Forum Replies Created
-
Posts
-
I use it mostly because I leave it generating 24×7 all day long in the background on my server without worrying about a web browser crashing or using too much RAM, etc., saving every 15 minutes, so I can stop it at any time, but it just runs indefinitely. And I validate once or twice a week (usually on weekends) on data from multiple brokers, to only pick strategies that pass Monte Carlo on all of them for several different date ranges (last three years combined, 3yrs ago alone, 2yrs ago, 1yr ago, etc.). If it makes it through the “funnel” of about 12 validation steps, then I import them into EA Studio and make a portfolio EA. For me, that whole validation process across three brokers is so much faster to script with a single file that I double click than to do it for each symbol and each broker in EA Studio. But I still use EA Studio in the end to combine the passed strategies into portfolios that I swap out on weekends. All of that could most certainly be done without the Express Generator, but for me it automates it with less mouse clicks. It’s not so much because it generates more strategies faster persay (although it does), it’s more that it makes my time more efficient processing them afterward.
Mostly from the videos posted on the forexsb . com forum, and the discussions over there. I’ve had a long IT career, so it was easy to figure out, even before the videos were made. You’ll figure it out fast with your skillset. 😉 And there are lots of tips and instructions now on the other forum if you get stuck.
You’d be better off adding Express Generator and using that instead of Chrome to generate strategies before spending more money upgrading hardware I think. In your situation you’d be looking at a new CPU, motherboard, and probably RAM as well. A high core count Intel 13th Gen or Ryzen 9 would definitely make a bigger difference, but at a greater cost of investment, and you’d still want to switch to Express Generator over opening more browser tabs to maximize the performance. Express Generator is just SO much faster and more efficient, on any platform, that it probably makes more sense to invest in that first and hardware later. At the end of the day though, more cores is going to get more done when multitasking, no way around it. Also, you can run Express Generator on more than one computer if you have them, dividing up symbols between different machines. Scripting everything is SO much easier and faster than configuring multiple browser tabs. I generate strategies on 28 pairs daily just by double clicking a script and letting it run.
Hey folks!
I’ve decided to attempt the FTMO Challenge, and I’d like to do it with no other tools but EA Studio, just to prove if it can be done. 😉 I decided to document my journey here, successes and failures (because it might take a few attempts!), and also get some advice and input along the way if folks want to chime in or even try it with me.
The objectives are here: For a brief summary, here are the goals (for a normal risk account):
- 10% profit target in 30 days
- 10% maximum loss in 30 days
- 5% maximum daily loss
- Minimum of 10 trading days
- No trading weekends or Sundays
- All trades must be closed before end of Friday each week, no open trades over the weekend
- No trading during 3 bull news 2 minutes before or 2 minutes after (this is problematic for EA trading),
calender of restricted events is here: - Ideally no higher than 1.5% risk per trade, because other competitions that follow similar rules also include this one, although I don’t think FTMO has this particular rule
So, that’s a pretty tough challenge! If you pass it, the requirement to hit 10% profit in a month is lifted, but the others are still in effect to maintain your account, so you can lower your risk per trade substantially for the long term with real money.
Here are my current frustrations that I’m trying to work around:
- I signed up with a free demo first of course, before attempting the real challenge. They give you a demo $50K MT4 account with a broker called FFTraderSRO. The maximum amount of history for every currency pair and timeframe is about 16 days, and they can’t give you any more history (I asked). The demo is only good for 14 days too, so you can’t wait to gather more data. This of course makes EA generation frustrating. I downloaded DukasCopy history and decided to go for M15 timeframe. I also downloaded what data I could from FFTrader. I went to Symbol settings and attempted to get DukasCopy to match by entering the spread, swap day, swap long, swap short, commission, etc. But so far I’m not getting anywhere near the same results with MT4 backtesting vs. EA Studio (limiting the dates to the same range of course). How do you handle generating EAs for multiple brokers that have different history?
- The main symbols are AUDUSD, EURUSD, GBPUSD, NZDUSD, USDCAD, USDCHF, USDJPY, XAGUSD, and XAUUSD. There are many others, but they all have a .lmax afterward and I assume that means these trades would get copied to an account with LMAX. This may be the key for me to solving my first problem. Maybe I can download more history from LMAX and trade the *.lmax symbols on FFTrader instead and get better results. I haven’t had time to test this yet. It would also solve the risk of having too many currencies that are all USD.
- There is no way to set an investor password on the demo account, however you can work around that with FX Blue by using their MT4 EA on a chart to sync instead.
- I’m really struggling to get enough quality EAs generated in 3-4 days. I have a pretty fast 7700K that is liquid cooled and overclocked to 5.2GHz, but it’s only 4 cores (8 threads with hyperthreading). If I go for less stringent requirements, I don’t get very good EAs, even if I get quite a few that pass Monte Carlo. If I go for more strict ones and add Walk Forward optimization as well, then I’m waiting 24hrs or more for a single EA, but 3-4 of them make a really nice looking portfolio. I really want to get 10 or more really good EAs though. Here are my current settings:
- I’m just not getting anything with those above settings yet. I got a few EAs per pair with less stringent requirements, and also added walk forward optimization to them before Monte Carlo, but they failed miserably on the 2 weeks or so of backtesting I could do on FFTrader. I’ll leave the reactor running a few more days. My demo will expire before I even start anything with FTMO on this round, LOL! It will take me a month to generate enough quality EAs with what I thought was a fast computer. But at least I’ve got a glimpse into the broker and some ideas to try LMAX as a demo broker with hopefully more history before requesting another demo with FTMO. Maybe there is a better way to use EAs that Petko has already generated instead of making my own? What do you have for a CPU Petko? Probably more than 4 cores?
