swordfish

Scalper72,

You did nothing wrong. It's the nature of the EA. You just need to set the right settings to either avoid these trades, delay these trades or recover losses quickly. I am also using Apari UK (5-digit) on Indo Run 1.5 with 1.5 conservative set file. This particular forward testing as shown is using Martingale true. The chart shows 3 clean trades in the Asian session last night,1 single sell order in the UK session followed by a basket of 3 buys. All closed out with a profit. I would love to see your screen shot so I can offer you a suggestion of settings.

http://i162.photobucket.com/albums/t269/julianblo/IndoRuntrades12_17.jpg

tomislav,

On Nov 10 in Canble Run thread, you stated, "Yes my friend I can relate I have been very happy as of late but yesterdays big price move by the US selling bonds was totally unannounced and look what has happened to GBPUSD, EURUSD etc. One really has to stick to trading this EA between 21:00 - 6:00 GMT."

I assume your comment was on Cable Run modified. What is your forward testing experience with Indo Run 24/5 trading using conservative settings?

I only have 2 days of forward testing on Indo Run. I lost track of what I changed in settings. Now I am back to conservative defaults.

Thanks,

After watching this FractalS&R for a while, guess what, Cable Run logic works just fine for trend trading. Those "steps" are formed upon the formation of the fractals. So is the entry point. If there is an initial breakout of support (red dots), indicating trend is potentially down but we don't know how long. Sell as soon as the resistance (blue dots) is formed. In the above example, you would have 8 sell orders. All got stopped out when price closed above resistance. 7 wins and 1 loss.

This post is not trying to mess with Indo Run. As stated in my initial post, fixed steps get slaughtered big time in a trending market. You either need to increase the step size in the day session or make it adapted to price action.

http://i162.photobucket.com/albums/t269/julianblo/ScreenHunter_03Dec160711.jpg

GBPUSD drops like a rock in day session. Perhaps for day session, we should reverse the logic to trade with a trending leg. Try this Fractal indicator on your 5 min GBPUSD. Sell the breakout of the support and add to sell as long as there is a breakout. Vice versa for buy. Steps are based on price (blue and red dotted lines). One serves as TP or SL for the other. You would have 11 sell orders by now.

http://www.multiupload.com/4FM7DJQRYD

http://i162.photobucket.com/albums/t269/julianblo/ScreenHunter_04Dec151328.jpg

Forward testing indo run 1.5 at FXDD for 2 days, conservative set file.

A lot of buy and sell limits but none was executed.

So far no trade, is there anything wrong?

There was a basket of 5 sell orders starting GMT 15:00 on Dec 13. All closed out with a net profit at GMT 00:15 on Dec 14. I was also using the conservative set file (excep GMT offset=1). Broker is Apari UK. FXDD is 4 digit, so the ATRon filter may act differently.

Paracool,

Night Walker EA is just a modification to Cable Run with step being dynamic, TP dynamic, and ATR filter for entry. I am not using any of the modificaitons. In fact, I am still running Cable Run 2 for my live account. My manual strategy worked out nicely last night. A buy order went into draw down and then I added another buy (my rescue trade) this morning 60 pips below. The rest is history. The attached paint-bar indicator from Joe Simpson may help determine when to add the rescue trade.

http://www.multiupload.com/PE3KILNXGY Try it on the 5min chart.

I am demo testing Expat's Indo Run1.5.

Ok, Try multiupload.

http://www.multiupload.com/VUAFU2VA9R

Backtesting might differ from forward trading for this strategy. Spreads, brokers, trade position, opening/closing, etc. Doesnt matter if its MT4, 5 or whatever... Backtesting confirms so far for me the general performance.

 

24hr trading or not is the individual choice. GMTOffset is needed for news or day / hour based filters.

 

Dynamic "Greed" I should rename it... Read back and you will find some hints on it. It will help reduce the DD and trade up as it can.

 

Cheers

Expat,

I read the whole thread in one sitting. Now I understand your Dynamic Profit, which is pretty mild as compared to what I was doing.

