Thanks for the reply. I do have 2 MT4s installed I will delete one and see what happens.
I have a general data-related question for Smart Live.
I'm testing the same bot on two different platforms - Smart Live (live account) and Alpari (demo account). It became clear that Alpari had a much higher probability of taking trades, which wasn't related to the different spread.
It seemed my Smart Live MT4 wasn't updating as frequently as Alpari so I added a 'tick counter' to my EA. After around three hours Smart Live had recorded 2778 ticks, and Alpari 7602. Those results have remained fairly consistent, with Smart Live seeing approximately 36% of the tick data which runs through Alpari.
I understand that updates are inevitably going to be discrete data snapshots and providers are generally making their own prices so there are always going to be differences, but I wondered if my results were unusual?
Both MT4 applications were running on the same computer, which I gathered from other people shouldn't be a problem, but which I see above now might be. In fact, the Smart Live MT4 application now refuses to start at all on my machine (even without Alpari's MT4 running), so the tick frequency issue is a bit of an academic question at the moment, but it seems that fewer tick updates are going to increase the likelihood of variance from backtest results.
I'm new to forex trading and bots so maybe Alpari are just being unnecessarily volatile by nature but the upshot is that my bot seems to do much better on Alpari (spread-adjusted) because it sees more qualifying entries and exits on that platform. Do I need to look at my network routing, getting a VPS nearer Smart Live or accept that this is just how things are when comparing Smart Live and Alpari?
I'm testing the same bot on two different platforms - Smart Live (live account) and Alpari (demo account). It became clear that Alpari had a much higher probability of taking trades, which wasn't related to the different spread.
It seemed my Smart Live MT4 wasn't updating as frequently as Alpari so I added a 'tick counter' to my EA. After around three hours Smart Live had recorded 2778 ticks, and Alpari 7602. Those results have remained fairly consistent, with Smart Live seeing approximately 36% of the tick data which runs through Alpari.
I understand that updates are inevitably going to be discrete data snapshots and providers are generally making their own prices so there are always going to be differences, but I wondered if my results were unusual?
Both MT4 applications were running on the same computer, which I gathered from other people shouldn't be a problem, but which I see above now might be. In fact, the Smart Live MT4 application now refuses to start at all on my machine (even without Alpari's MT4 running), so the tick frequency issue is a bit of an academic question at the moment, but it seems that fewer tick updates are going to increase the likelihood of variance from backtest results.
I'm new to forex trading and bots so maybe Alpari are just being unnecessarily volatile by nature but the upshot is that my bot seems to do much better on Alpari (spread-adjusted) because it sees more qualifying entries and exits on that platform. Do I need to look at my network routing, getting a VPS nearer Smart Live or accept that this is just how things are when comparing Smart Live and Alpari?
Are you sure you didn't put the indicator in the wrong directory (and/or have now deleted it?
Yeah in the experts-indicator folder. Its not been deleted either god knows whats up eith it.
Trading Gurus
Active member
- Messages
- 159
- Likes
- 2
Hi Kyoto,
Alpari's MT4 demo uses their Pro feed. It is not a good guide to what you will see on a micro or classic account.
We run dozens of MT4 installations on one VPS, so multiple MT4s is not of itself your problem. We've also tested Smart Live with EAs running on a VPS in both U.K. and U.S. data centres with no issues.
I fear more information is required before a solution to your problem will present itself.
Jim
Alpari's MT4 demo uses their Pro feed. It is not a good guide to what you will see on a micro or classic account.
We run dozens of MT4 installations on one VPS, so multiple MT4s is not of itself your problem. We've also tested Smart Live with EAs running on a VPS in both U.K. and U.S. data centres with no issues.
I fear more information is required before a solution to your problem will present itself.
Jim
Thanks for that Jim - I didn't realise there was this kind of data difference between Alpari's demo and non-pro live accounts. I'm in Asia at the moment and thought it could be routing with Smart Live - even though Alpari seems fine - but perhaps it's simply that one feed is better by design.
I was just reading that link you included a couple of days ago. I know there are lots of problems with the Strategy Tester but I didn't really think about tick frequency as potentially being one of them until I ran up against the issue.
I suppose it's always going to be an inexact science, but using the Strategy Tester is proving too inexact for my liking. I'm going look into writing a replacement for it to backtest my strategies.
Trading Gurus
Active member
- Messages
- 159
- Likes
- 2
Hi Kyoto,
If you can live with one minute bars you might wish to save reinventing the wheel, and instead take a look at Forex Strategy Builder. If you really need to backtest using accurate tick data you're going to have to look beyond MetaTrader I'm afraid.
Maybe we'll have to find ourselves an Asian data centre for testing purposes too! You're in Japan?
Jim
If you can live with one minute bars you might wish to save reinventing the wheel, and instead take a look at Forex Strategy Builder. If you really need to backtest using accurate tick data you're going to have to look beyond MetaTrader I'm afraid.
