How we see the market

This is the oldest note on this page, written before there was much to show. It is not about a strategy or a piece of the stack. It is about how we see the market, because everything we build afterward inherits that view, and it is worth writing down plainly before the work buries it. A market is not a number that goes up and down. It is a system — humans, machines, and incentives, negotiating a price in public, in continuous time, and never quite settling.

Start with the object itself. A modern market is a limit order book: a queue of intentions, most of them conditional, many of them fleeting, resolved by rules that are public and mechanical. Price is not a fact that exists and then gets reported. It is the output of that queue clearing, moment to moment, against whatever flow arrives. Which means the thing you are trading is not really the asset — it is a stream of decisions other people are making about the asset, compressed into a single number and a depth profile. Read that way, "the price moved" is a summary of a negotiation, not an event in itself.

The participants in that negotiation are not the same, and pretending they are is where a lot of bad intuition comes from. A pension fund rebalancing on a schedule, a market maker quoting both sides to earn the spread, a macro desk expressing a view on the yield curve, and a latency-sensitive strategy racing to the front of the queue are all "the market," but they want different things, operate on different clocks, and leave different footprints. Some are trading because they have information. Many are trading for reasons that have nothing to do with information at all — a mandate, a margin call, an index reconstitution, the end of a quarter. The structure of who is forced to trade, and when, is often more legible than what anyone believes.

That heterogeneity is why we treat the market as adaptive rather than efficient in the textbook sense. Prices incorporate information quickly — quickly enough that most simple edges are gone before you notice them — but the participants themselves are learning, crowding into what worked and abandoning what stopped working. An advantage is not a property of the world like a constant; it is a temporary imbalance between what a price implies and what is actually going on, and it decays precisely because other people are looking for the same thing. Any honest view of the market has to account for its own erosion. The edge you find is, by construction, one that others will eventually find too.

This is also why we are skeptical of prediction as the organizing goal. Forecasting the level of a price is a hard, low-signal problem dressed up as a solvable one, and the payoff for being marginally better at it is usually smaller than the cost of being confidently wrong. We find more durable ground in structure: the mechanics of how orders interact, the constraints participants operate under, the way risk and liquidity get repriced when conditions change. You do not have to know where the price is going to know that a given arrangement of flow and constraint is unstable in a way you can act on. The question we keep returning to is not "what will happen" but "what is the current state, and what does it make more or less likely."

Two facts sit underneath all of it and never go away. The first is cost: every interaction with the market has a price — the spread you cross, the slippage you cause, the fees you pay — and any view that ignores those is describing a market that does not exist. The second is uncertainty: you are always acting on a partial, noisy read of a system that can move against you faster than you can respond. Those two together are why we think risk is not a module you add at the end but the frame the whole thing hangs on. Survival is a precondition for compounding, and most ways of being right slowly are indistinguishable from being wrong if a single bad draw ends the game.

So this is the ground the rest of the work stands on. We see the market as a live, adaptive system of people and machines, priced in public, honest about its costs, and indifferent to how badly anyone needs to be right. What follows on this page — the research, the infrastructure, the harness we are building around a human trader — is downstream of this view, not the other way around. We wanted to write it down first, plainly, so that everything after it can be read as an answer to a question we had already decided how to ask.