Polymarket prices “US recession by end of 2026?” at 12¢ — a 12% market-implied probability — against Wall Street forecasts of 20% to 40%. That gap looks like a mispricing, but once the contract’s five-month window and its resolution lag are adjusted for, a defensible true estimate lands near 8–10%, meaning the value sits on NO rather than YES, and the edge is small at roughly 2–4 percentage points.
The headline comparison is seductive and wrong. Goldman Sachs has published US recession probabilities between 20% and 30% during 2026; J.P. Morgan and EY-Parthenon have been quoted at 35% and 40%. Polymarket says 12%. The single most important input is that those bank numbers are next-12-months forecasts while the Polymarket contract resolves by December 31, 2026 — roughly five months from July 27, 2026. This Deep Dive walks the window adjustment, the resolution mechanics that push the number lower still, the internal inconsistency between two Polymarket contracts, and the four signals that would break the read.
The Bet at a Glance
• Contract: “US recession by end of 2026?” on Polymarket — YES trading at 12¢, roughly $2 million volume, about $20,700 liquidity (Polymarket market page, July 27, 2026)
• Market-implied probability: 12% for YES; 88% for NO
• Our true-probability estimate: 8–10% on a window-adjusted basis
• Edge: approximately 2–4pp on the NO side — small, and inside the noise band for a market this thin
• Cross-check: Polymarket’s separate “Negative GDP growth in 2026?” contract trades at 7%, and “US economic state at end of 2026?” prices a soft landing at 60%
• Anchor forecast: Goldman Sachs put 12-month US recession probability at 20% on January 15, 2026, raising it to 30% in mid-March on the oil shock
• Primary disconfirmation: two consecutive months of negative payrolls, which would invalidate the window math entirely
Methodology
The estimate is built by taking published 12-month recession probabilities from named institutional forecasters, adjusting them to the contract’s actual five-month resolution window on a roughly uniform hazard basis, then applying a further discount for the resolution mechanics described below. Inputs are Goldman Sachs’ published 2026 forecasts (January 15 and mid-March 2026), J.P. Morgan and EY-Parthenon figures as quoted in trade coverage, the live Polymarket market page snapshot of July 27, 2026, and two adjacent Polymarket contracts used as internal consistency checks.
Three caveats matter. First, a uniform hazard assumption is a simplification — recession risk is not evenly distributed across months, and if the labour market is deteriorating the back half of the window carries more weight. Second, the bank forecasts are not all dated identically, so blending them introduces staleness. Third, this market is thin: about $20,700 of liquidity against $2 million of cumulative volume means the quoted price can move several cents on modest size, and any edge under roughly 5pp should be treated as inside the noise.
The market and the price
Is YES at 12¢ on Polymarket’s US recession market value? Not on our read. Prediction-market prices in cents translate directly to probability: 12¢ implies a 12% chance the contract resolves YES. The apparent edge comes from comparing that against Goldman’s 20–30% or J.P. Morgan’s 35%, which suggests YES is underpriced by anywhere from 8 to 23 percentage points. But those are twelve-month figures. The Polymarket contract has roughly five months to run from July 27, 2026. Scaling a 20–30% annual probability to a five-month window on a uniform hazard basis produces a range of about 8.3% to 12.5%. The market’s 12¢ sits at the top of that band, not below it. Before any adjustment for how the contract actually resolves, the honest conclusion is that Polymarket is priced fairly to slightly rich — and the retail instinct that it is “obviously too low” is an artefact of comparing two different time horizons.
| Source | Probability | Window | Date | Five-month equivalent |
|---|---|---|---|---|
| Polymarket YES | 12% | To Dec 31, 2026 | July 27, 2026 | 12% (as quoted) |
| Goldman Sachs | 20% | Next 12 months | January 15, 2026 | ≈8.3% |
| Goldman Sachs (revised) | 30% | Next 12 months | Mid-March 2026 | ≈12.5% |
| J.P. Morgan | 35% | Next 12 months | Earlier 2026 | ≈14.6% |
| EY-Parthenon | 40% | Next 12 months | Earlier 2026 | ≈16.7% |
| Polymarket: negative GDP growth 2026 | 7% | Calendar 2026 | July 27, 2026 | 7% (as quoted) |
Sources: Polymarket market pages, July 27, 2026; Goldman Sachs published forecasts; J.P. Morgan and EY-Parthenon figures as quoted in trade coverage. Five-month equivalents use a uniform hazard assumption and are indicative only.
