Bank Run

Banks are a bet on the curve and on credit. Right now the curve is cooperating and the credit file is smoking — which is why this book is long the deposit franchise and only tactically long the trading books. First, my own discipline. Monday I called the broker bounce unsponsored and said I'd re-arm GS long only over 906.50. Tuesday GS opened 912, held 906.59 all session and closed 911.77 on 0.80x. That is the trade my line told me to take, so I'm taking it, and MS tagged 189.12 through the 187.16 pivot before closing 186.91 right on it. But understand the size of what happened and what didn't: GS has now put in a clean higher-low sequence (859.03, 870.82, 881.32, 906.59) and reclaimed its 20-day, and it is still 3.0% under a 50-day at 939.56; MS is 4.5% under 195.70 and couldn't hold the pivot. Five straight sessions across this complex on sub-1.0x relative volume. That's a post-earnings catch-up inside a downtrend, not a turn. GS long, stop just under 899.50, targets 931 then 939.56, and I don't add — the moment it loses 906.50 I'm out and flat, and I'd rather own GS than MS until MS closes over 189.12. The ratio is still the spine of the book. KRE/XLF 1.2989 is a fresh high for this move and it got there on a day when both legs closed red — that is what relative strength actually looks like. Above its 20-day (1.2818) and 50-day (1.2930). The caveat is unchanged and I will keep printing it: only 4 of the last 20 sessions closed above both averages and the ratio sits in the 26th percentile of six months. Reclaim, not regime. KRE itself has now been rejected twice at 73.42/73.13 while GS was running, which tells you Tuesday was rotation INSIDE the complex out of the safe deposit trade and into the beaten-up brokers, not money arriving. KRE long stays small: over 73.50, or a retest of 71.86/71.03, stop 71.07, targets 73.95 then 74.93, and the 67.79/67.96 double bottom measures to ~79 if 73.50 breaks. On credit, C's September card file is the hard data and it is deteriorating — 30-day delinquencies 1.35% from 1.29%, 90-day 0.65%, net loss rate 2.18% — and yet C reclaimed the 119.76/119.98 shelf and closed 122.69. Bad news, a bid, and no breakdown means the consumer tail is priced for now. So I do not press C puts at the low; puts only on a close under 119.76, and I fade 125.99-127. BAC is the same consumer-credit story with a $1.21B sanctions penalty and a securities class action on top: RSI 27.9, 8.7% under its 50-day, and six sessions pinned between 50.39 and 51.08 on 0.58x volume with a rising floor off 49.26. Coiled, not resolved. Short only under 49.26, fade 52.24-52.64, no long. The name that keeps me honest is SCHW. Down 9.9% in twenty sessions while KRE is up 3.9% — 13.8 points of spread in a month — and it just printed a fresh lower low at 89.21, the only name in the whole complex doing so. A deposit franchise is exactly what should be bid on a steepening curve; SCHW is not bid. Either it's idiosyncratically broken (client-cash yield, fee compression, retail brokerage competition) or the deposit story is thinner than my ratio says. Either way I want a small short leg as a hedge, entered on a bounce into 91.50-92.25, stop 93.10, targets 87.63 then 85.35. Size it as insurance, not conviction. COF is the trade that's working and I'm leaving it alone: above its 20-day (201.15) and 50-day (203.10), $6.69 against $5.45, Discover synergies, and one of only two names averaging above 1.0x relative volume on this bounce. Long 200.70-202.60, stop 198.70, targets 208.45 then 211.49, no chase over 205.50. Earnings season is behind us until January, so from here it is purely the curve and the credit file. One line still governs the whole book: JPM at 335.79, dead center of 325.07-344.40 and straddling its 20-day. Over 344.40 I add to KRE and COF and stop fighting the brokers. Under 325.07 I cut KRE, COF and GS without argument — that's a sector-wide warning, not a JPM trade.