Thursday, March 26, 2009

FTM Entry Time Study

Haven't written for a week because 1) I attended SXSW from last Thu - Sun (go Japanther go!) and 2) I have been in the midst of a 4-day drawdown and did not feel like writing. I gave back about 33% during this time period. Why was the strategy performing so poorly as compared to before? Good question!


I MAY have identified the reason why I was having so many consecutive losers: my entry time was whack. If you take a gander at the table above, you will find that, given the present sample, the FTM's winning % when playing individual stocks seems to drop precipitously when the first entry is at made 10:50am or later. The market has been so strong recently that stocks were not pulling back as early in the morning as they were before. When they did, that is, after 10:50am, they had more or less blew their loads for the morning and had no steam left. This phenomenon does not seem to be isolated to the last four days however: looking back I found that many of my losers occur on first entries after 10:50am. So we will see going forward if avoiding first entries after 10:50am will also help me evade a rapid drawdown such as the one I recently experienced. Most likely I will need to collect more data until I can be assured that this phenomenon persists.

Peace & love.

And for you metalheads out there!

Wednesday, March 18, 2009

Present Trading Plan

Since I have had trouble the last few days with entering positions that do not fit my criteria to the tee, I wanted to explicitly note my current trading plan, so that I know from now on exactly what plan I will follow until repetitive experience teaches me otherwise. The current state of the plan--as I commit to following it--is:

MY TRADING PLAN
  • I run screens on Yahoo! Finance between 9:50 and 10AM to find very weak/strong stocks, i.e. > 3% off open.
  • I trade from 10:10 - 11:45AM.
  • I place trades off 5-min charts.
  • I take only Floor Trader Method setups, such that:
  1. Pullback pierces Bollinger Band
  2. I enter on a break of the piercing bar
  3. Pullback is on decreasing or constant volatility
  4. At least one of the EMAs is rising
An example of a Floor Trader Method setup.

  • I move my stop to breakeven when price advances 2X entry setup.
  • I exit my entire position on breaks of levels formed on 2-minute chart if stop has been moved to breakeven (Home Run exit). See an example.
  • I risk a fixed amount of equity per trade.

Monday, March 16, 2009

FTM Equity Curve and some trades


Above is the equity curve by trade for the FTM over one month. The first two weeks I spent getting used to the setup and then I kind of fell into a groove. I wonder over the long-run what the performance characteristics of this setup are, i.e. win%, avg. winning R, expectancy, etc...

Went 2-0-0 today with avg. win. R of 2.50. Being really selective, thus am not in some winning trades, and not in some losing ones as well. The ones I do pick though, I feel have a good chance of generating many Rs, cf. BG trade below. I chose not to take the ITW, STT, or second FE setup because either there was an increase in volatility, the stock was not coming off a new high, or the MAs went negative prior to the entry. Although a stock coming off a new high shows strength, I do not know if it is necessary criteria for a long setup. I will have to further investigate that because I do not want to be passing on good trades.

Best to you!


Thursday, March 12, 2009

Fresh equity curve & FTM performance update


So I have verified that my code for the Exponential Crossover system is entirely correct and have begun to experiment with the system's variables a little, just to get used to how they generally affect system performance. Above is an equity curve generated for Comex Gold between 1975 and 2005. The system was always in the market, long or short, and parameters were as follows:

SlowLag = 260
FastLag = 50
Heat = 4% (percentage of closed balance risked on each trade)
ATR Multiplier = 2

The max. drawdown during the run was a rough 75% with an ICAGR (i.e. compounded return) of 12%. That's not a very good ratio. Over 30 years, the system turned $1M into about $30M.

Yesterday, I lost money trading. My record was 2-4-0 and I lost 1R, which isn't too bad. My record probably should have been 2-2-2, but I did not move my stop to breakeven on two trades that came within 1 cent of 2x the initial setup. Had I done that, I would have been up 1R yesterday. Notwithstanding yesterday's small loss, I started today feeling pressure to make money, which for me is not as profitable of a mental state. I got angry a few times, especially when I missed a good JPM trade, because I was afraid to lose money, to fail. My emotions were in vain because although it wasn't a huge day, I did just fine, even without the JPM trade. My record was 3-1-2, but my average winning R was only 1.51. That higher win% helps! Today I made 4.86R.

Looking at my weeks I am well on my way to my second profitable week in a row. Woohoo!

Tuesday, March 10, 2009

Still truckin' Mach II

Despite being up 20 handles by 10am, could not find many decent setups. My record was 2-0-0 with avg. R of 2.65. My two winners are below: you'll notice that the HOT kept going higher. After he consolidated and then stopped me out on the downside, there was another buy signal, which I did not take, simply because it was the afternoon. The market is very strong today, so maybe trading all day would be okay. However, I would like to work on my programming.

There were some winners that did not quite meet my setup criteria; but, then there were also several losers that did not as well.

Peace!

