Million Dollar Pips live forward test

Trading style: scalping, very short trade duration, multiple concurrent positions possible; due to the low average number of pips per trade, the results for this EA will certainly vary greatly from one broker to another.
Currency pairs: EURUSD (USDJPY and GBPUSD versions exist as a separate purchases worth $99 each)
Timeframe: M1
Expert Advisor price: $99
NFA compliance: full (FIFO setting has to be enabled)
Refund policy: 30 days, no questions asked
Read more at the Million Dollar Pips website

Buy Million Dollar Pips

Birt’s forward test

EURUSD

Settings: all default, except: Risk 1.5, FIFO disabled; starting with 11.10.2011, Stop Orders, Hard Stop Trailing and AutoAdjust_ECN are set to false
Started: 07.08.2011
Broker: PrivateFx
Account type: live, micro
Starting balance: $300

GBPUSD

Settings: all default, except: Risk 1.5, FIFO false, Stop Orders false, Hard Stop Trailing false and AutoAdjust_ECN false
Started: 24.10.2011
Broker: PrivateFx
Account type: live, micro
Starting balance: $300
Current version: 1.2.0

USDJPY

Official forward test

Started: 05.05.2011
Broker: ThinkForex
Account type: demo
Starting balance: $1000

Million Dollar Pips official demo forward test

User accounts

Started: 21.10.2011
Settings: risk=3, max_orders=4, FIFO=false, group_orders=true, AutoApplyECN=false
Broker: IBFX
Account type: live standard
Owner: Fred

Million Dollar Pips Fred's IBFX live account

Started: 06.09.2011
Settings: unspecified
Broker: IBFX
Account type: live
Owner: Professor53
Note: the account is using a proprietary VPS, search for Professor53′s comments below.

Million Dollar Pips Professor53's IBFX live account

Backtests

After seeing the comment posted by Eli below I decided to run my own backtest. I used Dukascopy tick data with a fixed spread of 1.5 and reset all the EA settings to their default values. As you will see below, the result was insanely good (although with a significant drawdown, probably because I used the default risk), but I believe backtesting this particular EA is not realistic due to the very low pip targets. In all likelihood, it will behave very, very different from a broker to another and testing with Dukascopy data is totally irrelevant since you’re most likely not going to run it on a Dukascopy account.

Million Dollar Pips 1.6 2007-2011 tick data, spread 1.5, default settings

Show backtest summary »

Symbol		EURUSD (Euro vs US Dollar)
Period		1 Minute (M1) 2007.03.30 16:01 - 2011.08.05 19:59
Model		Every tick (the most precise method based on all available least timeframes)
Parameters		Configuration="================ Configuration"; Show_Diagnostics=true; Show_Debug=false; Verbose=false; Silent=false; NumOrders_Level_Info_1="Range of 0..4, adjusts Number of Trades"; NumOrders_Level_Info_2="0=Max Profit% 4=Max Num.Trades/Rebates"; NumOrders_Level=0; Additional_Channels_Info_1="Range of 0..8"; Additional_Channels_Info_2="Safely improves trading frequency"; Additional_Channels=1; Slippage_Info_1="Lower value improves profit rate,"; Slippage_Info_2="higher: trades more but less accurately"; Slippage=3; Username=""; Magic=112226; Max_Simultaneous_Orders=8; FIFO_Info_1="if Simultaneous_Orders>1 ,"; FIFO_Info_2="u.s. citizens should turn it on"; FIFO_Info_3="(required for NFA compliance)"; FIFO=true; OrderCmt="MDP_EURUSD_1.1.6"; Pessimistic_Testing_Info_1="Simulate bad live account conditions"; Pessimistic_Testing_Info_2="during backtesting (i.e. Slippages)"; Pessimistic_Testing_Info_3="(this has no effect in FIFO mode :"; Pessimistic_Testing_Info_4=" +turn off FIFO for "optimistic")"; Pessimistic_Testing=false; Funky_Exit_Info="Exit trades sooner"; Funky_Exit=true; AutoApply_ECN_Info_1="Use ECN account settings automatically"; AutoApply_ECN_Info_2="when ECN conditions detected"; AutoApply_ECN=true; Money_Management="---------------- Money Management"; Min_Lots=0.001; Max_Lots=1000; Risk_Info_1="Fixed risk % of balance per order"; Risk_Info_2="NOTE : Trading volume is very variable"; Risk_Info_3="as stop losses are in different distances"; Risk=1.5; Risk_Mode_CommPips_Info_1="Risk is commensurate with target pips"; Risk_Mode_CommPips_Info_2="otherwise it's only determined by %"; Risk_Mode_CommPips=true; SL_TP_Trailing="---------------- SL / TP / Trailing"; Trailing_Resolution=5; Use_Stop_Orders_Info_1="Set true if 1) your broker doesn't deal with"; Use_Stop_Orders_Info_2="opening slippage when executing STOP orders"; Use_Stop_Orders_Info_3="and 2) broker stop level is <= 2.0 pips"; Use_Stop_Orders=true; Hard_Stop_Trailing_Info_1="Trailing hard Stop/Loss and Take/Profit"; Hard_Stop_Trailing_Info_2="Otherwise, doing it in programmatic manner"; Hard_Stop_Trailing_Info_3="and SL/TP is just for covering"; Hard_Stop_Trailing=true;

