The Bias Ratio

Some hedge fund return series look too smooth: a long, steady run of small positive monthly returns with almost no losses, which is exactly the pattern that shows up when a manager marks illiquid positions at whatever value keeps the return series looking clean, rather than at an independently observable market price. The bias ratio was designed as a quick, purely statistical way to flag this pattern without needing to inspect the underlying holdings.

It works by comparing how many monthly returns sit in a narrow band just above zero against how many sit in the same-width band just below zero. Genuine returns, especially from a liquid, honestly-marked strategy, tend to be roughly smooth across zero — small gains and small losses in similar numbers. A return series with an unusually large excess of small gains right above zero and very few small losses right below it is a statistical fingerprint consistent with a manager smoothing away small losses at valuation time, which is precisely the mechanism behind several well-known fund frauds where NAV was manipulated rather than legitimately earned.

The bias ratio is a screening flag, not proof of wrongdoing — some genuinely conservative, low-volatility strategies produce a similar shape honestly. It's meant to prioritize which funds deserve a closer look at their valuation policy, not to replace that closer look.