Tracking Error (ρ / NRMSE)
Education only · our voice · free public data
Definition
Two numbers that measure how faithfully a free proxy series reproduces a reference series: ρ (shape/co-movement) and NRMSE (size of the gap).
How to read it
When we rebuild a paid indicator from free sources, we need to know how good the copy is. Two complementary metrics answer that. ρ (rho, correlation) measures whether the proxy moves in the same direction and rhythm as the original — a value near 1.0 means they rise and fall together. NRMSE (normalized root-mean-square error) measures how far apart the two lines sit on average, scaled so you can compare across series — lower is better, and 0 means a perfect match. You want high ρ AND low NRMSE: high ρ alone can hide a consistent level offset, while low NRMSE alone can hide timing mismatches. Together they tell you whether a Class-A reconstruction is trustworthy enough to replace a vendor feed.
How practitioners use it
Used as context among multiple indicators — never as a standalone signal to act.
Less common professional uses
Power move: if ρ is high but NRMSE is elevated, look for a fixable constant bias (a scale or offset) rather than discarding the proxy. Caveat: tracking error is measured over a historical window — a proxy that tracked well in calm markets can decouple in a shock. Beware small-sample ρ: a short overlap window can flatter or punish a proxy misleadingly.
Sources & provenance
Portal reconstruction methodology
This page is educational content published by Pachira Aquatica Global LLC. It is not investment advice and not a recommendation.