3 – M2L1 02 Quant Workflow V3

A hypothesis that forms the basis of a trading strategy must survive several phases of increasingly rigorous testing. This process can be different at different companies, banks, or hedge funds, but usually, there is an initial exploratory research phase during which ideas are generated and validated in a fairly basic way. The goal here is to come up with promising ideas relatively quickly. In the research phase, you will focus on determining what set of positions to enter, on which assets at which times. Such that you have the potential to get positive future returns. Later on, you’ll incorporate other factors critical to a full-fledged trading strategy. How much money to invest in each asset, under what conditions to exit positions, factoring in the costs of making trades, and what risks constraints to impose? The process of rigorously simulating the entire flow in an automated fashion using historical data is called Backtesting. It is important to resist the temptation to jump right into full-fledged backtesting. We’ll talk about the reasons for this when we discuss overfitting in depth.