In my opinion, the best way to reduce DD is to reduce the number of open trades in the basket (Duh!). The only way to reduce number of orders is to increase the step setting to catch the perfect top or bottom. If you look at the chart at different sessions, GMT 21-7 (or 22-8 standard time) usually has mild swings. It works ok for 17 pips step as in Cable Run. Once it passes this window, the swings becomes large ( with trending swing legs). A step of 30 or even 50 pips will help reduce the open orders. I do this manually at GMT 8 if there are open orders (with drawdowns of course). I place a market order when I see the step is larger than 30 pips. I also apply martingale after this window. So far, I can always close out open trades during the day session. I am hoping an EA can take care of this while I sleep (GMT 8 is 3:00 in the morning for me). But immediately I see the problem with fixed grid for this strategy. For example, what if the price turns around at 29 pips step? I would miss a potential rescue trade. Let me introduce a concept for everyone to think about.

When I watched Cable Run working, many times I saw it almost touched the limit order but turned around and missed the trade. Also, it often passed the limit order too much and got filled at a disadvantageous price. Often, it almost made 15 pips but missed it. Granted, this is a grid system and sometimes the entries and exits were in fact excellent. But there is room for improvement. I compared Cable Run entries with entries based on stochastic (15, 3, 3). As you know, stochastic works best in a choppy market to catch oversold and over-bought levels. I was surprised by the performance of this simple and well-known concept. Please take a look at this PowerPoint file based on 5-min charts during GMT21-7. I call it the Pinball system because it bounces between oversold and overbought levels. Trading rules are in the first slide. As compared to the grid system, here are the pluses 1) it will not miss a trade, 2) It will take on additional orders as in Cable Run to leverage your losing positions, 3) it will not miss the profit target, 4) it will catch the beginning of a trend and the profit can be maximized if we implement a trailing strategy and 5) best of all, it will take a small loss so you can trade another day.

Download Pinball Visual Back Test.ppt from FileFactory.com

Try this if the link doesn't work:

http://www.filefactory.com/file/b4a3010/n/Pinball_Visual_Back_Test.ppt

You may count the pips during this 12-day period. My Cable Run didn't perform this well.

Below the backtest of my current live and demo set (see first post). There is more in it. e.g. with Dynamic Profit still to try...

 

http://img826.imageshack.us/img826/5996/indorun15tsdnewsfilter2.gif

 

Symbol GBPUSD (Great Britain Pound vs US Dollar)

Period 15 Minutes (M15) 2010.01.03 23:00 - 2010.10.08 22:00 (2010.01.01 - 2010.10.09)

 

Bars in test 20191 Ticks modelled 12393890 Modelling quality 90.00%

Mismatched charts errors 0

 

Initial deposit 10000.00

Total net profit 12281.01 Gross profit 23615.12 Gross loss -11334.11

Profit factor 2.08 Expected payoff 7.45

Absolute drawdown 666.54 Maximal drawdown 3404.38 (18.22%) Relative drawdown 22.24% (3045.92)

 

Total trades 1649 Short positions (won %) 778 (64.52%) Long positions (won %) 871 (63.61%)

Profit trades (% of total) 1056 (64.04%) Loss trades (% of total) 593 (35.96%)

Largest profit trade 245.16 loss trade -164.25

Average profit trade 22.36 loss trade -19.11

Maximum consecutive wins (profit in money) 14 (455.96) consecutive losses (loss in money) 10 (-826.57)

Maximal consecutive profit (count of wins) 1429.74 (12) consecutive loss (count of losses) -826.57 (10)

Average consecutive wins 3 consecutive losses 2

 

Cheers

Expat1967,

Many thanks for your contributions. This is my first time following your thread since I've been busy trading the Cable Run. I've been working with a programmer friend to improve Cable Run and found that back testing trades do not match live trades. He had to re-write the EA in MT5 and do back testing in MT5 to get an accurate performance results. Any comments?

Also, is Indo Run meant to be running 24 hrs and letting filters decide when to trade? If so, then GMToffset is irrelevant, isn't it?

One comment on dynamic TP. I call it the "greed" factor. I think increasing TP with additional orders will give more profit in back testing. In live trading, I've seen many trades couldn't close out because TP was set too high by the EA.