Executive Summary
BUIDL flipped CALM to WATCH at score 18; USDC shed $3.25B in 48 hours while spUSDT mirrored it. PSI walked 93.6 to 97 in BEDROCK; Bank Run Gauge hit -53.3 STRESS anyway.
BUIDL's DEWS flicked CALM to WATCH at score 18 on $2.87B of float, driven by supply velocity, the only top-of-leaderboard item not labeled suppressed. PSI walked from 93.6 to 97 across six days, dominant band BEDROCK throughout, midpoint 95.2 against last week's 95.8. The headline risk wasn't an event; it was a tokenized treasury fund twitching while everything else stayed inside the lines.
The week's thread was issuer rotation. USDC printed the two largest movers, -$1.73B on May 15 and -$1.52B on May 14, $3.25B gone in forty-eight hours. spUSDT mirrored it with +$1.16B, +$1.15B, +$1.25B, +$1.25B across May 12, 13, 16, and 17. USDG opened Monday with +$778M in twenty-four hours, three-quarters of its $1.01B weekly intake, before fading from the daily headlines.
The counter-move was upgrades hiding inside downgrades. FDUSD dropped B- to C on both May 15 ($393.29M) and May 16 ($393.45M); TUSD touched C and recovered to B- the same day on May 15 at $484M; sDAI climbed C+ to B at $190M on May 16 and again May 17. Twenty-four grade transitions against forty last week, but the heavyweights are the ones who moved.
Capital flows: $350.38B to $353.55B, +0.91% on the week and +3.38% versus prior. Active depeg observations fell 94 to 72; unique signals held at 25, with SUSD's 4439 bps gash narrowing to 3598 by Sunday but never closing on $33.60M of float. Blacklist activity exploded to 675 events worth $65.36M against last week's 86, USDT on Tron carrying the single largest $8.61M strike on May 15.
The strangest line on the page: Bank Run Gauge printed -53.3 STRESS on May 15 while PSI held 95.4 BEDROCK on its 52nd consecutive day, the kind of split-screen reading you usually only see in pre-quarter-end repo prints. Watch BUIDL's score next week; if supply velocity keeps it at WATCH and USDC keeps shedding billion-dollar days, the BEDROCK band starts looking less like consensus and more like a misweighted average.