How to determine if strategy entry criteria beats random entry

[DeltaRisk]

Can I ask you a question:

How do I run Monte Carlo using Excel?

https://support.office.com/en-us/ar...mulation-64c0ba99-752a-4fa8-bbd3-4450d8db16f1

Here's a google. It might help.

I have adjusted programs, not Excel.

 

[ironchef]

https://support.office.com/en-us/ar...mulation-64c0ba99-752a-4fa8-bbd3-4450d8db16f1

Here's a google. It might help.

I have adjusted programs, not Excel.

Thank you for your help. I am studying it now to see if I can understand what it is and if I can use it.

Regards,

 

[DeltaRisk]

Thank you for your help. I am studying it now to see if I can understand what it is and if I can use it.

Regards,

Margin(haircut) is much easier to figure out even with simulations.

MonteCarlo will take a few days to figure out. I think I might know of one program that costs like $20 you could run.
I'll have to dig it up, but it can help.

Have you ever considered going professional? There are tools available to professionals that retail customers cannot access.

 

[jk90029]

One idea I had was I could randomly pick entries in the stocks that had signals and use the same exit criteria. For example, if stock ABC had 5 signals, I would randomly pick 5 entries in ABC and continue for the rest of the stocks. I could run this simulation 100 times and see how many of those runs my original strategy outperformed. If it beats more than 90 or 95 of those random entry runs, then I likely have an edge.

Are there other approaches that are more robust?

Mostly I agree. However you should outperform the 1)expense(commission) + 2)index change(not just breaking even account) + 3)tax rate (yearly 1040).

Probably the difficulty comes from 1)and 3) is easy.

In order to follow just 2) to track the index without outperforming, then simply buy SPY and hold long time.

 

[ironchef]

https://support.office.com/en-us/ar...mulation-64c0ba99-752a-4fa8-bbd3-4450d8db16f1

Here's a google. It might help.

I have adjusted programs, not Excel.

I studied the article last night. Yes, it will do Monte-Carlo using a random number generator. The problem is they did not discuss how to make large number of runs. A typical financial simulation requires 100,000 of runs.

Anyway, it gives me something to work on.

Thanks again.

 

[ironchef]

Have you ever considered going professional? There are tools available to professionals that retail customers cannot access.

No. I don't have the background training nor the smart to be a professional.

Regards,

 

[fan27]

Mostly I agree. However you should outperform the 1)expense(commission) + 2)index change(not just breaking even account) + 3)tax rate (yearly 1040).

Probably the difficulty comes from 1)and 3) is easy.

In order to follow just 2) to track the index without outperforming, then simply buy SPY and hold long time.

1 and 3 make sense. I will add those in to my simulation. But I don't think a strategy should necessarily be required to outperform an index. For example, say I have a strategy that is only in the market 15% of the trading days in a year. It's absolute returns may be lower than an indesx but it allows me 85% of the trading days to allocate my trading capital to something else. I think a better comparison would be to compare the sharpe ratios of the strategy vs index but add to the equation somehow "time in market".

 

[fan27]

You must be a programming wizard!

How are you able to test so many different strategies and combinations? What programming/coding are you using? I use Excel/VBA and it takes a lot of effort for me to test a strategy each time. My current homework assignment is testing the random entry strategy to see if it indeed beats TA for the equities I am trading.

Regards,

Hardly a programming wizard, though I have been writing software for my day job for the past 16 years.

For the past 10 months I have be writing a machine learning backtesting platform in GoLang. I hope to make it commercially available within six months. The idea for the service is it will be able to do all the heavy lifting in terms of finding profitable trading strategies but the user will have to tell it what features to test (indicators, price, events, etc). The user will then be able to generate platform specific code (EasyLanguage, NinjaScript, ThinkScript, etc) for the strategies they are interested in.

fan27