Monday, March 9, 2009

Still truckin'

Record today 5-5-0 with an average winning R of 2.42. Had five straight losers in the morning, but made up for everything and much more with one--well, actually two--trades in WFC. (See below.) Observation for today being that obviously, losers are an inevitable part of of this game, but why not try to eliminate as many was you can? So far, this strategy produces about 50 - 60% winners. Don't know if I can break that 60% barrier or not. Regardless, with big 4R & 5R winners, and sometimes much more, the 1R losers don't matter that much.


Also, completed the Exponential Crossover project (see previous post) and am now pretty much finished with the Support and Resistance project. My backtesting software is coming along and I am enjoying programming immensely! I trade from about 10am - 12pm and try to work on programming the remainder of the day.

Peace!

Wednesday, March 4, 2009

using System.FloorTraderMethod;

private void CalculateATR()
{
decimal atr1;
decimal atr2;
decimal atr3;
decimal maxATR;
for (int i = 0; i <= Date.Count; i++) { if (i == 0) ATR.Add(High[i] - Low[i]); else { atr1 = Math.Abs(Close[i - 1] - High[i]); atr2 = Math.Abs(Close[i - 1] - Low[i]); atr3 = High[i] - Low[i]; maxATR = Math.Max(atr1, atr2); maxATR = Math.Max(maxATR, atr3); ATR.Add(maxATR); } } }


Last night, after much delay and excuse-making, I finally began programming the Exponential Crossover project on the Trading Tribe website! I am working in C#. I am amazed how foreign programming language used to seem to me, but now I am certainly becoming handy with it. I'll post some of the project: it's so apparently technical, but in actuality very organic. I like that paradoxical element about it.

I quit trading today around 12:00pm because I did not get enough sleep last night and I caught some great trades during my sweet spot from 10:30 - 11:30pm. I will have to figure out how to trade the afternoon because my present strategy of playing the strongest/weakest names for the day do not seem to work nearly as well after 2pm. Also, institutions usually put their orders in around 3:30pm, which does not give much time for a pullback I need for an entry.

Did well today: 6-4-1 with an average winning R of 2.10. Hit a home run in CAT with a 4R trade and should have had a monster in WFC (probably 6.5Rs) but put my stop too close and only made 3R. I will look into giving stocks a buffer above/below the pullback level on the 2-min chart, so that I do not get shaken out. Also, have found that more "orderly" pullbacks generate winners more often and often BIGGER winners. If you take a look at some of the lackluster trades below, such as MDR, MEE, and VLY, the pullback directly prior to the entry is pretty "messy" whereas the great trades, such as CMC, CAT, and to some extent WFC, are more or less streamlined. I'll start to keep an eye out for that.




Also, wanted to show how I am booking profits with the home-run exit strategy. CAT below is an example. I use a binary exit and move my stop up/down based on levels developed on a lower timeframe, which is a 2-min chart in this case. The ascending red lines show the movement of the stop.


Best to you all!


Tuesday, March 3, 2009

Still refining...

Record today was 3-3-0 with an average winning R of 1.61. Not too sexy, but did finish positive. Two positive days in a row! Yippee! I did not book any for 2R today but instead went for home runs, i.e. moving stop to breakeven after 2R target is breached and using a lower timeframe to determine levels to move stop down to. I will be trying this method out while paper trading to see if the frequency of outsize winners makes that exit more profitable than an R-based, singles-and-doubles exit strategy.

Also, if the signal bar (BB-piercing bar) has a greater range than previous bar, that MAY decrease the odds of winning. This makes sense though in light of the nature of increasing volatility and its effect on the odds of winning. It seems that any abnormal increase in volatility makes a play suspect. Also, increasing volume in the direction of the pullback may also decrease the odds of winning. I will be investigating these further.

Here are some of today's highlights.

Monday, March 2, 2009

Refining the Floor Trader Method

Slowly increasing my percentage winners. Today my win percentage was 40% with an average R of 2.90. I would like to be 60 - 70% correct on average, if possible. Since I have included the Bollinger Bands in my method, I have neglected the nature of the pullbacks, i.e. how many bars are included and changes in volatility. Looking back at my losers from today, I see that they most often occur on increasing volatility or on pullbacks with large numbers of bars. This is not new information, but I need to be reminded after including the Bollinger Bands for entries.

Straightforward Bollinger Band entries seem to work much better than categorizing entries in relation to the EMAs, e.g. Level 1, Level 2 & Level 3: BBs get me into more low-risk trades in the direction of the trend. After the BBs generated a number of false signals on the 3-min chart, I am only entering on the 5-min chart, which seems to be more reliable.

As for exits, today was a very bearish trending day and moving stops down would have been exponentially more profitable than booking at multiples of R. (See the WGOV trade below.) I am considering trying to hit home runs all the time, that is, once I am in the money 2R, moving my stop to breakeven and then just letting the stock work for me. I do not know if I can handle the subsequent gyrations in my P&L curve or not, nor is it apparent if this exit strategy is more profitable in the long-run. I will have to see if > 2R winners occur with enough frequency to justify trying to hit home runs all the time.

Here are some winners & losers from the day. You can probably figure out which is which.