D.Terminal  /  Correlation & rotation
Correlation & rotation

Find out what is really one position

Portfolios drift into concentration quietly. Six names that felt diversified in January can be one trade by March. The correlation matrix, the normalised performance chart and the heatmap answer three versions of the same question: what is actually moving together right now?

LiveRolling matrix
AnyBasket, normalised
Lead/lagVisible on one axis
Crypto+TradFiSame view
The correlation matrix over a chosen rolling window.
The correlation matrix over a chosen rolling window.

The correlation matrix

A live matrix over a rolling window you choose. Three things it is genuinely good for:

Normalised performance

Put any basket on the same axis from a common starting point and watch the lines separate. This is the fastest way to distinguish the name actually leading a sector move from the three following it a day late — a distinction that is invisible on separate charts with separate scales.

Normalised performance chart with multiple tickers from a common start
Any basket, normalised from a common start — lead and lag become obvious.

Rotation as shape

The heatmap sizes every instrument by volume or open interest, crypto and tradfi together. Where the matrix tells you what moves together and the normalised chart tells you what leads, the heatmap tells you where the money went — usually before a watchlist notices.

Choosing a window

Correlation is entirely a function of the window you measure it over, and a matrix with a fixed lookback quietly makes that choice for you. The window here is yours to set, which means a 7-day and a 90-day view of the same basket can be compared — and when they disagree, that disagreement is the information.

Questions & answers

Yes. The rolling window is configurable, which matters because correlation is a function of the period measured — a short and a long window disagreeing is itself a signal.

Yes. The heatmap and the normalised performance chart both cover crypto and tradfi instruments in the same view.

Putting an arbitrary basket on one axis from a shared starting point, so lead-lag relationships that are invisible across separately scaled charts become obvious.

Three days, no card.

Full terminal, every exchange, every panel. Sign in with Discord or an email address and start immediately.