AI Quant Base

Total change

For positive starting equity E₀ and ending equity Eₙ, total change is Eₙ / E₀ − 1. It measures the endpoint change in a comparable, appropriately adjusted sequence.

Total change (%) = (E[n] / E[0] − 1) × 100
Days = (last UTC date − first UTC date) / 86,400,000
CAGR (%) = ((E[n] / E[0])^(365 / Days) − 1) × 100

Annualization with an explicit convention

CAGR uses actual elapsed calendar days and a fixed 365-day year. It is not the arithmetic mean of returns and does not forecast future performance. The teaching sequence 100 to 121 across 730 days gives a 10% CAGR. Values with magnitude above 1,000,000,000,000% or outside finite numeric range are not displayed.

Peak[t] = max(E[0], …, E[t])
Drawdown[t] (%) = (1 − E[t] / Peak[t]) × 100
Sampled maximum drawdown = max(Drawdown[t])

Sampled drawdown and dates

The tool shows decline as a positive percentage. For the teaching values 100, 120, 90, 121, the peak-to-trough decline from 120 to 90 is 25%. If several troughs tie for the maximum, the first occurrence is recorded. Repeated equal peaks retain the earliest peak date. A non-decreasing sequence has 0% sampled drawdown and no drawdown interval.

What a sampled result leaves out

Unobserved intraperiod losses may be larger. Unequal sampling intervals are permitted and flagged; chart spacing follows actual dates. No interpolation, cost deduction, flow adjustment, benchmark comparison or trade-level statistics are performed.

A reproducible record

Exports include observations, basis declarations and calculation version. JSON reimport recomputes results. This arithmetic check establishes consistency with the supplied sequence, not the provenance or reliability of the underlying data.

Search the library

Search products, workflows, guides and tools.