Bars in test	1610964	Ticks modelled	56219367	Modelling quality	99.00%
Mismatched charts errors	0

Initial deposit	10000.00
Total net profit	149692548.28	Gross profit	237886860.68	Gross loss	-88194312.40
Profit factor	2.70	Expected payoff	11540.56
Absolute drawdown	6138.15	Maximal drawdown	4125000.00 (9.74%)	Relative drawdown	63.67% (8884.05)

Total trades	12971	Short positions (won %)	8884 (46.48%)	Long positions (won %)	4087 (46.37%)
		Profit trades (% of total)	6024 (46.44%)	Loss trades (% of total)	6947 (53.56%)
Largest		profit trade	569000.00	loss trade	-95000.00
Average		profit trade	39489.85	loss trade	-12695.31
Maximum		consecutive wins (profit in money)	55 (1606.65)	consecutive losses (loss in money)	83 (-1236.85)
Maximal		consecutive profit (count of wins)	5722000.00 (19)	consecutive loss (count of losses)	-2370000.00 (44)
Average		consecutive wins	6	consecutive losses	7

This entry was posted by birt on August 7, 2011 at 7:37 PM, and is filed under Live Forward Tests. Follow any responses to this post through RSS 2.0.You can leave a response or trackback from your own site.

Comments (475)
Pings (1)
Related Posts

#442 written by Gary February 10, 2012 (3 months ago)

Thanks for that Jean-Paul, very informative although it has made me a little paranoid to think what the big companies will get up to, especially with someone like me who doesn’t really know what to look for.
I’m also using Universal FX at the moment, (but just the demo version) and was planning to try them out with a live account. I just need to research further for a compatible VPS provider.
When you say I need a VPS service nearby, do you mean that geographically?
I am based in the UK, and Universal FX have an office in London, they recommend a VPS provider by the name of UK2 group and they quote an average latency of around 4ms. (I know…. it used to say OXO on the buses but they never sold it!!!) can anyone confirn that?
Does anyone know of any reason that I shouldn’t give them a try?
Reply Quote
- #443 written by John February 16, 2012 (3 months ago)
  
  4ms latency from that VPS to your broker is fine, but the most important question is what their latency is from the broker to the liquidity providers!? this is where performance will vary greatly. anything under 10ms is ideal so you’ll have to find out how close they are to the liquidity providers.
  Reply Quote
#444 written by Joao February 11, 2012 (3 months ago)

I tried it and no profits. This bot works more and less for the vendors. May be they have special conditions at certain broker, but for retail traders is very dificult to get profits. I read several foruns and the conclusion is that the market has un antidot against this robot, like stops with certain value, certain time to keep the trade, requotes and the slippage. Besides this is necessary to rent a VPS and my opinion is the bot don´t earn for the VPS.
Reply Quote
#445 written by Douglas February 11, 2012 (3 months ago)

Perhaps I am naive but why would a true ECN or STP broker want an “antidote”? I was under the impression they wanted successful traders = more lots traded, more money for them. Birt, any comment?
Reply Quote
- #446 written by Anthony February 11, 2012 (3 months ago)
  
