Chris

Tradesight September 2010 FOREX Results

One of the things that I have resisted in the past is "track record" type of stuff, especially in the FOREX market where so much is dependent on which hours you have available to focus on the market. It's not real fair, in my opinion, for you to try to match my trades if I'm staying up until 3 am your time and you have to go to bed at 11 pm. That's why we prefer to teach people how to trade rather than just to have them follow the trades.
However, at some level, there has to be evidence that the calls that we make each day in the Messenger using the Levels create a working system, even if the idea is that you should tweak that system to work on your own time.
So, I spent a lot of time thinking about what would be a fair way to track the results. When you do tracking, it is important to lay out the exact methodology of how you are tracking the results so that someone can look back going through the reports and come to the same number. At first, I was considering factoring in everything about trade size and time of day, but again, that starts to factor in items that mean that some people might not be able to follow directly.
And then I came to a very basic statistical conclusion. In the end, what we just want to do here is count pips with the LEAST complex set of parameters. Conversations about trade size, etc., is all important and should be overlaid into these results. But without making everything a matter of "well I was full size on this one but you should have been half size," we can start by just saying "What would have happened if you would have taken all of these calls, sold half at the first target, kept the stop where it should be, and adjusted based on what is in the Messenger?"
And once I stopped trying to make it complicated, well, it got uncomplicated.
Here are the rules.
1) Calls made in the calendar month count. In other words, a call made on August 31 that triggered the morning of September 1 is not part of September. Calls made on Thursday, September 30 that triggered between then and the morning of October 1 ARE part of September.
2) Trades that triggered before 8 pm EST / 5 pm PST (i.e. pre Asia) and NEVER gave you a chance to re-enter are NOT counted. Everything else is counted equally.
3) All trades are broken into two pieces, with the assumption that one half is sold at the first target and one half is sold at the final exit. These are then averaged. So if we made 40 pips on one half and 60 on the second, that's a 50-pip winner. If we made 40 pips on one half, never adjusted our stop, and the second half stopped for the 25 pip loser, then that's a 7 pip winner (15 divided by 2 is 7.5, and I rounded down).
4) Pure losers (trades that just stop out) are considered 25 pip losers. In some cases, this can be a few more or a few less, but it should average right in there, so instead of making it complicated, I count them as 25 pips.
5) Trade re-entries are valid if a trade stops except between 3 am EST and 9 am EST (when I'm sleeping). So in other words, even if you are awake in those hours and you could have re-entered, I'm only counting things that I would have done. This is important because otherwise the implication is that you need to be awake 24/6.
You can go through the reports and compare the breakdown that I give as each trade is reviewed.
Each month, I will copy those instructions, then give a basic breakdown, and then discuss the specifics. We also will include a "worst losing streak" number. This month, we had a period late in the month losing 5 in a row. I always focus on that because people need to understand that when figuring out their trade size. 5 losers in a row = 125 pips that you are underwater out of the gate if those were the first trades you did.
Tradesight Pip Results for September 2010
Number of trades: 38
Number of losers: 16
Winning percentage: 57.8%
Worst losing streak: 5 in a row (September 27-28)
Net pips: +308
One other thing we want to consider in reviewing the calls each month is the overall environment, and we have a very easy way to consider this, which is to look at the ADR (Average Daily Range) changes during the month. Remember that the ADR takes the daily range (high minus low) and averages that over each day for the last six months or so. Whether the ADR moved up or down during the month tells us whether ranges were, on average, better than average or not.
For example, on August 1, the ADR on GBPUSD was 173 pips. By the end of August, the ADR on GBPUSD was 167. A 6 pip drop in the ADR over one month when you are averaging six months of data is pretty significant. That makes sense given that August ranges were the worst of the year, as expected.
So how did ranges fare in September? At the end of the month, average ranges were all within 3 pips of where they started the month. Meanwhile, the GBPUSD, for example, moved up 2. So basically, we can read into this that ranges were slightly better than they had been in August, but not noticeably so. This fits with the fact that I haven't yet adjusted my size back up because we never really saw even a week where the pairs traded above ADR 3 out of 5 days. In many ways, this activity still feels like summer volume activity, maybe just a little bit better.
One of the reasons that I decided to start tracking this is that I know once we have several months of these logged, there will become some obvious statistical relevance to the data. For example, I would anticipate that in a month where ranges improved, our win/loss ratio will push up closer to 65% (might even get higher sometimes) over September's 57% and our net gains should be higher.
Remember that trading systems aren't based on luck, they are based on doing as close to the same thing over and over that you can based on a system that works and letting it play out. A winning system nets you money not just because of what happens in one week or one month, but what happens over an extended period of time that covers good and bad environments for the market.
And now, on to October and the last quarter of 2010.

by Chris