Why the resolution mechanics push the number lower
The window adjustment closes most of the gap. The resolution mechanics close the rest.
A recession contract resolving “by end of 2026” requires the recession to be identifiable within the contract period, not merely to have begun. The National Bureau of Economic Research, which dates US business cycles, declares peaks with a substantial lag — historically many months after the fact. A downturn that begins in October or November 2026 would very likely not be recognised in time. Even under a mechanical two-consecutive-quarters-of-negative-GDP reading, the second estimate for Q4 2026 GDP would not exist until well into 2027.
That means the effective window is shorter than five months. A recession would essentially need to have started already, or to start imminently, to resolve this contract YES. On that basis a true probability in the 8–10% range is defensible, and 12¢ looks marginally rich rather than cheap.
The internal inconsistency worth noting
Polymarket’s own board contains a useful cross-check. The separate “Negative GDP growth in 2026?” contract trades at 7%, five points below the headline recession market at 12%. Those two contracts should be close cousins — most plausible paths to a 2026 US recession involve negative GDP growth in 2026.
A five-point spread between them is not necessarily an arbitrage; the definitions differ, and the GDP contract is far thinner at about $31,500 of volume. But it is evidence that the 12% figure carries a premium relative to the mechanical measure sitting next to it — which is consistent with the read that YES is the slightly expensive side. Meanwhile the “US economic state at end of 2026?” market prices a soft landing at 60%, which is the same picture from the other direction.
What the named forecasters actually say
Goldman’s published position is not a recession call. Its January 2026 outlook put full-year 2026 GDP growth at 2.8% and Q4 year-over-year growth at 2.5% against a 2.1% consensus, with unemployment stabilising at 4.5%.
“Our strongest conviction views for 2026 are our above-consensus GDP growth forecast.”
— David Mericle, Chief US Economist, Goldman Sachs (Goldman Sachs, January 15, 2026)
Mericle also noted that “tax cuts, real wage gains, and rising wealth should sustain solid consumer spending growth.” The bank’s own labour figures cut the other way, however, and this is the strongest part of the bear case: Goldman put underlying trend job growth at 11,000 per month against a breakeven rate of fewer than 70,000 jobs needed to hold unemployment steady. A labour market adding a fraction of its breakeven rate is not a robust one, and it is why the market’s 88% NO price should not be mistaken for safety.
The bank forecasts have also moved a great deal within 2026 — from 20% in January to 30% in mid-March as oil prices rose — which is a reminder that these are opinions with wide error bars, not measurements.
“We no longer see a U.S. recession, but expect material headwinds to keep growth weak through the rest of this year.”
— Joseph Lupton, Co-head of Economic Research, J.P. Morgan (J.P. Morgan Research, May 27, 2025)
That Lupton comment is dated May 2025 and is included as historical context on how quickly the house view has shifted, not as a current reading. The direction of travel matters more than any single print: the institutional consensus has oscillated between “no recession” and 40% within eighteen months.
The sentiment gap: what the crowd expects versus what the market prices
There is a striking divergence between public expectation and market pricing, and it is worth naming because it is the most likely source of one-sided flow into this contract.
Social discussion of a 2026 US downturn is dominated not by probability analysis but by housing affordability, and much of it is openly hoping for a downturn. A single finance video on the theme drew over 2.9 million views across recent weeks, with the highest-voted comments reading “Me waiting for this collapse so I can finally purchase” (55,666 likes) and “I wish a 79k dollar starter house wasnt 300k” (16,364 likes).
