RNOW · live · P(recession now)
–%
The probability the US is in an NBER recession right now — the hidden present, before NBER announces it.
A real-time measure of US recession risk. Updating continuously from public data and Kalshi’s prediction markets, RNOW reads the core economic data and the Kalshi signal to answer a simple question: are we in a recession right now — before the NBER says so?
The recession-nowcast probability we publish — over the last 30 days, updating every 5 minutes off the live feeds.
How much each signal counts toward the published RNOW — % from live Kalshi markets, the rest the core model and financial data with no market equivalent.
RNOW reads several live Kalshi markets, each wired into the block it informs — the recession market (markets block + forward R12), the GDP-growth market (output), and the unemployment market (labor). These are the strikes where price discovery is happening right now.
Most of the index is composed from live Kalshi markets, each wired into the block it bears on: the GDP-growth market sets output, the jobs and unemployment markets set labor (the latter framed against the Sahm trigger), and the recession market feeds the markets block. They reprice every day the books trade.
The rest is the part with no Kalshi equivalent: a statistical model of the core monthly data (jobs, output, income, sales) and the financial block — yield curve and credit spreads. It's what keeps RNOW a real nowcast rather than a pure market aggregator. An unemployment-based trigger sets a hard floor.
The Kalshi recession market asks a slightly different question — a recession sometime this year. We translate it onto our exact question (are we in one right now) using a mapping fit on history — so we lean on the direct GDP and jobs markets, not the bridged one.
Kalshi-led and rolling: we blend this year's and next year's recession markets (KXRECSSNBER) by how much of the next twelve months falls in each, so the horizon stays a true ~12 months out instead of shrinking as the calendar year runs down. Weighted up to 60% by how liquid the book is, and cross-checked against the classic yield-curve indicator. Separate from the "right now" reading above.
Everyone wants to know if we’re in a recession — and nobody finds out until it’s official. The NBER dates recessions six to eighteen months after they begin. In between, people reach for proxies: a single jobs report, a yield-curve headline, one scary GDP print. Each captures a sliver; none tells you the state of the economy right now.
So the most important macro question — are we in it? — gets answered late, and in fragments.
RNOW answers it continuously. It reads the core economic data through a statistical model, blends in financial, labor and output signals and the prediction markets, and prints one probability that updates as the data does. It tracks the turn as it happens, not after.
NBER announces recessions 6–18 months late. RNOW nowcasts the present from public data.