New EA for Backtesting -

All, 

I am looking for a simple EA to be coded based upon the VEGAS system. I have attached a link to a detailed description of the system. 

However, i would like many of the variable to be 'inputs' to enable modifications and tuning of the EA.  I have searched for this EA but cannot find it -  

Your help would be appreciated -  the basic EA should be pretty simple to code.

I have also cut & paste the detail below of the parameters and the entry / exit points etc.

 The document below is from the Vegas team 0 I have added some RED words to help target the EA.

More background;

http://www.forexfactory.com/showthread.php?t=4113

thanks

 Darren

(London)  

Step 1.

First, you need a charting service. Since most all electronic trading platforms have charts with technical indicators, this shouldn't be a problem.

Create a 1 hour chart on whatever currency pairs interest you. Barcharts or candlesticks really make no difference. Overlay on this 3 things: 1) a 169 period [1 hour] ema [exponential moving average], 2) a 144 period [1 hour] ema, and finally 3) a 12 period  [1hour] ema.

 I would like to test this on all timeframes ie NOT just 1 hour

The 144 and 169 ema's create what I call the "tunnel". The 12 ema is an extremely valuable filter that you will want to have there all the time. I will talk more about this in the filter section.

Step 2.

Memorize or write down and keep next to your trading screen the following fibonacci number sequence: 1,1,2,3,5,8,13,21,34,55,89,144,233,377. For trading purposes, the numbers of  interest are 55, 89, 144, 233, and 377.

Step 3.

Wait for the market to come into the area of the "tunnel". When it breaks ABOVE the upper tunnel boundary, you go long. When it breaks BELOW the lower tunnel boundary, you go short.

I would like to have a variable of the level below / above that the entry is made for example ENTERLEVEL  so either +10  or -10   to be variable

Step 4.

Stops and reverse are placed on the other side of the tunnel.

Optional 
 

Step 5.

As the market trades in your direction, you take partial profits at the successive fib numbers respectively, with the final portion of your position left on until one of the following conditons occur: 1) market hits the last fib number [377 pips]  from the ema's, or 2) the market eventually comes back to the tunnel and violates the other side.

Example: GBP/USD is trading at 1.8500. The ema's are as follows: 144- 1.8494, 169- 1.8512. The market breaks 1.8494, and you sell at 1.8492. Your stop and reverse is now at 1.8512. Over the following hours, market starts to go down. 40 minutes after you put position on, cable is at 1.8440. You can use for computation purposes either tunnel boundary or the median of the tunnel. Ema's are still the same, so if you use the median, 55 from 1.8503 is 1.8448. You should have taken part of the position off at 1.8448. Market does nothing rest of day. Stop can be moved down to protect position or left alone at tunnel.  You are now looking for price to be 89 pips away from the ema's. Since 55 was already passed, it no longer concerns us in this cycle. A couple of days later, cable is at 1.8300 and the median of ema's is 1.8410 [1.8400 - 1.8420]. You should be out of another portion of the position at 1.8321. Market bottoms here and in the next 2 hours, cable screams to 1.8535. Your remaining short position is covered at upper tunnel boundary of 1.8420, and you are now long from this point as well. Since you are long, you would now take partial profits at 1.8475 and 1.8509.

I would like all of these to be variable as well -  also a traditioial Trailing loss might be better to be used.
 

This is a fairly typical example. 

If you were to just stick to this basic model, you account would grow very well over time.  Las Vegas was built with far fewer percentages in the casino's  favor.

In case you haven't figured it out, this model cuts your losses very short. By definiton, you can't lose very much on a single trade from your initial entry position.. On the other side, you take some quick profits at the 55 level which satisfies the scalper in you, and you have positioned yourself for bigger profits in the long run should the market keep going in your favor. By definition, you are letting profits run.

The Achilles heal of this model is when the market chops around the tunnel and gets you in and out multiple times for small losses. I will cover how to deal with this in the filters section.

That's it. This is the model. Fairly simple in its design, and easy to remember. Has all the things every local wants in a model, except the quick 2 pip scalps, which you can't do anyway. Cuts losses, and lets profits run. Yet for its design simplicity, the thought behind is more complex. Time to talk about that.

