Fractional Differencing for Stationarity - Quant Researcher Interview Question

Problem Description

Fractional differencing is a technique used in quantitative finance to transform non-stationary time series into stationary ones while preserving memory, unlike standard integer differencing which erases long-term dependencies. By applying a real-valued differencing order $d$, this method balances the trade-off between stationarity and information retention, making it crucial for feature engineering in machine learning models. Task Implement a function solution(prices, d, threshold) that perfor