I have to be careful with the word, because it has been doing two jobs and only one of them is honest. To obviate is to make something unnecessary, to render the question moot. It is not to take the other side of the bet. This matters because there is a weak version of what I am about to say and a strong one, and they are easy to confuse. The weak version is simply good allocation: stop paying creed-prices for crowded, decaying moats, and move your concentration laterally into positions the consensus has stopped examining. That is sound, and I will come to it, but it is not obviation. Relocating your capital to the unloved asset does not make anything moot. It is a bet that the crowd has mispriced, no more. It is going sideways.

The strong version earns the word: you obviate a moat not by storming it and not by declining to pay for it, but by changing the terms so that crossing it stops mattering. Galileo did not out-argue the Church on cosmology. He built an instrument, and the Church's authority over the question became beside the point. That is the move with historical force behind it: not competition, but the instrument that makes the competition irrelevant. The microscope did it to the doctrine of disease. The telegraph did it to the geography of arbitrage. The thing that captures returns across generations is rarely the position that won the old contest. It is the instrument that dissolved the contest, the lens that makes a previously opaque system legible and whose value compounds with every observation rather than leaking with every disclosure. That is obviation, and it is what I mean by it.

So the move I am describing has two registers, and I want to keep them apart rather than let one borrow the authority of the other. The first is obviation proper: own or build the instrument that renders the incumbent's moat moot. That is where the deep generational returns sit, because the instrument does not compete with the secret, it ends the conditions that made the secret valuable. The second, lesser and more immediate, is the allocation tactic: while the instruments are still being built, refuse to fund decaying opacity-moats at depth-moat prices, and relocate your concentration to the overlooked. The first makes something unnecessary, whereas the second merely declines to be a sucker. I think both are worth doing, to be clear.

This is also why the obvious winners read to me as old models. SpaceX is an old model. The megacap platform is an old model. Not because they are failing, but because everyone agrees about them. Consensus is the tell – the instant a thesis becomes obvious, the edge inside it is gone, fully priced and believed. What remains is rent, the slow extraction of value from a position no one is competing to take because everyone has already conceded it. Rent is what conviction pays itself once the thinking stops, and rent on a decaying moat is the worst trade in the market: you are paying a premium for permanence precisely as the permanence evaporates. The obviating instrument is not trying to win that position. It is building the thing that makes the position irrelevant.

If the edge has left the obvious, it has gone somewhere. My claim is that it has gone in two directions at once. It has gone laterally, into the spaces the convicted have stopped looking at, the overlooked, the unglamorous, and the structurally ignored. This is where mispricing lives, because mispricing requires someone to not be looking. And it has gone, more deeply, into the instrument, the lens being built before the world knows it needs it, which is where the largest returns will eventually settle. The consensus has trained an entire generation of capital to look in exactly one direction. Both the lateral and the instrumental move begin by refusing to.

A dynamic market full of obviating players also happens to be healthier, less strangled by rent, more responsive, and more alive. It is the emergent property of the better trade. Domination is a bet that the world will stop changing, whereas obviation is a bet that it will not. I am betting it will not.

I hold this loosely, which is the point. The deepest failure of the conviction regime is that it cannot survive its own correction, because it never priced one. It built no room for being wrong, and so when it is wrong, it does not adjust, it shatters. That is the fragility in the title. Conviction looks like strength right up to the moment it looks like nothing at all.

We are at the beginning of that turn, I think. The creed is intact, the temples are full, and the obvious winners have never looked more inevitable. Which is when the move is cheapest, and exactly when no one wants to make it.