PART 3. THE FILTERS

Filters are used to increase overall profitability and/or reduce overall losses. If a filter does not do one of these two things, then I do not use it. What good is a filter if it raises your profitability by 10% but only gets you into 1/3 as many trades? What good is a filter if it reduces losses by 10% - 20% , but also reduces profitability on every trade by half? I think you get the point.

Here are the filters the vegas team uses. [Yes, I have a team. There are 3 of us. We trade GBP/USD, USD/CHF, and the S&P e-mini futures contract. Each has a specialty. Mine is GBP/USD. We are each responsible for our main pair. One of us is always at the screen when markets are open. Positions are covered by other partners when away. We only tunnel trade.]

1.)

Put the 12 ema [1 hour] on your screen with the rest of your indicators. When everything is at the same price [tunnel, current market price, 12 ema] sit up and take notice. When the market breaks away from the tunnel, there is a very high probability of a strong market move coming.  I don't need Gann, because this gives me time, the square of time, and price all in equilibrium. When it breaks, it goes.

 Again,  i would like this to be variable -  so 12 EMA but can i test with other EMA's

Need proof? Well, go back on your favorite currency pair and check it out. In the first quarter of 2005, this filter alone produced 20 trades, 19 which were profitable in USD/CHF. In fact, as I write this, 1 trade is still on from about 3 handles ago. Since I am not responsible for Swissy, I'm not the guy pushing the button, only monitoring it when I'm at the screen [changing stops when needed, etc.]. But, the position is still on.

All currency pairs i would like to test 

This filter is so profitable, we increase the size of our trading position when we see it develop and then happen.

When you go back and check it out, you will notice many times how it just misses a move by a few hours. It is an extremely profitable filter.

We also define "same price" as being within 5 pips or so of being equal. Sometimes it turns out the signal is exact, but I don't think you have to split hairs on this. Within 5 pips is good enough for us.

2.)

We do not initiate new currency trading positions based on tunnel trading during the Asian time-frame. Anything between 5pm NY and Midnight NY is ignored for entry of new positions. Positions that are on are monitored as normal, i.e., everything else is the same. We will take profits if fib levels are hit. If we miss a move, then we miss a move. A missed move is just an opportunity cost. Chop-chop in Asia will eventually cost you more money than it is worth.

3.)

News days that can have a significant affect on prices are ignored. That's right, we skip them for entry of new positions. Currently there is only 1 day per month which qualifies, and that is US Non-Farm Payrolls [NFP] which comes at 8:30 am NY time the first Friday of each month. Positions that are on are monitored as normal.

4.)

When the tunnel is very narrow [most of the time], do not just put stop on the other side of tunnel. If you do you get whipsawed to death. Use the hourly charts and the most recent hours of support and res. to make the call.

If you are a newbie to trading, you will find this to be the most troublesome filter. If you are not familiar with trendlines, triangles, flags, pennants, and support and res. levels, then go get the eduation and come back. Simple but necessary advice.

I don't mean to infer that just because you know this technical stuff it's going to be a walk in the park. It's not. Let's make one thing perfectly clear. EVERY model has its vulnerable spot that seem to increase losses. For tunnel trading, this is one of the scenarios. Putting in the right stop is an art, not a science.

5.)

We look for clean moves [1 bar] through the tunnel. This means your into profits almost from the get-go. You will not always get the clean moves. The longer the market stays in the tunnel chopping around, the higher the probability our entry decision will be made on a break of support or res. instead of the tunnel boundaries.

6.)

We do not trade minor [contra-major] trend signals in a strong up or down market price trend. If the GBP/USD is in a strong price uptrend, we will not initiate new short positions on a break of the lower tunnel boundary. Why? Because the probability of success in getting past 55 from the ema is not very good. Past history tells us that, so I'm not looking to be the hero here and say "This time it's different." When market comes back through the tunnel on the upside, we will get back in on the long side.

If I have to tell you when the market is in a strong price move, I don't think you have been paying attention to the price movements of late.

In a range-bound market, which we define as a market between 3 - 5 handles [or lower] in a 5 week time-frame, we trade both sides.

Now, that's all we use. Can you use more? Can you invent your own? Can You change some of the definitions? Yes, absolutely. Invent your own filters, use an Elliot Wave filter, anything you think will help your trading.

