{"slug":"z-score","term":"Z-Score","category":"Quant & Stats","definition":"Z-score measures how far a value sits from its average, in units of standard deviation. A z-score of +2 means the reading is two standard deviations above its mean — rare territory if the data is roughly normal. Traders apply z-scores to price, funding rates, and basis to flag stretched readings that historically tend to snap back.","street":"It's how you tell 'this funding rate is high' from 'this funding rate is historically unhinged'.","depth":["The formula is (current value minus the mean) divided by the standard deviation, over your chosen lookback — and the lookback is the whole game, since a 90-day z-score and a one-year z-score can disagree about the same reading. Rough map: within ±1 is normal noise, past ±2 is stretched, past ±3 is rare. It works on anything with a history — price distance from a moving average, funding rates, basis, the ETH/BTC ratio.","The trap is treating ±3 as an automatic fade. Crypto's distributions are fat-tailed — extreme readings show up far more often than a normal curve promises, and in a regime shift the mean itself relocates, so you can short a 'three-sigma' breakout the entire way up. Use z-scores to rank how unusual a reading is, not as a standalone trigger. A stretched funding z-score into resistance is a setup; a stretched z-score alone is a statistic."],"live":{"label":"Spot the stretched funding rates","href":"/tools/funding-rates"},"related":["mean-reversion","funding-rate","basis","atr"]}