New Year, New Ideas

So as the new year rolls in I am trying to learn new things that I find interesting and that could help me become more profitable. I have still maintained my relative out performance as of late, in the past 3 trading sessions I have been +4%, 0, +1.5%, the thing that is annoying is that on the flat day, I would have outperformed by 1% had I not taken a losing day trade on the futures. But I can't be that upset, i limited my loss at 1% and it could have gone in my favor and given me even better performance.

I had the privilege of speaking with Eric Utley who is a day trader and one of the two guys who runs Trading Addicts along with Jeff Kohler. I always enjoy hearing about real world experience, even if it is just to hear that they too take losses and make mistakes. He seems to wait around A LOT for a trade but said that he has a win % of 70%. I would wait around all day too waiting for those types of entries, I just wish I knew exactly what he is looking for that is such a high probability to him. If I knew I was taking a trade with 70% probability I could easily be mentally prepared for a loss knowing it was a minority of the time.

So I ran into a friend who is also into trading and he told me about an automated trading program for forex. I have always had an interest in these programs but it seems like unless you are at a hedge fund like Renaissance Technologies your program likely doesn't work. So far this program I paid 139 for doesn't work at all, luckily I am only using it on a demo account so it isn't real money being lost but I did lose 139 in paying for it. Supposedly there is a 60 day money back guarantee so I'll just go that route.

But besides that it seems like you should be able to automate trading which could be very handy if you work a regular job and at the same time had a program making you money. I am on a forum that is dedicated to these programs and most the things I see I laugh at. They have so many indicators and lines all over the place which I don't use. All I use are support and resistance lines, VWAP and sometimes pivot levels. It seems like if you have a profit and loss system that has a positive expectancy and a fairly simple order entry you could make some money. Of course much smarter people than myself have likely thought that too. Apparently the winner of a programming competition that turned 10k into 130k in 3 months has phd's in math and physics and his programming is based on Neural Networks so I am fairly disadvantaged. However as I have watched these programs run on the demo accont most of their entries are down right stupid and are at places where I definitely would not have manually taken a trade.

My idea would be to use a simple system that detects a trend change intraday based on say a 1m or 5m trend then compare it to the 30m trend and only go in the way of the 30m trend. Or just do a system based on a daily chart and trade the changing trends. The other more complicated system that I have an idea for would be to detect flag formations on various time frames and then take a trade once it breaks one direction or the other. You can see this happen on charts decently often, if you could determine with decent accuracy a flag and a break and use a stop loss of less than your profit you could possibly have a long term profitable system.

For example, say it is forex, and you have a flag that is about 50 pips in height, generally this is a continuation pattern and a break signals a resumption of trend for at least the height of the flag. Lets say I think it generally continues for 2x the height of the flag and once it breaks the flag it should not re enter the pattern, so I give it a stop loss of say 10 pips and my target is 100 pips. That means I could be wrong and get stopped out 90% of the time and still make money. Of course if I got stopped out 90% of the time I'd want to see if I was getting a lot of false signals or if my stop loss was too tight, but the point is theorteically in the long run it should make money. If you were able to be correct and reach profit 30% of the time you would make 230 pips of profit every 10 trades, depending on lot size and frequency that could be a lot of money.

We'll see if I can program these ideas into reality because it seems like people much smarter than myself can't do it but I will learn a great deal in the process and who knows maybe I'll come up with something. ;-)