Trade simulation with historic data: worthwhile?

Beginner Questions
1

I’ve been relatively quiet here but learning a great deal… much thanks to everyone for that… what I am faced with now, perhaps more than ever, is just a sheer lack of personal experience and chart time.

So I am wondering if there is a way to address that. I managed to find a trade simulation EA for MT4, it’s a bit cumbersome (from my unlearned perspective, I’m not used to using anything like it) but it quickly illustrated just how wrong I could be with regard to thinking I understood the market.

To save a long explanation, I tried a trading strategy or two and quickly discovered after a few simulated “weeks” of trading that really, I was my own worst enemy unless I set a sufficient stoploss, for instance. Not a bad lesson to pick up in a couple of hours! This was very important to learn, and perhaps in some minor way, put me a month ahead of where I’d be otherwise.

So… this is my “noob~like” Utopian Vision for improving my skills and instantly becoming a pro (que fantasy music here, and “entering a dream” visual effects)…

First Day: the 1970’s. Having futzed around considerably I start with MT4 just chock full of historic data. I’m ready to go at the computer and I’ve turned the screen monochrome green and black to match the time period. I put Earth, Wind and Fire on in the background, and with Boogie Wonderland playing in the background, I trade, trade, trade and trade some more using my favourite strategies. Eventually the 1970’s exhaust themselves, and I emerge with a fair bit of trading experience. It’s not the greatest result, but many bad habits are lost.

Second Day: the 1980’s. Time to put the computer on just 4 colors, put on a New Wave Greatest Hits and start in all over again. No ponytail or suspenders, but I can imagine it, as I trade, trade, trade and trade some more. 1987 comes and goes… I’m still trading. Am I good yet? Well not really. But getting better.

Third Day: the 1990’s. I’m at it again. The screen has ok color now but it’s dialed down to 1024 x 768. Nirvana is playing on the radio and I’ve set my Motorola Microtac cellphone out on the desk again where it used to be (yes, I’ve still got it!) In my own mind, I’m gathering 25 years of experience on the charts! It remains to be seen if that’s anything like a real 25 years, or just the first month of trading over and over again, 300 times…

Fourth Day: the 2000’s. Dot Com bubbles bursting all around me, markets rising and crashing… Nickelback is singing about how they want to be rock stars. There is the strange sensation of being in HG Wells’ Time Machine… and then suddenly, hurtling headlong into the future… the charts slam to a crawl. I come to my senses, the computer fans spool down, and it’s early September 2011. Dazed, I step outside into the sunlight. Am I good at forex yet?

So, that’s the imagining here. Likely there’s quite a bit wrong with this. For instance, unless I follow treasuries, the dollar, futures and all that I’m really missing a huuuge portion of the big picture. Maybe a critical portion.

But… has anyone done anything like this?

If you did, did it make a difference?

And if so, did you have a favourite simulator?

Apologies if this has all been covered somewhere else before.

2

Do i think trading with historical information is good? Absolutely. It helps you hone your technical analysis skills faster than doing forward testing. Will you be missing out on the news and fundamental factors yes. But you can practice your system and discipline trading. I haven’t found a satisfactory simulator though. A couple on FF are free to download and put into your MT4 but there are a lot of limitations. I personally have been on the hunt for a good simulator myself. Ninjatrader platform has a built in simulator that you can control the trends up or down or leave neutral and let it do its thing. That way you don’t have to worry about news or whatnot just practice your system. Its free, but i just downloaded it a few days ago and have no idea how to get it to chart :stuck_out_tongue:

3

The [U]only[/U] way to save a lot of time, i.e. years, is using a pretty good sample data over some years and let a EA trade or simulate trading. Plus it’s right, you will see instantly what most of the herd learns just by blowing an account. That’s probably the best part of it.

What we see going on at charts with some dozen trades is not reliable enough. You need a couple of months, better years, to check a system within a lot of different market scenarios like bull trend, bears, ranging, high/low vola, etc. to really get to a point where you can judge this system being profitable or not and what the risks are.

However, to get experience with EAs, simulations, backtests, etc. etc. takes time itself. It took me around two years and I am still learning.

4

Without potentially sounding like I know what I’m doing (because I don’t!) ~ I’ve dabbled with EA’s a bit and had some… interesting results.

There are no shortage of EA’s on the internet, which require little more than loading up, chasing down some optimised parameters that someone was generous enough to share, and let it run… I’ve done that for a fair number of them.

While most perform… questionably, some actually seem to work! I was rather surprised. I’d share specifics but I’d rather not link to other forums, and I’m certainly not one to be handing out tips at this point. I shall, of course, share what I’ve got if I’ve ever ‘got something’ good enough.

