Transferring Success in Simulator to Demo

Since I graduated in May I've been lucky enough to have a lot of free time to work at trading. In that time I've done some reading (Methods of a Wall Street Master, Trading In the Zone, part of Markets In Profile) and spent a lot of time looking at charts. It was sort of an epiphany for me when I came across Feb's System II which employs some of the same methodology found in Methods of a Wall Street Master. Without going into the specifics too much (you can search the forum if you're curious) I finally had that feeling of trading with a system that seemed to click with me.

Without going any further I'll say I've only demo traded it for the month of July so far. Since I'm not doing much else I've tried to be very diligent about it. Monday-Friday I begin trading the Euro at 8 AM and watch it pretty closely until 12. I would say the benefit of this is that already I'm starting to get a better "feel" for this currency and how it moves.

After about a month I'm beginning to feel like what Mark Douglas calls a "boom and bust trader". I seem to get a few wins which bring my equity up then many small losses which eat away at what I've won. At best I can't seem to do better than breakeven and it's possible as time goes on I may consistently lose.

What makes this most frustrating is that I've been doing well when I practice previous months in Forex Tester. If you aren't familiar with the program it essentially lets you trade any data from the last 10 years or so in realtime or sped up. I've tried to be extremely honest in trading Feb's system this way. I know how easy it can be to "cheat" on your results, rewinding the tape to change the way you traded something. The months I've done so far have been encouraging and I feel I've learned so much.

So finally here is my question: why do I come up with good results trading in the simulator but not in the demo? Some obvious ideas right off the bat:

-I only have one month of results trading this system "live" in a demo account. This isn't enough to draw any meaningful conclusions. The reason I want some feedback anyway is because I'm feeling frustrated as I've come out ahead every month I've traded in the simulator. I know liquidity dries up in the summer in the stock market, does this happen in FOREX and could this have an effect?

-Trading a simulator, though useful, is not equal to trading a demo account which is not equal to trading a live account. As strict as I've tried to be in trading the simulator it's not the same as trading in realtime. Also, because I put so much pressure on myself to succeed, trading a demo account for me probably approaches the stress others experience trading live. In the simulator I can actually relax and just trade without fear. Reading Trading In the Zone has gotten me thinking about that a lot. Still I think after getting the hang of it in the sim I've adapted what I've learned to the demo.

Well thanks to anyone who read all that. Any thoughts are appreciated. If you want to tell me I'm a dummy and I should wait more than a month before I start drawing conclusions well that's allowed too.

Ian

you possibly aren't considering things like changes in spread compared to what you have backtested with in the historical data. plus the change in speed of real time versus sim trading is probably going to mess with your head too.

I think change in speed could definitely be a factor. The thing is I'm used to seeing fully formed one minute bars one after another in the sim. While I actually tend to trade much more quickly there this may be helping me because it leaves less time to question what I'm doing as I'm not watching the bars take shape but pop in one after another every 5 seconds or so. Then when I need to change stop loss or something I pause the sim thought not for more time than it would take for another bar to fully form in realtime.

I don't think spread is an issue because it's actually larger in the simulator than it is for my Oanda account.

slow the sim down to see how you go when it's not running at light speed. then you can get an idea of how it feels when everything takes longer to resolve.

Cool thread, appreciate it.

I've been using simulator on the 1' TF for almost two years, and find it much better than demo. Occasionally I'll toy with a trade in demo, but know that the real work is in doing simulator. I average 3-5 trades in a half hour, and used to do 5-24 in a couple of hours. I use a shotgun approach to mm to build consistency -- that is, I'll avoid mm, except for a 90-pip mental SL. It costs more margin calls / unacceptably big losses, for the benefit of inspiration to improve the method in various ways. And, I trade from the moment I open and set up the chart; I probably should switch to different starting points, am too lazy I guess.
Thus far, I've been able to shape my discretionary trading method's money management, chart study (drawing devices (mostly TL's) on candles) and built-in psych controls. I never could have done this in demo.
And going over the same years never gets old, since I go with what I put on the chart, not what I may (not) remember. I just have to do a simulated year each day at a time, which takes months realtime, skipping some days to avoid gaps, overlapping history and such.


I can see slowing down the simulator, but find top or near-top speed trains the mind to recognize patterns better. If I see a place to draw, then I slow it down. It's fairly systematic that way, which helps reduce variables.
I was thinking about demarcating when I'm available to trade realtime, and speed up to those points and then go slow on the simulator; put it off, since not even close for realtime.

despite fearing to derail this currently clean thread i'll ask : how can you only manage 3-5 trades / half hour in sim? a half hour in sim for me is about 1 to 2 hundred trades...

My brain is alot slower, perhaps. I study to get a trade setup for up to five minutes a trade. And, I may get refreshments or something. Hey, it's discretionary, cut me some slack

hmm, k.

brings me to a new thought... perhaps the OP is used to faked tick data generating candles? if you base a positive expectancy on faked data then it's only going to be a world of hurt when you experience actual real data. i'm not familiar with Forex Tester, so i don't know if it simulates tick data or if it even deals in ticks at all or just generates completely formed candles.

I don't know, but I do know there's a world of difference between the simulator and real feeds. Time distance between candle moves seems to be too regular in simulator, I think. I attributed it to why Ronald Raygun seems to use his own tick databases to run his own simulators; maybe I'm wrong, and there's no way to record the noted time distance.
I do find that the reduced ability of a simulator to put ticks on the screen into a TF bar, can be a benefit. It is so because, before, when I was on demo or real, I would play off my gut feel rendered from the time distance. Gut feel, as one'd suppose, isn't very exact; so, I now turn my attention more to how I technically interpret the chart.

This is definitely possible. I know that the 1m tick data is faked. I actually take trades off 5m charts and use the 1m to manage my trades once I'm in. I'm pretty sure the 5m data is genuine it's just the movement before the bar closes that is system generated. Could this still have an effect?

yup of course it will have an effect. if you are trading fake tick data then you will find a way to game the simulated data. you either need to get genuine data, or ignore the intra-candle movements and only trade candle closes to get around the issue.

I didn't realize the 1m data would affect the 5m data if I was trading closed 5m bars. They provide real 1m data for 20/month. They also have tick data for 30/month but I can't really afford that for long.

that's different. if a closed candle is what it is then cool. if, however, you are trading intra-candle with generated tick data then you are asking for trouble. if you are only trading closed candles then you are probably fine. not being familiar with forex tester i can only assume. only you will know, being that you are familiar with the software.

p.s. i'm sure you could find a way to import data without having to pay for it. mt4 brokers give their data for free. with a bit of creative thinking and coding i'm sure you could get free data into the software.