The Week the Three-Month T-Bill Quietly Dropped

No Fed speech, no FOMC, no CPI surprise. VIX -6%, the 10Y-3M curve steepened 18 basis points, credit spreads tightened across the grade. The market made up its own mind about what’s coming.

Between Monday May 11 and Friday May 15, the U.S. Treasury 10Y-3M spread moved from 0.72 percentage points to 0.90 — a 25% steepening of the most-watched recession-risk gauge on Wall Street in five trading days. The 10Y end of the move was essentially flat (4.42 to 4.47). The arithmetic forces the conclusion: the 3-month T-bill yield fell roughly 13 basis points across the week, from around 3.70 to 3.57. The market was buying short bills.

It did this without an obvious policy catalyst. The Fed had no scheduled FOMC meeting, no Powell speech, no surprise SEP release. The week’s CPI print, when it came, did not deliver a downside shock. There was no liquidity event. The bull steepening of mid-May 2026 was the market quietly making up its mind that more easing is coming than is currently priced into the dot plot.

What the five days looked like

Five trading sessions. Every dial that prices risk moved looser.

Figure 1: The 10Y-3M Treasury spread widened from 0.72 to 0.90 percentage points over five trading sessions. With the 10Y yield essentially flat across the same window (4.42 → 4.47), the steepening was driven almost entirely by a fall in the 3-month T-bill yield — roughly 13 basis points lower, from approximately 3.70 to 3.57. Source: FRED series T10Y3M, DGS10.

The 10Y-3M curve is the cleanest tell, because the 10Y end barely moved. The compression of the 3-month T-bill yield — which trades almost entirely on Fed-cut expectations within a 90-day horizon — is what drove the 18-basis-point steepening. Markets do not flatten the front end of the yield curve without a reason. The reason is: more cuts, sooner.

Figure 2: VIX daily close fell from 18.38 on Monday May 11 to 17.26 on Thursday May 14, a 1.12-point or 6.1% decline. Long-run VIX median is approximately 17.5, so Thursday closed below median into easy-regime territory. The decline was monotonic across four sessions with no intraweek spike. Source: FRED series VIXCLS.

VIX closed at 18.38 on Monday and 17.26 by Thursday — a 1.12-point, 6.1% drop, dipping below the long-run median of roughly 17.5 into actual “easy regime” territory. That is not a calm-after-storm flatline. That is active risk-on, with the equity vol market reaching for less hedging.

Figure 3: High-Yield Corporate OAS compressed from 2.79 to 2.76 percentage points (3bp tighter); Investment-Grade Corporate OAS compressed from 0.78 to 0.76 (2bp tighter). Both moves are directionally consistent with VIX decline and a market that has reduced its institutional-risk premium. Source: FRED series BAMLH0A0HYM2 and BAMLC0A0CM.

Credit confirmed in both grades. High-Yield Option-Adjusted Spreads compressed from 2.79 to 2.76, with an intraweek pop to 2.82 before resolving tighter on Thursday. Investment-Grade OAS compressed two basis points, from 0.78 to 0.76. When equity vol and credit spreads move in the same direction at the same magnitude, the market is not telling you about one specific asset class; it is telling you about the path of policy.

SOFR, the secured overnight financing rate that anchors the front end, drifted from 3.60% on Monday to 3.56% by Thursday — four basis points of softening in the actual funding market itself, not just in expectations.

Why it matters

The Fed’s December dot plot guidance pencilled in two cuts for 2026. As of last Friday’s close, fed funds futures implied something between two-and-a-half and three cuts before year end — the market is pulling forward roughly half a cut versus the official guidance. None of this is screaming. It is exactly the kind of slow, quiet repricing that happens when no individual data release is dramatic enough to make news but the cumulative weight of soft data is shifting expectations.

The stuff that probably matters: April payrolls came in soft, May initial jobless claims have ticked up modestly, manufacturing surveys for the past three months are running below 50, and CPI year-on-year is now closer to 2.6% than to 3.0%. None of those individually justify a re-rate of Fed expectations. Stacked together, with a Treasury market that started the year worrying about more inflation rather than less, they justify exactly what we are seeing: a slow, broad de-risking with no headline cause.

Two normal explanations and one harder one

Normal explanation one: technical mean reversion. Weight I’d assign: ~35%. VIX entered the week at 18.38, modestly elevated relative to the long-run median. Spreads were at the wider end of their two-month range. Curve was on the flatter side of the recent distribution. Pure mean reversion would deliver something like this even with no fundamental news. The 18-basis-point curve move strains this explanation — that is more move than mean reversion should produce in a week without help — but it does not break it.

Normal explanation two: post-storm relief. Weight: ~25%. The prior two weeks featured a notable run of Supreme Court activity (Mifepristone telehealth, A.J.T. v. Osseo, federal Bruen-derived ATF rule revisions). Markets carried a low-grade institutional-volatility premium during that stretch. The rulings concluded, the news cycle moved on, and the premium quietly evaporated. This is a credible piece of the move but cannot be the whole story, because the 3-month T-bill drop is not a relief-rally move; it is a forward-expectation move.

The harder explanation — and the one I’d weight: the front end started front-running a dovish pivot. Weight: ~40%. Real economic data over the past six weeks has consistently come in slightly softer than the Fed’s December baseline — softer payrolls, softer ISM, softer non-shelter inflation. No single data point has been dramatic. But the bond market does not need a dramatic data point. It needs the gradient. The gradient is now consistent enough that the natural play is to pull forward the cut timing. This is what bull steepenings under non-recessionary conditions look like: a quiet, persistent compression of the front end while the long end stays anchored to long-run real-rate expectations.

What to watch next week

• The 3-month T-bill yield itself. If it continues falling — toward 3.50% or below — without a Fed speech or FOMC, the market is fully pricing in a June cut. If it bounces back to 3.65%+ on any moderately strong data, the move was technical.
• HY OAS direction. Sustained tightening below 2.70 would be credit confirming the dovish-pivot read. Reversal above 2.85 without a news catalyst means the calm was technical.
• The May FOMC minutes (release date pending). If the minutes reveal genuine internal Fed disagreement on the path of cuts, the market’s pull-forward is justified. If they reaffirm the December baseline, expect a brief reversal.
• The next CPI print. A downside surprise validates the move and probably pulls fed funds futures to three full cuts. An upside surprise reverses everything.

Reproducibility note

Daily VIX (FRED VIXCLS), 10Y Treasury yield (DGS10), 2Y Treasury yield (DGS2), 10Y-3M spread (T10Y3M), 10Y-2Y spread (T10Y2Y), HY Corporate OAS (BAMLH0A0HYM2), and IG Corporate OAS (BAMLC0A0CM) come from FRED’s public API. SOFR is from the New York Fed’s reference-rate publication. All values cited are daily closes between May 11 and May 15, 2026. The 3-month T-bill yield is computed by subtraction (10Y yield minus 10Y-3M spread). Fed funds futures probabilities are sourced from CME FedWatch; the interpretive call — “the market is pulling forward roughly half a cut versus official guidance” — is mine.

Counts and series in this report are computed from public market data sources (FRED, NY Fed, CME FedWatch) for the period of May 11 through May 15, 2026. Interpretive readings (“front-running”, “technical”, “relief rally”) are flagged as such. — Andrii

Originally published at https://sovenyr.substack.com.