  It is true that they want successful traders, but ONLY if those traders aren’t causing them to have extra expenses. MDP users require HUGE in/out data capacity. The brokers don’t want to have to upgrade their servers to handle the monster information flow to handle a brief moment to cater to a few traders. They would rather watch them fail and hope that they stop using MDP…..plus make a huge amount on commissions while they wait for the MDP user to get a clue and stop trying.
  Reply Quote
#447 written by Joao February 11, 2012 (3 months ago)

Think on this. If you earn, who loses ? Or the broker ( market Maker) or the interbank market. Interbank market have the resources to fight against the robots, the traders and so on. One side are the traders, the other side are the brokers and interbank market. Where is the power, the money, the knowledge ? There are ant-vírus, may be there are anti-robots, simply we don´t know, but one thing we know, many traders have had losses with the robots, and the testes that are made to the them, say very few are profitable. Check to see that the same robot has one perfomance in the live account of vendor´s and a diferent perfomance at others live accounts, big diference, not small. Why ? For me the statements are falsified.
Reply Quote
#448 written by Joao February 11, 2012 (3 months ago)

It is well known that robots have success only for a limited time. After that, usually the market “recognizes” their system, and they start losing. I think it is quite clear fapturbo reached that point. I think it is much too late for using it; from now on it will mostly lose.
Reply Quote
- #449 written by John February 15, 2012 (3 months ago)
  
  that’s if you’re using a broker that is a market maker! if you’re using a true ecn environment, that is not the case. the ecn broker makes their commission on a flat rate depending on the size of the lot you’re trading. they could care less if you’re winning or losing as long as you’re trading! it’s the same with stock brokers. they make money on you whether you are making or losing money as long as you’re buying and selling shares of stock.
  Reply Quote
#450 written by John February 15, 2012 (3 months ago)

MDP can work in the right environment. Too many people only look at latency between their VPS if using one and their broker. Although it is important to have low latency between you and the broker, it is even more important to know what the latency is between your broker and the liquidity providers!! That is the key for any scalping strategies whether it is an EA or some manual indicator. This is why people report that such and such scalping ea will be broker sensitive. The main reason the vendor shows better results than most is because they have their bot running in the most advantageous environment. You could have less than 5ms latency between your VPS and the broker of choice, but the latency between your broker and the liquidity providers could be up over 100ms latency which will most certainly affect the performance of scalping strategies! Just my 2 cents. happy trading
Reply Quote
#451 written by Douglas February 15, 2012 (3 months ago)

Took alot of losses on MDP last night…reduced risk from 0.5 to 0.25
Reply Quote
#452 written by cd February 20, 2012 (2 months ago)

The user accounts test was actually using HotForex, not ThinkForex.
Reply Quote
#453 written by Jo March 1, 2012 (2 months ago)

This’s the news for anyone who’s new in forex or unaware with what may happen with your trade with your broker.
Today I decided to thoroughly check last night and Feb trades on my platform after I read many comments about the trade that’s closed beyond stop-lost you or your robot put in. It was surprissed me to find out that many of them were closed beyond set stop-lost. Some of them 3.2 pips beyond set stop-close.
I contacted my broker (GoMarkets) and I was very disppointed to learn that no guarantee the trade will close at set stop-lose. Some robot, i.e MDP, took big lot in one trade. Just imagine if your robot take 5.1 lots, which happened in my case, we will lose an extra $163 in one trade. From my experiences in a day robot (MDP) can take up to 20 trades. Just imagine how much extra loses we need to pay.
Please be aware !!!!
Reply Quote
- #454 written by birt March 1, 2012 (2 months ago)
  
  I’m afraid that’s quite the norm, actually – most brokers will not guarantee stop orders because it’s quite impossible, there’s no certainty that liquidity will be available at your selected price. This is very “visible” for MDP which trades during periods of very high volatility. If you see a broker that guarantees stop orders, I suggest staying away, something is fishy there.
  Reply Quote
- #455 written by birt March 1, 2012 (2 months ago)
  
  By the way, I deleted your duplicate post from the FFM article because it did not belong there.
  Reply Quote
- #456 written by Anthony March 1, 2012 (2 months ago)
  
