Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
For this particular system, the only quit point I am considering is the single contract drawdown. This is simple, and pretty robust. If I am trading the system years from now, with many contracts (my hope, of course), I still have that $5,000 maximum drawdown limit.
In the past, I have used the Monte Carlo simulation results to help me decide when to quit. I have also kicked around, but never implemented, a temporary quit point based on market volatility. When the market gets super crazy, it is best to take a break.
I don't think there is a wrong metric or combination of metrics to use to decide when to quit trading. There is probably no "one size fits all" optimum, either. The key, in my mind, is to select some criteria that you are comfortable with, write it down, and then follow it exactly. Then, if your system fails, there should be no tears. You knew the system could break, and you quit at a pre-ordained spot.
I think where people get in trouble is in not having a "quit point," or their quit point is when their money runs out. Speaking from personal experience in the late 1990's, having to quit trading when your money is gone is not a pleasant way to quit.
1) Running monte carlo simulations for N days of trades, with N varying from 1 to aBigNumber, and taking the desired % cutoff (i.e., "after N trade days 90% of my simulations ended up with an equity above $X"")
That's what seemed tedious to me, and I figured the following was your simple way to get at the same idea:
2) Using the formula: N-trades * AveTradeResult - ZScore * StDevOfTrades * Sqrt(N), where the ZScore corresponds to the %-cutoff you are aiming for.
My confusion is this:
If (1) = no confusion (psych!).
But I thought (2), and can't intuit how that formula gives you a result "close enough" to the result of (1). Can you explain that formula, or wikipedia me to the right spot to do my own digging?
I'm trying to learn just enough statistics to improve the evaluation and monitoring of my strategies, without geeking out unnecessarily. I'll look into the central limit theorem - it can up recently in my self-study but I didn't look too closely.
The plot I show is generated by day by day Monte Carlo analysis. It is not as tedious as it sounds, if you set up the spreadsheet and run a bunch of macros.
You should be able to run the formula in #2 to get pretty close to the same results to the simulation in #1. It won't be exact (there will be a difference, for example, if the Monte Carlo loss limit is hit), but it should be close. Here is a link to another system I ran both the Monte Carlo and the equation method:
Before I continue this discussion, I realize that this short post is probably way too involved for the casual reader. So, I'll summarize this work in a sentence: Before you change or quit a poorly performing trading system, make sure that you aren't …
I honestly don't know where I got that equation, so I can't point you anywhere. As @record100 says, it is probably due to central limit theorem or some other simple statistics tool (Simple is all the statistics I am good at!).
I'm actually a little pleased that I've absorbed enough of your logic to anticipate the day by day Monte Carlo approach, while also a little disappointed that I missed it in your other thread - since I read it from start to finish! I must've not had my head wrapped around what you were doing, so it went in one ear and out the other. Now that I've thought it through for myself the post you linked to makes perfect sense.
I'm working on some Monte Carlo macros in Open Office, and will play around with making the under/over performance tracking curves.
I'm curious, what were you expecting, and why? It would be interesting to hear your take on things.
I can say that using more than the max drawdown as quit criteria was driven in part by the small size of the max historical drawdown. I definitely do not always use max drawdown as my criteria.