- Is there any way to automate pausing or stopping EAs during specific time periods? There are times I’m traveling or working and I’d have to have all EAs paused for 2-3 days before I would get a chance to start them again, just to avoid 2 minutes before or after a critical news time when I can’t be at the computer or remote into it–or risk being disqualified for a trade opening or closing at that critical moment. FTMO has an EA that can close out all open trades at specific time frames, like 5 minutes before an event, but that seems even more brutal with greater potential losses. I’d rather just pause EAs and resume at specific times and automate that in advance, rather than close open or pending trades.
Happy to have any suggestions or feedback! Has anyone else attempted this, and what was your experience?
Haha, yeah, must be something weird with their search functionality. Most registered brokers will list their NFA number and you can search for that too.
I think it’s a bit more complex than that, but I’m not sure. I think it’s more like Binance.com who also runs Binance.us, the latter not being offshore for US citizens even though the parent company is. I think to be registered with the NFA and CFTC means following their rules on those accounts for US citizens, which includes FIFO sadly. MT5 accounts can be netting or hedging, and I haven’t opened one to find out yet, but I suspect it is netting only and FIFO compliant for US citizens, meaning it’s no more offshore than Forex.com. But I’ll have to wait to hear from someone who has opened a live account with them and tried. 😉
That’s strange, they’ve been registered since 1996 or so. Must have been the search parameters, the .com not searching right or something weird. Glad you found them!
Oanda is registered with the CFTC and the NFA according to the bottom of their website:
“OANDA Corporation is a registered Futures Commission Merchant and Retail Foreign Exchange Dealer with the Commodity Futures Trading Commission and is a member of the National Futures Association. No: 0325821. More Information is available using the NFA Basic resource.”
I have accounts with both them and Forex.com and preferred them over Forex.com for trading conditions. I have not yet opened an account on Trading.com yet to compare.
Trading.com is probably the best at the moment. They are the only broker registered with the CFTC and NFA that offer MT5 for US citizens to my knowledge. I have asked IG, Oanda, and Forex.com for 6 years to add MT5 and none of them have any plans to do so.
Ingot Brokers is one of a handful of entirely unregulated brokers that allow US clients, no FIFO requirements. They support MT4 and MT5. Some of the others I have used are AAFX, FXChoice, TradersWay, and Coinexx. All five have been decent brokers for me, with various options for crypto, indices, leverage amounts, funding methods, etc.
Hey Petko!
Yes, I have done very well in crypto as well the past three months. I haven’t done as much with stocks yet, mainly because of the broker choices US citizens are limited to. I look forward to seeing your new courses.
Hey folks!
I haven’t written here in a while because I’ve sort of given up on strategies for the time being. They backtest well, but don’t trade live very profitably, especially since around October. I’ve tried for months to use only robust strategies that pass multimarket, Monte Carlo, etc. and yet have been unsuccessful in live trading. Many people using EAs have had similar complaints, so it could just be that the market is vastly different in the last quarter than from the previous year or two for strategies to do well. Maybe in time we will be able to develop more robust strategies again, but for the moment I’m working on other things outside of EA Studio. I do think that trading a portfolio of some kind with many currency pairs and hedging at a low risk % per trade of your account balance / equity is wise, because it reduces your overall risk compared to a single strategy or currency pair. That concept is still sound, but trading strategies today that were generated from previous data is not going very well without adding filters or a recovery method or other ideas to it because the current market is so different. So many currency pairs are just ranging right now without strong trends. Backtests of strategies will look good when it was trending more, but live trading will not while the pairs are ranging, etc. On the other hand, EAs that look promising right now but use grid trading, Martingale, etc. are also extremely risky and will usually blow an account at some point when a ranging market goes back into a very strong trend. Hence why I mention filters and other methods outside of EA Studio are where my focus is at for the moment.
Two prop firms that I really like so far are City Traders and The5ers. Both are different than FTMO because they allow trading during news and leaving trades open over the weekend, and they have a much longer time period to achieve your goal (6 months to a year). Do a little research on them and see if they fit your style of trading or not. For me it’s slower paced and lower risk per trade than trying to hit x % profit per month as a crazy goal, and I like that style of trading better. City Traders in particular has a very active Discord group for support.
I don’t know about direct contact outside of the forum and if it’s allowed or not. I don’t see a private message feature for users here. You can find me on Telegram as @AaronPriest.
Best of luck trading!
FTMO has had a lot of issues lately switching to a cheaper service with their liquidity provider (LMAX)–many complaints on their Facebook page and in other forums. I have passed two demo accounts with them, but not bothered going for a paid challenge yet until I read fewer nightmare stories every week. They brutally butchered one of my own demo accounts, converting it from USD to GBP then claiming I had a $20K loss when they switched it back to USD, etc., etc. Growing pains I assume with so many new signups and trying to find cheaper rates for the amount of traffic they are now getting. I don’t think they are a scam or anything, but I’m giving them time to become a more mature service again before spending money with them on a challenge. I have two funded accounts with other prop firms in the mean time. So, not much of an update I guess, but that’s where I’m at. 😉
OK, apparently I was mistaken about the number of trades each strategy opens at a time. I probably reused a magic number or accidentally gave two portfolio EAs the same magic number making it appear that each strategy could open many trades at once. That is good to know, thanks!
Thanks Petko,
More than 1 image per post would make sense with what I’ve seen regarding posts getting blocked or missing. Thanks for looking into it!
I’ve been using the Forex Historical data app with custom symbol settings, but there weren’t very many pairs listed there. I had my best results with gold using it though since the data was more complete.
So, to be sure I understand what you wrote, there is no way to limit the number of active trades each strategy opens at a time. But running the reactor on half the amount of bars will result in potentially twice as many trades being open at once by those strategies (all other things being equal)?
