Backtested with ES, but trade with MES

Desktop SDK - Strategies
1

Hi members,

I have created a strategy with ES as the instrument. However, as I have only a small capital, I need to trade with MES. When I tried to backtest the strategy with MES, the result was worse than the one with ES.

Given the situation, may I please seek your advice on the execution?

A few plans have popped up in my mind, from most preferred to least preferred:

1. Follow ES rules, and trade MES with market order: the entry/exit rules still follow the strategy on ES. When the entry/exit orders have been triggered on ES, there would be market orders immediately sent to MES for execution. In this way, the impact of the price difference between MES & ES would be minimized. I still haven’t figured out the coding part yet.

2. Follow ES rules, and trade MES with ES price: the entry/exit rules also follow strategy on ES. The entry/exit orders would be directly executed by MES (i.e. add a data series on MES and place all orders there). One big downside is that the price difference between MES and ES could sometimes cause some orders to fail to be triggered. For example, the strategy may place a buy-stop order at $1000. ES may hit the price (as the price is generated by the entry rules from ES), yet MES does not. Hence, the result would be different from my backtest performance.

Happy to know your thoughts! Thank you.

2

  • Generally ES and MES execute the same in the live market

  • so you can simply mirror your trades to MES from ES with a trade copier that allows different pairs etc…

  • or add in the code to do that…multi inst strategy exec the trades etc - do both and compare back… price entry and exit etc ES and MES both trade in parallel ES is the leader - MES follower

  • or try in the live paper trading ES and MES side by side for 1 week at least… to get a feel for it… 1 system each

  • backtest - timeseries can have a bid effect

  • a small basket of instruments can assist also…

  • replay test both of them

  • paper trade walk forwards

  • then reconsider a game plan

note in a high frequency system the stats can be very different from, sim to live like 80% the same - in a live low freq it should be around 95%

possible handy links here: