Quant
Why prices look random — the random walk
Here's an uncomfortable idea that every quant has to make peace with: over short horizons, price movements are close to random. Not perfectly — that's where edge lives — but close enough that most of the "obvious" patterns you spot on a chart are noise your brain dressed up as signal. Understanding the random walk is what separates a real edge from a story.
The coin-flip model
The simplest model of a price is a random walk: each step, flip a coin — up a bit on heads, down a bit on tails. String thousands of those together and you get something that looks exactly like a price chart, complete with "trends," "breakouts," and "support levels." None of it was designed. It's just what accumulated coin flips look like.
Why markets push toward randomness
There's a force actively erasing easy patterns: other traders. If a chart pattern reliably predicted the next move, people would trade it, and their trading would move the price until the free money was gone. The more obvious and well-known an edge, the faster it's competed away. This is the grain of truth in the efficient-market idea — not that markets are perfectly random, but that exploitable structure is scarce and self-destructing.
Accumulated coin flips look exactly like a price chart — trends, breakouts and all. So "I see a pattern" isn't evidence of an edge; randomness produces gorgeous fake patterns for free.
What this means for finding an edge
- Your default hypothesis is "it's random." The burden of proof is on the edge to beat that, not on randomness to disprove itself.
- A pattern that looks great in hindsight is worthless — the random walk generates beautiful hindsight patterns all day. Only out-of-sample and robustness testing can tell edge from noise.
- Real edges are small and perishable. A tiny, statistically real deviation from the coin flip — sized carefully and defended constantly — is the whole game. Anything that looks big and obvious is probably overfit.
The random walk isn't a counsel of despair; it's the honest baseline. Markets aren't perfectly random — if they were, no one could win — but they're random enough that the job is finding the faint, fragile signal in an ocean of convincing noise, and never mistaking one for the other. Respect the coin flip, and your risk-adjusted results start telling the truth.