V.  SUGGESTED MODEL SYSTEM

Do I really need to mention money management?

I didn't think so.

At a minimum you should be able to do 3 units to implement tunnel trading. Use the 55, 89, and 144 levels to take 1/3 off at each level. If you can do 4 units, use 55, 89, 144, and 233. 5 units is the preferable level, and you use 55, 89, 144, 233, and let one unit ride until crosses over tunnel boundary or it reaches 377.

Of course, you can make your units any size you want. For smaller traders, a unit size may be 10,000. If you do not have the money to trade 30,000 of something, then I would advise you to save up and come back when you do. If your account has $2,000 in it, you can easily implement tunnel trading with 10k units.

One of the greatest advantages of this model is its flexibility in its design to allow you to choose the level of risk/reward you desire in trading. You can make this as aggressive or as conservative as fits your style. I will give an example of each. These are just examples, I'm not saying you have to do this. I'm only giving you these two to stimulate your brain. In the following day and weeks I am confident you will find an appropriate level for yourself.

Example 1 - Very Aggressive

Tunnel is pivot level for buy/sell. Above tunnel, buy breaks, sell at fib numbers. At 233 an 377, fade the move for retracement. Below tunnel, sell rallies, buy at the fib numbers. Use previous fib numbers in the move as stop loss points. This is very aggresive, and woul be appropriate for very short-term traders who have a time-frame of day-trading.

Example 2 - Very Conservative

Uses basic tunnel system with 12 ema. Only initiates on this signal. Looking for best possible probability trade. Willing to give up more profitability in return for less risk. Trades three units. Uses fib numbers 55, and 89 for 1/3 each. Leaves the other unit on until 233 or market price crosses over tunnel boundary. Allows trader to catch short-term [1-5 day] profit points, and also allows him/her to ride the major trend if one develops.

Like I said, these are just two of an infinite number of risk/reward senarios you can develop using this model. This is not some rigid system, where you have to do this or that. It is adaptable, with no right or wrong answers. This is why many locals from soybeans to bonds to gold and silver, oil, etc. use it. I've seen some people who have transformed this into a model you wouldn't recognize without knowing what tunnel trading offers.

When you get right down to it, once you have adapted it into your own trading style and personal risk model, tunnel trading will give you all you want. Momentum to catch the bigger moves over time, early profit points that allow you to catch short-term movements, and the lowest risk you can possibly have in a trade, because you are only risking 10 -25 pips on each trade. If your odds of success on each trade were 50-50 [they aren't this low], over time you would make a fortune. If you don't believe me, then do the math.

Precisely because of this flexibility tunnel trading is the best model I have ever seen.

VI. YOUR HOMEWORK

You really need a good charting service to go back and look at the history of the currency pairs you trade. I have mentioned several times of  fxtrek on the forexnews forum. If you have another, great. But for those of you who only get their charts from the trading platform where you trade, most will not allow you to bring up historical data. You can get a free 7-day demo of intellicharts at www.fxtrek.com. They only do forex charts. They offer spot forex on dozens of currency pairs, with hundreds of technical indicators over the last 30 years, with any time-frame you want. Therefore, you can go back and look at 30 years of 1 hour charts on whatever pair you wish. After 7 days the price is US $100/month.

If I was in your shoes, I wouldn't make a trade without some kind of validation that what I have said really is true. That is why I am asking you to do some kind of historical  homework on the 1 hour charts. You can see for yourself what tunnels do, and why the fib numbers are so important.

VII. I STOP BUT IT'S NOT THE END

I could ramble on about a lot of things regarding the tunnel method, but you now have the basics to get started. Once you get your trading style and risk model defined, you can start thinking about additional filters and signals for refinement. Of course, you are free to use the model the vegas team uses as well. I would like to end by giving you my e-mail address. It is trafficcap@hotmail.com  Please feel free to e-mail me anytime. I will respond as quickly as life allows me.

I hope I have been of some help. For some I hope this has opened your eyes to a model that delivers. For others, I hope you have picked up an idea that may be of use in the future. For those who think I'm nuts and full of ****, that's OK too. I think I may have stood next to you in the pits. Your screamings have always provided humor.

Best of trading,

Vegas