I am interested in subscribing to designer models.
However, I am worried about the robustness testing of the models.
I am new to quant but imagine it isn't that hard to produce a nice looking backtest, for someone experienced in the field.
I was wondering if P123 could display some standard robustness checks alongside all the other standard data displayed about the model?
I don't know what the best standard (and easy to implement from a P123 perspective) robustness checks would be, due to my lack of experience.
Maybe display some of the other important elements too, e.g. commission, slippage?
Or maybe my concern can be addressed another way I don't know.
It would be nicely for example to see a small summary of the behaviour of backtest in 3 or 4 subdivisions of the Universe to see robustness of system plus some kind of rolling back test. Slippage and comissions are fixed by portfolio in designer models, if I recall it well.
OOS results in conjunction with all the rest of standard metrics (risk, profit,...) are nice to be check but all of these are in summarized data.
All interesting ideas. Currently DM marketplace is not generating enough revenue to justify additional investment. Perhaps the designers could help up understand how to increase adoption?
You can always request something from the designer. It is one way to get what you want without the platform having to spend time and resources to make it happen. The designers have an incentive to provide the information anyways. I have replied to similar things in the past. Maybe it is not worth their time to do all the time, but maybe they can have it handy for anyone who asks so it would be a one time thing to prepare. Asking for different start dates and slightly different rebalance timeframes is a good idea. I would also say increasing the portfolio size but some models are meant to pick a small set of opportunities so it might not be widely applicable.