All that said, I’m not one to rush in. If I’m not sure why something works, I’m not using it. I made that decision when I encountered a strategy called ‘martingale’ ~ without disparaging anyone who uses that, it’s too scary/risky for me. Results, even years of results aren’t enough for me… I need to know what’s under the hood.

Another thing I’d noted… some of the EA’s I’d come across clearly didn’t stick to the strategy they purported to use! Which baffles me. Again I have to be a bit vague, and I don’t want to be directing people here and there, or speaking ill of anyone. But it has been quite common.

So I’m very shy of public EA’s, especially any where I can’t see the code. Even if it seems to backtest ok. On top of that, if anyone ever published an EA that became a ‘big hit’ and tens of thousands of people (or even one bank!) jumped on the bandwagon, might it erode the strategy? I’m told there aren’t enough of us to make a difference. But it doesn’t matter what I think, it only matters what the owners of successful EA’s think. If they believe it shouldn’t get out, rightly or wrongly it doesn’t matter… well, that’s that then.

I’m dead sure that some top forex bank traders and analysts moving billions around would read this forum too… wouldn’t you, if in that position? If only out of interest? Engineers read EE Times, for instance. I’ve thought of starting a thread: “Forex professionals moving billions, say hello here!” … it might be a total dud or fill with liars, but on the other hand we just might hear some actual Tales from the Dark Side…

I’m willing to put in the time doing this manually, especially with strategies that are running perhaps only 50 trades per year. I’m not about to try to code up anything myself at this point.

Hence the continuing search for the Holy Simulator of Forex Wisdom… it may not be the Holy Grail, but be a rather useful Relic nonetheless…

5

Hi Desmond,

For starters you can look up forextester, it has good reviews in this forum, but it cost money and has indicators problems. MT4 visual mode is of course free and I THINK (because I haven’t tried forextester2) you can perform just as good with it and some program like excel. Either way you can’t test 25 years of data in either of one in 4 days, but I doubt there is a better way.

Good luck with your search of your holy simulators grail.

6

Desmond, I highly recommend Ninjatraders paid version of their platform. I use Ninja myself in my trading, and it has become a regular weekend habit of mine to download weeks or even months of data on any one of the 11 markets I trade.
The data is tick data, and I literally can open up a few different time frames, enter orders on my chart or via the DOM (the futures level 2 screen), and watch the market unfold EXACTLY as it did on the days i downloaded.

it’s about $60 a month for ninja, but IMO, it’s worth it for that service alone.

I’m not just talking historical data…but a tick by tick recording that you can “play”.

To give you an example, I am in the process of downloading the previous week’s AUD/USD and USD/CAD data right now. I will then fast forward each day until it comes to the U.S. session…and then watch the market on a 5 min screen, at 6X regular speed (it can go up to 500X regular speed), and place my orders for the first few hours of the U.S. session.

They will trigger just as they normally would in real market conditions. Then, when the london session ends, I can speed up the recording to 500X normal speed, and get through about 20 hours of market time in 7 minutes. At that point, the U.S. session is starting again, and back I go to trading at 6X market speed…which gets me through the entire 4 hour U.S - London crossover session in about 45 minutes.

It’s a phenominally powerful tool for manual backtesting. i would venture to say it’s about as good as forward testing, because if you take it seriously, you will feel the same emotions and deal with the same psychological issues as in real time demo trading.

Yes, it costs money, but if you are going to be trading with a methodology that relies on your discretion and experience to be successful with, you’d better shell out 60 bucks. I mean…really… you can get 1 year of screen time in a few weeks. This is absolutely priceless.

Jay

P.S. I really don’t know if MT4 offers live playback simulation like ninja does. i know it has historical data…but can you play it back in realtime to test manual strategies?

7

Thank you everyone for the wisdom pouring in… I’ll be looking over pretty much everything.

One of the things I was thinking of doing was trading daily instead of H4 or hourly, being in California I’d have to become completely nocturnal or keep waking when the forex level alert on the smartphone goes off ~ and go off it does, over and over… then twenty minutes later, then again once in deep sleep. I’m assigning a pip value per day to quality sleep… Anyway, trading daily that’s certainly less than 3650 candles per decade. I wonder if anyone could ever really know if they were good at trading daily without being 150 years old?

60 dollars a month for a trading simulator isn’t much at all, I agree. A decent simulator would save me 60 dollars worth of trouble in a matter of days. Though I’m cheap enough to wait until I’m back in town after some traveling this month. I think. I may get impatient… One of the MT4 ‘free’ simulators works ok, but it’s been a bit touchy and rather complicated.

So far I’ve found these threads here, that seem to have some data too:

http://forums.babypips.com/newbie-island/36440-any-good-forex-simulators.html

http://forums.babypips.com/newbie-island/33531-trading-simulator.html

Any and all further commentary appreciated. I’m looking to do this right. I’m not any good, but quite serious about becoming good, however long it takes.