The ETH/BTC ratio is one of the most-watched charts in digital-asset trading. It also produces some of the most expensive losses in our experience, because the textbook mean-reversion setup keeps appearing without delivering the mean reversion the chart promises.
The pattern is seductive. ETH/BTC has bounced inside a range for most of the last five years. When the ratio approaches the lower edge of the range, the fade looks obvious. The chart suggests an asymmetric trade. The execution reality is meaningfully harder.
Three reasons it is harder than it looks
Regime breaks happen at the worst moment. The range that has held for most of the last cycle is anchored in a specific market structure: a relatively narrow set of large holders, a stable ETF approval landscape, and a fairly consistent institutional flow pattern. When any of those change, the range breaks and the apparent mean reversion turns into a one-way move that lasts months. Our position-sizing rule assumes the range may break at any time.
The carry is unfavourable. Holding a directional bet on the ratio means holding both sides through funding cycles and basis movements. The implicit cost over a six-month holding period is non-trivial, and most retail traders do not include it when sizing the trade.
Liquidity is asymmetric. Selling the ratio when it is high is well-supported by liquidity from arbitrage desks. Buying the ratio at the bottom of the range coincides with stress in the entire complex. Stop losses cluster. Execution slippage spikes. The trade that looked great on the chart trades worse than it ought to.
What we do inside the algorithm
The strategy takes ETH/BTC mean-reversion exposure only when three conditions hold simultaneously: realised volatility on the ratio is elevated (the market is already pricing meaningful moves), cross-asset correlation has not spiked above its trailing threshold (no broader risk-off in progress), and the funding rates on both legs are reasonably aligned (no leverage cliff to fall off).
When all three are present, the trade is asymmetric and worth taking. Outside those conditions, the chart looks great and the trade is a slow bleed.
The chart suggests the trade. The regime decides whether it is worth taking.
The practical lesson
The seductiveness of mean reversion is that it works often enough to feel like a system, and fails often enough to bankrupt the people who treat it as one. Discipline about the conditions under which you take the trade matters more than the entry signal itself.
This is true of every mean-reversion trade we have studied in any market, and the digital-asset version is no kinder than the rest.