NYSE TICK: Session Bias & Extreme Prints
The TICK index — NYSE stocks upticking minus downticking — summarized per session since 1997: the day's bias, its ±1000 extreme prints, and an era-proof five-state read with the forward-return base rates behind each state. The base rates are contrarian: washouts bounced, euphoria went flat.
Today's reading
As of the August 21, 2026 close, the NYSE TICK session read Balanced: average bias -1 (z -0.1 vs the trailing year), with no ±1000 extreme prints. Session extremes: +943 high, -444 low. Since 1998, balanced sessions preceded +0.63% average SPY returns over the next 21 sessions vs +0.70% for all sessions.
Sources, methodology & freshnessLast updated 2026-08-21 · Open ↓Close ↑
Since 1998, balanced sessions preceded +0.63% avg SPY returns over the next 21 sessions (64% positive) vs +0.70% baseline — 2,747 cases. That is historical context. It does not forecast the next move.
Intraday tape — the last sessions, 5-minute
The raw tape behind the daily aggregates: every regular-hours 5-minute bar of $TICK, plus the session-cumulative running total day traders watch for one-sided program days. Pick how many sessions to show; the data refreshes with the daily update after each close.
Session bias vs its own era
Each session's bias (blue) against its rolling one-year mean and ±1.5σ band (gray). The band IS the era adjustment — what counts as a washout or euphoric session moves with the market's structure, which is why raw TICK levels from different decades can't be compared directly.
Cumulative TICK — trend & divergence
The running sum of each session's bias, rebased to zero at the left edge of the window. The trend is the read: a rising line confirms the tape beneath a rally; price making new highs while cumulative TICK makes lower highs means fewer stocks are ticking along — the classic divergence warning. On the Max window the line mostly reflects TICK's structural era drift (strongly positive 2001–2007). Use the shorter windows for the trading read.
What followed each state — full history
Forward SPY returns from every session in each z-score state since 1998. The shape is the familiar contrarian U: washout sessions preceded the best average returns and euphoric sessions the worst — panic prints marked exhaustion, not acceleration.
| State | % of days | Next 5 sessions | Next 10 sessions | Next 21 sessions | N |
|---|---|---|---|---|---|
| Washout | 8% | +0.28% · 58%↑ | +0.48% · 59%↑ | +1.05% · 63%↑ | 566 |
| Bearish tilt | 21% | +0.31% · 59%↑ | +0.43% · 60%↑ | +0.71% · 63%↑ | 1,461 |
| Balancednow | 39% | +0.10% · 56%↑ | +0.29% · 60%↑ | +0.63% · 64%↑ | 2,747 |
| Bullish tilt | 27% | +0.15% · 58%↑ | +0.34% · 61%↑ | +0.77% · 65%↑ | 1,945 |
| Euphoric | 5% | -0.01% · 58%↑ | +0.03% · 62%↑ | +0.23% · 60%↑ | 372 |
| All sessions (baseline) | 100% | +0.17% · 57%↑ | +0.34% · 60%↑ | +0.70% · 64%↑ | 7,091 |
Forward returns on SPY closes. States use the rolling one-year z-score of the session bias, so they are era-comparable. Overlapping windows — descriptive tendencies, not signals.
Extreme-print days — modern era (2010+)
The sharper cut: days when ±1000 prints actually hit the tape, in the structurally stable 2010+ era. Heavy washout days (10+ −1000 prints) preceded roughly 2.5x the baseline return over the following week; surge-heavy days were flat short-term.
| State | % of days | Next 5 sessions | Next 10 sessions | Next 21 sessions | N |
|---|---|---|---|---|---|
| Washout day (5+ −1000 prints) | 13% | +0.39% · 61%↑ | +0.56% · 61%↑ | +1.40% · 66%↑ | 528 |
| Heavy washout (10+ −1000 prints) | 4% | +0.65% · 62%↑ | +0.55% · 59%↑ | +1.60% · 67%↑ | 181 |
| Surge day (5+ +1000 prints) | 8% | +0.04% · 53%↑ | +0.29% · 60%↑ | +1.29% · 65%↑ | 328 |
| No extreme prints | 39% | +0.20% · 60%↑ | +0.37% · 64%↑ | +0.74% · 67%↑ | 1,617 |
| All 2010+ sessions (baseline) | 100% | +0.25% · 60%↑ | +0.51% · 64%↑ | +1.06% · 68%↑ | 4,197 |
±1000 counts are 5-minute bars whose high/low reached the threshold. Restricted to 2010+ where the print-frequency regime is stable. Overlapping windows — base rates, not signals.
Since 1998, balanced sessions preceded +0.63% avg SPY returns over the next 21 sessions (64% positive) vs +0.70% baseline — 2,747 cases. That is historical context. It does not forecast the next move.
How NYSE TICK Works
- 1What TICK isThe NYSE TICK index is the number of NYSE stocks whose last trade was an uptick minus those on a downtick, printed continuously through the session. Readings near zero are two-way trade; big positive or negative prints mean program-driven, market-wide buying or selling hitting hundreds of stocks at once.
- 2Aggregate each sessionFrom 5-minute TICK bars (1997–present) we compute each session's BIAS — the average TICK reading, i.e. which way the tape leaned all day — and its EXTREME PRINTS: bars whose high reached +1000 (surge prints) or whose low reached −1000 (washout prints).
- 3Make it comparable across erasTICK's distribution has drifted hugely over three decades — the pre-2001 bias was persistently negative, 2001–2007 strongly positive, the modern era mild. So the state uses a rolling 252-session z-score of the bias: today versus the last year of itself, not versus 1998.
- 4Classify the state — and attach base ratesFive states by z: Washout (≤ −1.5), Bearish tilt, Balanced, Bullish tilt, Euphoric (≥ +1.5). For each, we show what SPY actually did over the following 5, 10 and 21 sessions — plus a second study on extreme-print days in the stable 2010+ era. Both are contrarian.