Designing an Automated Trading Program

[expiated]

Granted, I was away when EURAUD began signaling that selling the pair might have been a mistake, but that doesn’t change the fact that I still found the loss unacceptable. I genuinely believe that it is possible, with just a couple of slight final adjustments to NPP, to take the system to a level where it can sustain a daily success rate north of 90%, avoiding mistakes like the one I made with EURAUD.

Having attempted to make such adjustments, the setup has lost much of the messiness it was beginning to evidence in its last configuration. My first trade to test the theories behind the subtle changes I made was to buy GBPUSD at around 1.2788.

 

[expiated]

I have a second theory I am testing by selling USDNZD at 0.6781. According to this second theory, I should exit the GBDUSD position now, while the rate is at 1.2783, so it is at odds with the theory that led me to execute the trade in the first place. It will be of interest to me to see which of these two theories prevails.

 

[expiated]

According to this second theory, I should exit the GBPUSD position now, while the rate is at 1.2783, so it is at odds with the theory that led me to execute the trade in the first place. It will be of interest to me to see which of these two theories prevails.

Thus far, Theory II appears to be winning the day, so I have therefore abandoned my GBPUSD position, with there being a good chance I will now sell the pair instead.

 

[expiated]

I have therefore abandoned my GBPUSD position, with there being a good chance I will now sell the pair instead.

Selling the pair instead was a good move. Theory II would suggest selling AUDUSD at 0.1797 at this time, even though this would not be a move I would make intuitively, so I'm going to give it a try and see what happens.

 

[expiated]

I have a second theory I am testing by selling USDNZD at 0.6781.

This same forecast model indicates that the structure of USDNZD is now such that the pair is no longer bearish, so my best hope at this point is to get out of the trade near break even.

 

[expiated]

I finally got my average profit trade back up above my average loss trade (by a factor of about 3:2) but my percentage of winning trades was way down today.

No modifications were made to any of my chart setups, and by now, this is as it should be. I would have hoped to have reached this point long ago. All I want to do now is get better at using what I’ve already got.

The nice thing, or one of the nice things, about this system is that I know immediately when I need to get out of a trade and there is no temptation whatsoever to remain in a losing situation. The forecast models yield extremely precise prescriptions based on market conditions, and as soon as the diagnosis changes, so do my instructions.

 

[expiated]

I now have a one-hour chart setup that is a spinoff of the 5- and 15-minute formats I’ve been using to trade for the past two days. Each day I trade based off these current forecast models is a day in which this (hopefully final) configuration is more firmly set.

I’m still long EURJPY, with high hopes of eventually seeing the rate hit my take-profit target, at which point my daily success rate will be at 83%—much better than yesterday’s mediocre results—which should also improve today’s profit factor. But the ratio of loss trades to profit trades is still not good today…about 3½ to 2.

 

[expiated]

As best as I can tell, there are no more adjustments to make to my scalping system. The chart graphics dance in perfect step with price, in a manner of speaking, but I cannot capitalize on the system because other demands on my time prevent me from monitoring positions on an ongoing basis, which is what’s required.

As a result, I’ve racked up an abysmal percentage of profit trades so far this week. For me personally, anything under 85% is unacceptable, and I’m WAY below that. The only saving grace is that my average profit trade so far is greater than my average loss trade.

Anyway, that poor percentage of successful trades forced me to return to my pseudo-swing style of trading again today. I exited both trades made earlier with a profit and just entered a third, but here too, I think there are no more adjustments to make, so right now I’m going to go begin crafting the initial language for a rough draft of an inquiry seeking someone to code an expert advisor for MT4 that will execute my scalping system for me algorithmically while I’m busy doing other stuff.

The code will have to be semi-generic so I can plug in the precise parameters myself, thereby protecting my intellectual property, assuming that this can be done.

UPDATE: So far, today's trades are more in line with what I expect of myself...

 

[expiated]

Thursday I settled on the strategy I want to test as an automated system, so I am now in search of a coder to create the expert advisor (EA) for me.

RESULTS OF THE SYSTEM I SETTLED ON, THURSDAY​

The expert advisor (EA) would enter long positions every time the yellow moving average (in the above image) crosses the blue moving average from below (marked by the yellow-green boxes) and then exit the positions as soon as the yellow moving average makes contact with the red moving average from above (marked by the blue boxes)…

…and then enter short positions every time the yellow moving average crosses the blue moving average from above (marked by the red boxes) and exit the short positions as soon as the yellow moving average makes contact with the red moving average from below (marked by the purple boxes). [One of the marks for a short position entry is missing.]

The moving averages are proprietary (my own intellectual property), so I’ll need to enter the genuine formulas myself after I receive the finished product.

The genuine code will be something along the lines of what follows below:

MathMax((MathMax(High,High[i+1])),(MathMax(High[i+2],High[i+3]))), so I’m planning on having the coder use standard moving averages to test the EA as s/he creates it...

(i.e., (Close[i+1]+Close[i+2]+Close[i+3]+Close[i+4]+Close[i+5]+Close[i+6])/6)

 

[expiated]

It seems Craig's List is no longer accepting payments via PayPal (I haven't used their service in several years) and I was not having much luck finding a coder via Upwork, so I decided to give Fiver a try.

I got a quick response, but then the guy said what I was requesting was not a custom EA, which is what I payed for, but rather, required the industry standard at the next level up for double the cost.

I informed him that I did not have the money at this time, so he should go ahead and cancel my order and refund my payment. I've never used Fiver before, so I don't know if this is going to be a hassle or not. (I use Upwork all the time, and their system puts the money in escrow so funds are not released until after the job is done—which makes a lot more sense as far as I'm concerned.)