Maybe we'll have to find ourselves an Asian data centre for testing purposes too! You're in Japan?
Jim
Thanks for the link (again!) Jim. I'll have a look at that. As far as reinventing the wheel is concerned, for the relatively small jobs I usually find the best way of getting exactly what I want is to write my own tools - before becoming a full-time derivatives trader six years ago I used to work as a software developer, so it's in the blood really. And I used to write clinical research systems - which is turning out to be a very transferable skill when it comes to backtesting and writing analysis utilities for forex
It's occurred to me that the only accurate tick data that exists for properly backtesting MetaTrader-based bots is the tick data that each specific MetaTrader-based broker generates - and as I'm discovering - this can be significantly different - so it has to be approached on a case by case basis. I've downloaded, cleaned and data-mapped some better data into my test platform than MetaQuotes provides, but I'm thinking that writing an EA (since the DDE data set is too limited) to output real tick data for backtesting might be a way forward. It's looks possible in principle, but there's a difference between what works on paper and what works in an operational environment.
I'm based in South Korea. We have a very highly developed Internet infrastructure here (100 megabit connections like mine have been common for years and they're cheap), but when it comes to connectivity to Europe and the US, like Japan, it's obviously got some unavoidable latency issues. People are quite keen on trading here - there are two channels on my TV that show nothing but candlestick charts all day with commentary and analysis, but unfortunately I can't find a Korean broker that uses MT4.
edit: I've spent some time looking at Forex Strategy Builder now. It's a very interesting piece of software and it's a more polished version of what I had in my mind to build. Perhaps it's going to save me a job.
Last edited:
Tenapenny
Experienced member
- Messages
- 1,005
- Likes
- 46
Trading Gurus
Active member
- Messages
- 159
- Likes
- 2
Hi Kyoto,
It's in my blood too. I started doing optimisation runs for real robots on mainframes back in the 70s. I wrote my first backtesting software in the early days of the dot com boom.
If you're into Java you can get historical tick data courtesy of a Dukascopy demo account.
Surely with your skill set you want to find a broker with an API that's more robust than MT4?
Keep us posted on how you get on with it!
Cheers,
Jim
As far as reinventing the wheel is concerned, for the relatively small jobs I usually find the best way of getting exactly what I want is to write my own tools - before becoming a full-time derivatives trader six years ago I used to work as a software developer, so it's in the blood really. And I used to write clinical research systems - which is turning out to be a very transferable skill when it comes to backtesting and writing analysis utilities for forex
It's in my blood too. I started doing optimisation runs for real robots on mainframes back in the 70s. I wrote my first backtesting software in the early days of the dot com boom.
If you're into Java you can get historical tick data courtesy of a Dukascopy demo account.
Surely with your skill set you want to find a broker with an API that's more robust than MT4?
Keep us posted on how you get on with it!
Cheers,
Jim
Live trading update for those interested---had the live account a while now and been very impressed.
I had a trade this week that I believed would be stopped as it came within one pip several times for a few minutes and NO stopout. Have had a couple of requotes but only when the market was motoring. Admittedly I have been trading with small stakes but upping the anti worked fine and now comfortable to use this platform full time . Now my favourite and hope it continues (just need to trade better but i guess thats down to me) I trade mainly indices and stocks by the way.
I had a trade this week that I believed would be stopped as it came within one pip several times for a few minutes and NO stopout. Have had a couple of requotes but only when the market was motoring. Admittedly I have been trading with small stakes but upping the anti worked fine and now comfortable to use this platform full time . Now my favourite and hope it continues (just need to trade better but i guess thats down to me) I trade mainly indices and stocks by the way.
fibbinstarchi
Active member
- Messages
- 164
- Likes
- 4
You shouldnt be getting requote bollox
francisfinley
Established member
- Messages
- 902
- Likes
- 46
nicely put.
Smart Live Markets
Active member
- Messages
- 160
- Likes
- 1
Hi
In an ideal world you should not be getting requotes. Could you let me know what market you are experiencing this on please.
I am guessing it is on a market that is 1 pip wide. We could make them 2 pips wide and build in more tollerance.... you tell me what you would prefer. Good as MT4 maybe for a lot of things, it does have a few issues, which involves a little give and take.
We have put spread betting on MT4, not just 20 or so currencies but over 100 and another 700 markets as well. We offer what I beleive to be the fairest and lowest financing in the industry as well as some of the tightest spreads.
I wish I was allowed to deal with us, because I would !!
Paul
SLM
Smart Live Markets
Active member
- Messages
- 160
- Likes
- 1
Thanks for the nice comments.
Here is an exclusive for you guys..... we have launched an iPhone App for MT4.
So you can now spread bet anywhere using MT4.
Just go to the App store and search for Smart Live...the app is free.
We need to make a couple of small changes to it, but it works very well.
Your views would be appreciated as always.
Regards
SLM