A z-score measures how many standard deviations a current observation is from its historical mean. It provides a standardized way to assess whether a yield, spread, or auction metric is at an extreme relative to history.
Z-score = (Current value - Mean) / Standard deviation
Interpretation:
z = 0: the value equals its historical average
z = +2: the value is 2 standard deviations above average (approximately the 97.5th percentile in a normal distribution)
z = -2: the value is 2 standard deviations below average
On this site, z-scores appear in several contexts:
Morning Dashboard: each tenor's yield is shown with its z-score relative to its full history, identifying which maturities are at extremes
Auction grading: the auctions tool converts raw auction metrics (bid-to-cover, tail, bidder shares) into z-scores using an expanding historical window, then maps z-scores to letter grades from D- to A
ChatYCP: the AI assistant uses z-scores and percentile ranks when answering questions about where current data stands relative to history
Z-scores assume roughly normal distributions. For metrics with fat tails (common in financial data), extreme z-scores may be more frequent than a normal distribution implies. The expanding-window approach used in the auction grading system (documented in "Grading US Treasury Auctions") mitigates this by using all available history rather than a fixed window.