Correlation Coefficient

The Correlation Coefficient measures how consistently the chart symbol moves with a second, user-selected symbol, plotting the Pearson correlation over a rolling window as an oscillator bounded between -1 and +1. A reading near +1 says the two series rise and fall together; near -1, they move opposite ways; near zero, no stable linear relationship exists.

This build is the standard raw calculation: the chart symbol is compared against a comparison symbol input (the S&P 500 by default) over a 20-bar window, with an option to correlate log returns instead of raw prices. The Correlation line is drawn with a soft area fill, framed by dotted levels at +1, zero, and -1.

How to Trade the Correlation Coefficient?

• Near +1: the two symbols move in near-lockstep; holding both adds exposure, not diversification.
• Near -1: consistent opposite movement — the profile a hedge is supposed to have.
• Around zero: no dependable relationship; treat the pair as independent.
• Zero-line crosses: the sign has flipped; alert conditions cover crosses above and below zero.
• Drift from an extreme toward zero: a once-tight relationship is decoupling.

Correlation frames trades rather than triggering them: use it to size, filter, or veto ideas generated elsewhere.

Correlation Coefficient Settings

• Symbol (default SP:SPX): the instrument the chart symbol is measured against.
• Length (default 20): bars in the rolling window. Shorter reacts fast but whipsaws; longer is steadier but slow to register a real break.
• Source (default close): the price series used for both symbols.
• Use Log Returns (default off): correlates bar-to-bar log returns instead of raw values, stripping the trend effect that inflates raw-price readings.

Frequently Asked Questions

Should I correlate raw prices or log returns?

Log returns, for most decisions. Two instruments in long uptrends can post high raw-price correlation regardless of how their individual bars behave; Use Log Returns measures how the bars themselves move together instead.

How is the Correlation Coefficient different from beta?

Correlation grades the consistency of co-movement on a fixed -1 to +1 scale, but says nothing about size. Beta scales it: a stock can track an index tightly while moving twice as far on every swing — correlation captures the first fact and misses the second.

Does a high correlation make two symbols a good pairs trade?

Not by itself. Tightly correlated instruments can still drift apart indefinitely, so pairs traders use the coefficient as a first screen and test the spread separately. And the rolling 20-bar sample can swing hard — confirm on a longer Length before trusting it.