However, there are many concerns to be addressed in backtesting:
- A backtest can never exactly replicate the performance of an investment strategy in an actual trading environment.
- The quality of the historical data is questionable, since it is subjected to outliers by third-party data vendors.
- Look-ahead bias takes many forms. For example, listed companies may split, merge, or de-list, resulting in substantial changes to its stock price.
- For strategies based on information from the order book, the market microstructure is extremely difficult to simulate realistically, since it represents the collective visible supply and demand in continuous time. This supply and demand are in turn affected by news events around the ...