US Recession Probability Falls to 23% Despite Deteriorating LEI

---

Prediction Markets Say Soft Landing, But What Are They Actually Pricing?

The Conference Board's Leading Economic Index fell 0.2% in January 2026, extending a decline that has preceded every U.S. recession since the 1970s. February payrolls printed at -92,000, the first negative reading in months. And yet, initial jobless claims sit at a tranquil 210,000, creating a split-screen economy where the leading edge is cracking while the coincident data stays calm.

Prediction markets have chosen the calm screen. The "US recession by end of 2026?" contract has fallen from 32% to 23% over three days, a 9-percentage-point move that amounts to a soft-landing verdict. Kalshi prices the event at 22%, Polymarket at 24%, and the tight 2-point spread between platforms suggests genuine consensus rather than a platform-specific anomaly.

The drop feels decisive. But prediction markets have a pattern of moving late relative to leading indicators. In early 2007, recession implied probabilities fell even as the yield curve had already been inverted for months. The 23% price isn't necessarily wrong. It may, however, be suspiciously tidy given the volume of macro data pointing in the opposite direction.

---

The Leading Economic Index Is Quietly Telling a Different Story on US Recession Risk

The core tension in this market is taxonomic: traders are weighting lagging indicators over leading ones. The LEI is designed to move before the economy weakens. Its January decline, paired with deteriorating six-month momentum, is precisely the kind of signal that has reliably preceded downturns for five decades. That is not a coincidence. It is the indicator doing its job.

Beneath the headline LEI number, the components tell a more granular story. New manufacturing orders are soft. Consumer expectations remain depressed, with the Conference Board's short-term expectations index stuck below the 80 threshold often associated with recession risk. Temporary help services, a reliable canary for broader labor-market turns, continue to contract.

Then there is the "Vicious Cycle Index", a composite designed by Moody's economist Mark Zandi to track self-reinforcing deterioration across credit, employment, and manufacturing. It is elevated, suggesting feedback loops are forming beneath the calm surface of headline unemployment. The pattern it describes is familiar: firms slow hiring, consumers pull back spending, businesses respond with further caution. Once that loop activates, the labor data that markets are leaning on catches up quickly.

The distinction matters for anyone trading this contract. Jobless claims, the unemployment rate, and payroll revisions are lagging metrics. Firms stop hiring after conditions worsen, not before. The LEI's whole purpose is to see around that corner. At 23%, this market is betting that it won't.

---

The Strongest Case Against a US Recession by 2026

The bull case for a soft landing deserves honest engagement because it rests on data, not just sentiment. Start with the yield curve: the 2s10s spread sits at roughly +56 basis points, firmly positive after spending much of 2022 through 2024 inverted. According to RecessionPulse's March 28 assessment, the recession risk score stands at 44 out of 100, labeled "elevated" but explicitly not imminent.

Manufacturing is cooperating. The ISM Manufacturing PMI printed at 52.4 in February 2026, above the 50 expansion threshold. Continuing claims fell to approximately 1.82 million in mid-March, a move inconsistent with the broad-based layoff wave that typically precedes recession. The Sahm Rule indicator reads 0.27, well below its 0.50 trigger.

Goldman Sachs and J.P. Morgan have both lowered their recession probability estimates to between 20% and 30% for the next 12 months, roughly in line with where prediction markets are trading. The institutional consensus and the crowd-sourced price are aligned, which historically makes a contrarian bet harder to sustain.

The strongest version of this argument: the economy is in a low-layoff, low-hire stall. That is uncomfortable but not recessionary. If April hiring data rebounds even modestly and claims hold below 230,000, the LEI decline could prove to be a false alarm, a slowdown that never crosses into contraction. That outcome is what 23% implies, and it is plausible.

---

What Would Push This Market Higher

For the 23% price to be correct, the labor market needs to stabilize in the next two months. Specifically, payrolls must return to positive territory and initial claims must stay anchored below 230,000. If April and May data confirm the February -92,000 print was an outlier, the contract likely drifts toward 15% or lower.

The bear case requires the opposite. If claims begin trending above 240,000 for three consecutive weeks, or if March payrolls come in negative again, the feedback loop that the Vicious Cycle Index is tracking becomes observable in real time. At that point, the leading indicators and lagging indicators converge, and the market reprices sharply.

This contract resolves on December 31, 2026. That gives the economy eight and a half months to either validate the soft-landing consensus or prove the LEI right. With consumer confidence still fragile and the expectations index below 80, the margin for error is thin. The 23% implied probability is a bet that calm labor data can outrun a slow-motion leading-indicator deterioration. History suggests that race usually has the same winner.