  This is what is referred to as “slippage”. This is why a backtest and/or test on a demo account of any scalper is 100% unreliable. Slippage cannot be accurately emulated in a demo account and is not accurately emulated in a backtest.
  That said…..becareful. Slippage can be VERY nasty is fast moving markets. I have been slipped 21 pips before during news trades. It can get MUCH worse than that.
  That said, slippage is not entirely bad if you are with a good broker. A good broker will also give you positive slippage. FinFX is one of the brokers I have tried (and am still with) that has given me positive slippage multiple times. Maybe even more often than I have gotten negative slippage. Anyway….sorry you had to find out about slippage the bad way.
  Reply Quote
#457 written by Matthew March 6, 2012 (2 months ago)

Just thought I would share this with everyone: MDP do not have a REAL money account – only demo. I think his response says it all.
Edit by Birt: removed the emails posted by Matthew. Nothing very interesting other than the developer linking the demo account twice. In case anyone didn’t know already, there is no official MDP live account, just a demo which is doing much better than any live account would.
Reply Quote
#458 written by RG March 19, 2012 (1 month ago)

Interesting concept in knowing latency between broker and liq provider… nvr tought about that.. no wonder when looking at my EA Journal, I’ve seen fill delay about 3-4 seconds. Even though the ping from my VPS to broker is only 6ms.
So my question is, how do we know the ping from broker to liq provider. Ask them? I doubt they will provide such info… I wonder if an EA can be created to measure this delay…
Reply Quote
#459 written by Jean-Paul March 19, 2012 (1 month ago)

hi RG
latency is key!!! check pepperstone & beekfx (VPS).
check the speed o meter on :
http://forex-expats.com/recommended/
file: extestforexpat.ex4
with this ea you relatively easily measure ’round-trip’ lantency issuing, changing, canceling orders from you MT4 client. anything below 300ms is suitable for MDP. run the programm on real accounts only. because demo usually have less than 150ms.
last friday, running MDP on two different MT4 server on a VPS with both lantency <300ms resulted in one gaining $30 and one losing $30. running MDP for 6 month now and gained and lost and gained and lost. from experience seems very very instable, so that i dont consider putting down real money, only 'pocket' money.
good luck!
Reply Quote
- #460 written by RG April 22, 2012 (3 weeks ago)
  
  Hi Jean-Paul, Gaultier?
  Thanks for the EA to check the whole loop. I’m testing it next week to see how ATC performs. Hows your trading doing.. I’m still undecided which EA to get, after getting hammered in the past… its between top/oddbot, phibase, or the new neural net EA…
  Cheers!
  Reply Quote
#461 written by Peter March 25, 2012 (1 month ago)

Great review, love your articles Birt!
I have a question regarding the demo Accounts on Brokers.
Lets say i send an order at pm 5:00 to buy 1 Lot for 1.30.
Price moves up to 1.31.
My order arrives at the Demo Server.
Does my PC just send the order to the Demo server with the entry (1.30) taken from my PC (1.30) or does it check against the actual prices (1.31) when the order hits the demo server?
Reply Quote
- #462 written by birt March 25, 2012 (1 month ago)
  
  I’ve seen slippage on demo, so it must be that the current price is checked. Otherwise, you’d be able to send an order to buy at 1.25 even though the price is 1.30.
  Do some tests: try to send an order 1 pip away from the market price (in your favor) and see what happens.
  Reply Quote
#463 written by Eddie March 27, 2012 (1 month ago)

Hi Birt,
Could you please review forex Zombi EA? http://www.fxzombi.com/
Reply Quote
#464 written by socalengineer March 27, 2012 (1 month ago)

Hi Birt,
A few questions…
1. It seems you are compensated for people purchasing this EA, as when I click the buy link it shows that the referral came from xbirt. And also this EA is rated as the highest according to your own proprietary “Birt’s Index”. Do you have any affiliation with this company? I am hoping this is not the case, because this review website looks legitimate to me, and I would hate to find any bias in your review.
2. Which EA would you recommend trading with for maximum profits and least risk factor? I am assuming according to your Birt’s index it would be this EA. Please let me know your opinion, thanks Birt.
Reply Quote
- #465 written by birt March 27, 2012 (1 month ago)
  
