Deriv Synthetic Indices: A Research Guide

Synthetic symbols can have different volatility settings and tick behavior. A rule tested on one symbol should not be transferred to another without an independent evaluation.

Measure sample coverage, missing ticks, duplicates, digit distribution, payout, longest adverse sequence, drawdown, execution failures, and stability by time period.

Selecting the top performer from many tested symbols and settings can overfit the past. Reserve untouched holdout data and account for the number of hypotheses tested.