The cost a signal must beat is not a fixed number. It rises and falls with the market, and the biggest mover is volatility. When conditions turn turbulent, spreads widen and slippage deepens, lifting the bar a signal has to clear. The same setup that passes comfortably in a calm market can fail in a chaotic one — and that is the cost rule reflecting reality, not malfunctioning.
How volatility raises the bar
Two of the four costs a signal must beat are directly sensitive to volatility. The spread tends to widen when market makers price in uncertainty, so you pay more to cross it. Slippage deepens because price moves faster between the moment a signal fires and the moment an order fills. The safety buffer, which exists for uncertainty, is most justified precisely when volatility is high.
Why the same signal can pass and then fail
Imagine a signal with a fixed expected edge. In a calm session, the spread is tight and slippage minimal, so the edge clears the cost bar and the trade is valid. Hours later, in a volatile session, the identical edge now faces a wider spread and deeper slippage — the bar has risen above it, and the signal is rejected. Nothing about the signal changed; the cost of acting on it did.
A worked example: the same edge, two sessions
Numbers make this concrete. Take one signal with a fixed expected edge of 35 basis points (0.35%) and a flat exchange fee of 16 bps for the round trip. Everything else — spread, slippage, and the safety buffer the cost model layers on for uncertainty — moves with volatility. Below is the same signal evaluated in a calm session and again hours later when conditions turn turbulent. All figures are basis points (1 bp = 0.01%).
| Cost line | Calm session | Turbulent session |
|---|---|---|
| Expected edge | +35.0 | +35.0 |
| Exchange fee (round trip) | −16.0 | −16.0 |
| Spread crossed | −6.0 | −13.0 |
| Estimated slippage | −4.0 | −11.0 |
| Safety buffer | −3.0 | −7.0 |
| Cost-to-beat | 29.0 | 47.0 |
| Net after cost | +6.0 — passes | −12.0 — rejected |
Nothing about the setup changed. The edge is identical at 35 bps in both columns. But volatility roughly doubled the spread, nearly tripled the slippage, and widened the buffer — lifting the cost-to-beat from 29 bps to 47 bps. In the calm session the edge clears with 6 bps to spare; in the storm the same edge falls 12 bps short, and the cost model logs a transparent rejection. The discipline is in the math, not a mood.
Discipline in turbulent markets
The instinct in volatile conditions is to trade more, because the moves are bigger. The cost rule pushes the other way: bigger moves come with bigger friction, and only signals whose edge has grown faster than their cost should pass. Fewer valid signals during turbulence is the system protecting you from paying a high cost bar for an edge that did not keep up.
Common questions
Does a higher cost bar mean the signal was wrong?
No. The edge can be perfectly reasonable and still fail to clear a raised bar. A rejection in turbulence says the price of acting outgrew the expected reward, not that the underlying read was mistaken — see when a signal is right but still unprofitable.
Why does slippage move more than the fee?
The exchange fee is a fixed rate, so it is the same in calm and chaos. Spread and slippage are conditions-driven: they widen as uncertainty rises and fills land further from the quoted price. Spread vs slippage in the cost model separates the two precisely.
Should I lower the buffer to pass more signals in volatile markets?
That inverts the purpose of the buffer, which is largest exactly when uncertainty is highest. Loosening it to force trades through is how a flood of marginal trades sneaks in — the opposite of the case made in why fewer signals beat a firehose of noise.
This is why volatility belongs in every cost calculation, not as an afterthought. See how to read a signal cost breakdown to watch the spread and slippage lines move, and the cost-beating rule for the principle. To put a number on volatility itself, the tools pillar has a volatility dashboard.