  I am affiliated with almost all the EAs that you see on this website, a fact that is mentioned in several places including the about page. I make no attempt to hide that and I do my best to keep unbiased. It’s quite impossible to keep alive a site of this magnitude without being affiliated.
  If you plan to buy MDP, I would advise reading all the comments on this page and perhaps reading other forums as well prior to purchasing it. It can be a very profitable EA, but it is very, very hard to find a broker + VPS + settings combination that gets it to work profitably. Most brokers are running server plugins that will render it unprofitable. Due to its popularity, available liquidity is a concern so people with such working setups will generally keep them quiet. If you’re rather new to Forex, I would not recommend MDP as your first choice – chances are it will be unprofitable for you in the first combinations you try and then you will give up and ask for a refund.
  As for an EA recommendation, I make a point of not giving such direct advice. That’s why I am running the EA Top page – everyone can check out the stats of each profitable EA and decide which are those that match their trading style and meet their expectations. On top of that, picture this: Forex Combo System was performing totally awesome up to some 4 months ago and had kept up that performance for over 1 year; at that point it would’ve been a great EA to recommend, yet it started going into a large drawdown right after that – the point of this small story is that there’s no predicting the future in Forex and even if I tried, it would be impossible to say which particular EA is going to be the best during the following months or years based on the data that is available now.
  Reply Quote
#466 written by sam March 27, 2012 (1 month ago)

Hi Birt,
I have heard that Scalping EA like MDP do scalping with more than 10 lots with MT4 is very dangerous and risky, is it true? It advise to use Dukascopy’s JForex versions and FIX API Currenex version? I don’t have any idea on that, do you?
Reply Quote
#467 written by birt March 27, 2012 (1 month ago)

Well, it all depends on the available liquidity and your latency – MDP trades might get too much slippage with 0.1 lots as well as with 10 lots. Trading MDP is inherently risky (well, trading Forex in general is), no matter the lot size.
Regarding using JForex or the FIX API, you’d have to have the expertise necessary to implement an EA that uses a similar trading logic using the toolset available for these platforms. Alternatively, you might be able to find a programmer to do it for you – if you can afford to trade 10 lots you can probably afford to pay several thousands for porting it. However, keep in mind that there’s no guarantee that it will work once you ported it and that the license probably doesn’t allow you to port it to a different platform to begin with. Sure, in theory it should do better than on most MT4 brokers seeing that in essence you skip a link but I don’t know anyone who actually applied it in practice.
Reply Quote
#468 written by Chris March 29, 2012 (1 month ago)

Hi Birt
For JForex it requires a Java solution. Or you have a bridge for example FTT?
Chris
Reply Quote
- #469 written by birt April 7, 2012 (1 month ago)
  
  Yes, for JForex you’d need to port the code to Java. I recall someone tried using FTT without much success, it was mentioned in a comment somewhere.
  Reply Quote
  - #470 written by Chris April 9, 2012 (1 month ago)
    
    hI birt
    I have tried with FTT but it doesn’t work well. Because the responding Times are to long. The same with Sleeping.
    Reply Quote
#471 written by cd April 7, 2012 (1 month ago)

This ea are not performing very well so far in 2012.
Reply Quote
#472 written by Garth April 20, 2012 (3 weeks ago)

I’m using MDP and since I have changed the settings to what Birt has, it hasn’t made a single trade in over a week now… Has anyone else found that it seems to have stopped trading?
Reply Quote
- #473 written by Remus April 20, 2012 (3 weeks ago)
  
  One week is not enough to draw conclusions, but if it helps, I have not had it trade for about a week now either (and I did not make any changes)
  Reply Quote
- #474 written by birt April 20, 2012 (3 weeks ago)
  
  Mine hasn’t traded since 12.04 (so a bit over a week). There’s nothing unusual, though – if it has no setups, it doesn’t trade.
  Reply Quote
Type your comment
You may use these HTML tags: <a> <abbr> <acronym> <b> <blockquote> <cite> <code> <del> <em> <i> <q> <strike> <strong>

Notify me of follow-up comments via email.
« Previous 1 … 3 4 5
Comment Feed for this Post

Million Dollar Pips live forward test

Buy Million Dollar Pips

Birt’s forward test

EURUSD

GBPUSD

USDJPY

Official forward test

User accounts

Backtests

Recent Reviews

Recent Posts

Email subscription

Social networking