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)
I was referring to the monte carlo tab, on large data samples make sure to use the latest beta version of the journal with the new monte carlo engine, as it is a lot faster ( easy migration as described for other recent versions applies).
vvhg
Hic Rhodos, hic salta.
Can you help answer these questions from other members on NexusFi?
I am checking the MC tab often, but perhaps I cannot read it properly.
This is the MC simulation for the first sample posted yesterday. I have a basic understanding, that greener is better:-)
I was working late but maybe it will pay out.
I have tweaked the strategy a bit and performed some more backtesting.
I just run the test on all instruments I have loaded in Ninja and the result looks like this (from September 2011 to now).
Strategy is optimized for E7 but it somehow performs in other markets as well.
Fixed Ratio MM system used for position sizing in the backtest.
How big is the chance of curve-fitted settings in such case? Do you think is this something I can build upon?
The number of trades is a bit misleading: I am using multiple (variable number) targets for one entry.
40k starting equity - optimized for
40k starting equity - another
2k starting equity
The equity curve is a bit bumpy for my liking, but this is perhaps something I can work on...
Two things that caught my eye immediately: first, try to use the beta journal, the new monte carlo is a racehorse...
Second, when the max drawdown is bigger than profit it's really bad news....it should only be a fraction.
Run some MC sims and I'll try to explain them...set them to 500-1000 iterations.
25k start equity,
FixedRatio MM,
MC on 1000 iteration with same equity and sample of 200 trades
This is essentially the same strategy as in previous post, but I restricted the hours a bit so that it runs when the market is moving.
The strategy stability probably needs several simulations at different starting capital levels... I need to work on this.
I will read the other posts but here is something to consider. Not only do markets change in character which may give rise to a change in the underlying strategy, but trying to trade a stragegy that is built for trend following will get large drawdowns during choppy markets.
That is why most black box programs fail. They can't adjust to the different type of trading days. A very narrow range day which we experienced on Wednesday Jan 4th is not conducive to a trend following program.
I believe because of this your results can be better than those from a automated stragegy if you're patient and learn how to trade different market conditions.
Didn't get to it yet. I'll play with it today. By fixed position sizing you mean just one contract? This won't work since the strategy takes into account the traded size and is reducing risk at certain points. There have to be three targets to diversify the risk. One target just doesn't work.
The strategy takes into account the risk and adjusts the position size to it (2% rule each entry).
So I am wondering how to proceed in the MC analysis to have some meaningful results.
The start starts with 3 targets = 3 contracts (or the multiple of it based on total risk per entry) and ends with 3 targets = 10 or more contracts in some situations.
The results of the backrest are essentially money managed and risk managed. This is not just simple bot, that places order at some indicator trigger.
So in which mode do you suggest to run the MC?
-1000 iterations
-100 samples or more (I can easily put more there) - but here is the question whether it should be just random fraction of the entries or all of them
-now the the money management or fixed position size mode?
the previous picture is run with the money management mode.