Autocorrelation Lag-1 (20 bars)Cross-asset & regime

Timeframe M1 · unit ratio

What it measures

Measures the statistical correlation between each bar's return and the previous bar's return over 20 bars. A positive value means consecutive moves tend to continue in the same direction (momentum); a negative value means they tend to reverse.

How Janira reads it (bullish vs bearish)

Range -1 to +1. Positive values (especially > 0.2) indicate that up bars tend to be followed by up bars and down bars by down bars - momentum-like serial dependence. Negative values (below -0.2) indicate alternating behavior - each move tends to partially reverse the next. Values near 0 mean returns are serially independent, consistent with random walk behavior.

In plain language

If a ball rolling right tends to keep rolling right on the next bounce, that's positive autocorrelation. If it tends to bounce back left after each roll, that's negative autocorrelation. This measures which pattern the last 20 bars followed.

Scenarios

• Autocorr > 0.3: Serial momentum; consecutive returns are positively correlated; trending behavior dominates at the bar-to-bar level.
• Autocorr -0.2 to +0.2: Near-random serial structure; returns show little memory from bar to bar; no exploitable autocorrelation pattern.
• Autocorr < -0.3: Serial mean-reversion; each bar tends to partially reverse the previous one; oscillating microstructure.
• Autocorr transitioning from positive to negative: Momentum character is giving way to mean-reverting character at the micro level.

More cross-asset & regime indicators

Janira computes Autocorrelation Lag-1 (20 bars) deterministically from live price action, the same way for every reading - no discretion, no hidden weighting. This page explains the method; it is not a live reading and not advice.