Thread by @sourabhsiso19

@SunderjiJB My philosophy is :

Idea --> Rules --> Backtest --> Discard or Deploy --> Optimize --> Scale

https://twitter.com/SunderjiJB/status/1226892925939748864

1) Idea : It comes from market observation.
For eg : Is it profitable to buy stocks breaking previous day high within 15 min of market open.

2) Rule : Anything that can be said to a computer in a simple language
Buy if 15 min close > Previous day high

(1)

3) Backtest :
Test the same rule on historical data to see if our system has any positive edge

4) Discard or Deploy

If the rule is showing any positive expectancy then we can deploy it else discard and look for testing another set of rules/idea

(2)

5) Optimise:

If the backtested strategy is profitable, then we can try to optmise.
Optimisation is basically improving the strategy:
1) Can I avoid big losses ?
2) What are unfavourable condn’s for my strategy ?
3) Do I make losses intraday if I enter after 11 am etc

(3)

6) Scale

After optimisation, we should see that whether our strategy is scalable or not.
1) Can I deploy more than 100 crores ?
2) What would be the slippage ? etc

One can use stress test and liquidity test to check for scalability

#backtest #system #quant #trading