There is a simulator in the Van K Tharpe's book, "Van Tharp's Definitive Guide to Position Sizing."
This simulator is no longer available, but what it did, was allow you to enter trades you have, where you have a list of trades, with pNL for each. The simulator would somewhat SORT all of the trades in various orders, and THEN apply some position sizing schema to it.
You would then set some threshold of DD% to STOP a given simulation run(E.g. 25%) and consider that run a failure... and a target growth%(E.g. 300%), which will consider that run a success.
You could do this 100s to 1000s of times very quickly. What the report gives you... is the % of time the system failed, the % of time it managed to hit the growth target, the Avg profit per run, and other useful metrics.
Does anyone know of a similar type of software? Maybe an excel add-on, or some stand alone software?
Can you help answer these questions from other members on futures io?
That seems a Monte Carlo simulator, there are many on the internet. I think @kevinkdog can help with that, he provides a free simulator in his book "Building winning algorithmic trading systems"; the only issue is that it is not built for position sizing.
The following 2 users say Thank You to soacm for this post:
I can't provide link to it because of self promotion rules, but if you know how to macro code in Excel then adding position sizing is a relatively easy task.
The following user says Thank You to kevinkdog for this post:
So I've created a dashboard in Excel already using pivot tables. This allows me to take a bunch of parameters of a system(due to the way I've organized the data I have), and I can toggle on/off a bunch of different parameters.
Attaching an image, it is messy as I was just building out the basic functionality.
From here, yes it would be easy to apply a position sizing scheme, and show the resulting curves. But that would only show my resulting equity curves for ONE single run.
I am more after running a BUNCH of simulations that would randomize the trades(or some other method), and THEN apply position sizing... and spit out the final results so that I could compare. I would then be able to see how often a given position size failed(E.g. 40% of the time), succeeded, or something inbetween.
I suppose, absolute WORST case, I could potentially build a MACRO, that would do all of the above I described(I'm not particularly good w/excel), and have it iterate 100s-1000s of times, while storing the results of each run somewhere.
But I'm hoping I don't have to build all of that out, if something already exists. 2022-06-06_085751
