TSMC reports revenue monthly. April (NT$410.73B) and May (NT$416.98B) are already filed — NT$827.71B of the quarter is public. We lock Q2-2026 consolidated revenue at ≥ NT$1.23 trillion: it needs only ~NT$402B in June, below May's run-rate. This is the Prediction Series' core method — sum what's already public, lock the number the market is still 'waiting' for.
The lock needs only ~NT$402B in June — below May's NT$416.98B run-rate, against a +30% YTD trend. Two of three months are already filed; only June remains, and it files ~July 10, before the July 16 result.
We lock TSMC's reported Q2-2026 consolidated revenue at ≥ NT$1,230 billion. With April and May already public, this is a statement about one remaining month, not three.
| Month | Revenue | Status |
|---|---|---|
| April 2026 | 410.73 | filed (+17.5% YoY) |
| May 2026 | 416.98 | filed (+30.1% YoY) |
| June 2026 (needed for lock) | ≥ ~402 | files ~Jul 10 |
| Q2 total (locked floor) | ≥ 1,230 | resolves Jul 16 |
TSMC's January–May 2026 cumulative revenue is NT$1.9618 trillion, up ~30% YoY, on sustained AI-accelerator demand. April printed NT$410.73B (+17.5% YoY) and May a record NT$416.98B (+30.1% YoY). For the Q2 lock to fail, June would have to fall below ~NT$402B — a sequential drop the current order book does not support.
Sources: TSMC April & May 2026 monthly revenue reports (pr.tsmc.com / investor.tsmc.com); SEC EDGAR 6-K (CIK 1046179); verified 2026-06-20.
Sell-side notes still frame Q2 as forthcoming, but two-thirds of it is on the public record. Our edge is simply refusing to treat a near-determined sum as uncertain. The honest residual risk lives entirely in the single unfiled month — and we sized the floor below the established run-rate to absorb it.
The P-05 method: build the quarterly KPI bottom-up from the issuer's own monthly filings, set the lock threshold below the run-rate so a single soft month can't break it, and score against the single reported integer. No FX modelling beyond TSMC's own reporting.
A miss requires June revenue below ~NT$402B — e.g. a sharp NT$ appreciation, an earthquake/fab disruption, or an abrupt AI-order air-pocket in a single month. Each is low-probability against a +30% YTD book; together they leave ~7% of risk, which is our honest confidence gap.
RAOSCAFF locks P-08 on 2026-06-20. Two of the three component months are already public; the June filing (~July 10) completes the deterministic input before TSMC's July 16 Q2 results. We score the ≥ NT$1.23T threshold against the single reported consolidated-revenue figure.
The component months are public now — anyone can check the arithmetic before the lock resolves.