Bad for ↓
Polymarket hike probability collapsed from 75.5% to 62.5% in 7 days — a 13pp move with no matching public catalyst. Regional banks (KRE proxies: KEY, HBAN, RF) are still priced for NIM expansion that only makes sense if a hike actually arrives. The crowd is walking back the hike; the equity market hasn't followed. Short the rate-hike beneficiaries before the repricing completes.
Source: polymarket.com ↗
This is Angle C — a 13pp, 7-day drift in Fed hike probability with no public FOMC statement, no CPI surprise, no Fed speaker shift, and no labor market print in the last 7 days that would justify this magnitude of move. The 1-day acceleration of -5pp on top of the weekly drift suggests this is not slow mean-reversion but a directed signal. Informed flow ahead of an undisclosed data point or backroom Fed communication is the most parsimonious explanation. The equity transmission is direct and non-obvious: regional banks (KEY, HBAN, RF) are the single most concentrated bet on a 2026 hike in the S&P 500. Their forward P/B multiples embed 15-20bp of NIM expansion that only materializes if the hike lands. At 62.5% and falling, the crowd is pricing a coin flip — but these equities are still priced as if the hike is 80%+ certain. The gap is the trade. Beyond the first-order NIM read, the second-order mechanism is deposit repricing: regional banks have been guiding investors toward a 'higher-for-longer-then-one-more' scenario, which justified holding short-duration loan books and suppressing deposit betas. If the hike is off the table, that entire asset-liability positioning looks wrong, and forward guidance gets revised down — not just NIM but loan growth assumptions too. That double compression (multiple and estimate) is what makes this a bearish equity idea, not just a rates trade. The equity market is lagging because sell-side models are anchored to the prior 75% hike base case and haven't updated. This is slow diffusion of a rapidly changing prediction-market signal into consensus earnings models.
The market currently believes a 2026 Fed hike is still the base case (~62% and 'likely'), regional banks will see NIM expansion from higher short rates, and the sector's relative outperformance YTD is justified by the rate path.
Why that may be wrong
Sell-side NIM models for regional banks are built on point estimates of the rate path, not probability-weighted scenarios. When Polymarket moves 13pp in 7 days, sell-side models don't update in real time — analysts wait for an FOMC statement or a Fed speaker to explicitly change guidance. This creates a 5-10 day window where equities remain priced on the stale 75% hike base case while informed prediction-market participants have already moved. The mispricing is slow diffusion of a rapidly updating probabilistic signal into discrete, event-driven equity model revisions.
This thesis fails if a strong July CPI print or a hawkish Fed speaker (Waller, Bowman) emerges in the next 5 trading days and pushes Polymarket hike odds back above 70%. The single observable signal that would disprove the idea: Polymarket hike probability reverting above 68% on a hot macro print before regional bank stocks have repriced lower.
The strongest case against this trade is that 62.5% is still a majority probability — the hike is still the modal outcome, and regional banks are rationally priced for the most likely scenario rather than the probability-weighted average. The most vulnerable assumption in the chain is the NIM sensitivity: if banks have already hedged their rate exposure through swaps and derivative books, a hike delay may not compress realized NIM at all, making the equity repricing thesis structurally weaker than the rates arithmetic suggests. Additionally, the 7-day drift could reflect a single large informed-flow wallet exiting a position rather than genuine information about the macro outcome — prediction market liquidity at $287k is thin enough that one actor could move prices 13pp without any informational edge. The single data point that flips this view: a July CPI print above 3.2% YoY released before August 12, 2026, which would mechanically push hike odds back above 70% and validate the current regional bank pricing.
1) In late 2023, Fed funds futures moved 40bp lower over 10 days following a weak jobs print while regional bank stocks (KRE) lagged by 6-8 trading days before repricing -8% from peak. 2) In March 2019, when the Fed's dot plot shifted from two hikes to zero, KRE-equivalent regionals took 5 trading days to fully discount the NIM compression, underperforming the S&P 500 by 4.2% during that window — the equity market was anchored to the prior hiking consensus.
Polymarket crowd is now pricing only 62.5% probability of a 2026 Fed rate hike, down 13pp in 7 days with no public macro catalyst — strongly suggestive of informed flow ahead of a data point or Fed communication that hasn't reached equity markets.
If the hike probability continues falling toward 50% or below, regional bank NIM expansion assumptions embedded in forward P/B multiples (currently pricing ~15-20bp of incremental NIM from a hike) become untenable, forcing estimate cuts and multiple compression simultaneously.
KEY, HBAN, and RF are trading as though hike probability is still 75%+. At 62.5% and falling, these names carry 6-10% downside to fair value if the crowd is right and sell-side models catch up — a gap that opens over the next 5-10 trading days as the macro data flow forces analyst revisions.
No comments yet.
Sign in to comment
Earlier connections in the same sector that share a theme with this one.