Statistical Inference

Inference turns noisy market data into defensible claims. Estimation pins down parameters like mean return and volatility while quantifying their uncertainty; hypothesis testing asks whether a backtested edge is real or luck; regression relates one series to others and becomes the engine behind factor models; and stationarity — whether a series' distribution holds still over time — decides which of these tools even apply to a time series. Financial data violate textbook assumptions constantly: overlapping returns, changing regimes, and heavy tails. Learning to check assumptions before trusting a p-value is the core habit, and it rests on the probability language of the previous node.