That is an enormous reservoir of people who expect and want a recession, set against a market pricing 12%. Where retail sentiment is this lopsided and this emotional, the systematic pressure on a YES contract is upward, not downward — which is a further reason to think 12¢ contains a sentiment premium rather than a discount. Our analysis of the 2026 midterms market found a similar dynamic, where strongly held political expectation sat well ahead of what the contract price supported.
Where this bet breaks
Four observable signals would invalidate the read that NO is the marginally better side:
- Two consecutive months of negative non-farm payrolls. With Goldman’s trend job growth at 11,000 per month against a sub-70,000 breakeven, the labour market has almost no cushion. Two negative prints would move the recession start date into the contract window and break the hazard math outright.
- A Q3 2026 GDP print below zero. That would put a mechanical two-quarter reading within reach before year-end and materially raise the chance the contract resolves in time, regardless of NBER’s lag.
- The price moving through 20¢ on rising volume. A thin market with about $20,700 of liquidity can be pushed by modest flow, but a sustained move past 20¢ on genuinely higher volume would indicate informed money rather than sentiment, and the read should be re-estimated rather than defended.
- The two Polymarket contracts converging upward. If “Negative GDP growth in 2026?” rises from 7% toward the headline 12%, the internal inconsistency argument disappears and with it one of the supports for the NO lean.
TL;DR
Polymarket prices “US recession by end of 2026?” at 12¢, against bank forecasts of 20–40%. That gap is mostly an illusion: those are twelve-month numbers and this contract has about five months to run, which scales them to roughly 8–17%. Add NBER’s declaration lag — a downturn beginning late in 2026 would not be recognised in time — and a defensible true estimate is 8–10%. The value therefore sits with NO, not YES, but the edge is only about 2–4pp and the market carries just $20,700 of liquidity, so it is inside the noise. The bet breaks on two consecutive negative payroll months.
FAQ
Q: What are the odds of a US recession in 2026 on Polymarket?
A: The “US recession by end of 2026?” contract traded at 12¢ on July 27, 2026, implying a 12% probability, on roughly $2 million of cumulative volume. A separate contract on negative GDP growth in 2026 traded at 7%.
Q: Why do banks say 20–40% when the market says 12%?
A: Because they are measuring different windows. Goldman Sachs, J.P. Morgan and EY-Parthenon publish next-12-months probabilities. The Polymarket contract resolves by December 31, 2026, about five months out from late July. Scaling the bank figures to that window gives roughly 8–17%, which brackets the market price.
Q: Is YES at 12¢ value?
A: On our estimate, no. Window-adjusted and discounted for the NBER declaration lag, a defensible true probability is 8–10%, which puts the value on NO. The edge is small — about 2–4 percentage points — and the market is thin enough that this is within the noise band.
Q: What would make the recession contract cheap?
A: Two consecutive months of negative payrolls, or a negative Q3 2026 GDP print. Goldman’s own figures put trend job growth at 11,000 per month against a breakeven of fewer than 70,000, so the labour market has very little margin before the picture changes.
Q: Does NBER’s timing really matter for this contract?
A: Materially. NBER dates US business-cycle peaks with a long lag, often many months. A recession that begins in the final quarter of 2026 would very likely not be recognised before the contract’s resolution date, which shortens the effective window well below five months.
For related prediction-market analysis, see our Polymarket Fed decision breakdown and our Senate 2026 contract analysis.
This article is informational analysis only and is not betting or financial advice. Odds and prediction-market prices move constantly; every price quoted is a timestamped snapshot, not a live line. There is no such thing as a guaranteed bet — past results and model estimates do not guarantee outcomes. Do your own research.
Gambling carries financial risk and can be addictive. 18+ / 21+ depending on jurisdiction. Never bet more than you can afford to lose. If you or someone you know needs help, visit GamCare (UK), call 1-800-GAMBLER (US), or see our Responsible Gambling page.
