End of Day Journal (8-26-10)

Summary:

Un-believe-able.

Finished today at -$165 (same gig):

Comes out to -$0.825 per share.

Today I did something different on the EOD strategy application. Again, I finished the day thinking that there was no way the EOD app was going to be positive - I thought I had I traded close to the plan. First I did like I always do - cleared the charts, spread out the day in front of me, drew S/R and then flagged trade signals. It took me about 5 minutes and after an even briefer review, I knew that what I had just drawn was going to be profitable.

No way.

I decided that the entire day's data on the chart must be skewing my perspective. I erased everything, scrolled the chart back so that nothing beyond the 10:00 a.m. was visible, and took one candle at a time marking S/R and trade signals.

I want to offer a comparison between what I drew live and what I drew candle by candle (the top one is live, the bottom is candle by candle delineation):

Here's how they compare (starting from the top):

Live version:

1. 44.28
2. 44.08
3. 44.02
4. 43.98
5. 43.90
6. 43.83
7. 43.79
8. 43.70
9. 43.58

Candle by Candle:

1. 44.28
2. 44.01
3. 43.93
4. 43.83
5. 43.79
6. 43.75
7. 43.70
8. 43.50

 And for the sake of full disclosure, indicator S/R (NT's current day OHL, and prior day OHLC) are placed at 44.32, 44.26, 44.24, 44.07, 43.50, 43.35, and 43.27. This accounts for the 44.08 in the live version (I didn't draw lines over the NT indicators on the candle by candle, but forgot to turn the lines on for the live image). The 43.58 was a new low that I didn't mark on the candle by candle but I did trade on.

Ok - all that was to convince myself: S/R == S/R, seemingly no matter how I do it. I can be consistent with this. I am guessing that I could take anyone off of the street and get them to draw something very similar within 5 minutes.

Here is how the EOD trades came out:

$1.97 per share, for a difference of $2.80 per share.

Pfft.

Where am I going wrong? Is the EOD application some hopeless fantasy?

As it turns out, I matched the EOD position 68% of the time today (calculations are at the end of this post); which amounts to missing only a couple of key decisions, nearly all of them ones that gave me pause today while trading. Keeping market context would have saved the day ('iffy' signals mean less if they go against market sentiment, and they are stronger if they go with market sentiment). But all in all, I actually nailed several of the signals. Details follow.

I took a slightly different approach and focus today, and it worked very well for most of the day. I tried to tune into the candle action for the last 30 seconds of the candle and mentally reviewed possible scenarios and how I would respond to each one. For example, if the candle was nearing support, I would envision how the candle would have to finish in order for me to reverse. At candle change I was ready to go. This allowed me to basically ignore anything in between signals and trade relatively emotion free: I became the trading machine, just clicking the button when I was told to. I did fairly well with this, except when I forgot or when I didn't consider all the rules or market context - or somehow completely missed seeing an S/R line.  

After the review, I see substantial progress today. I am almost certain that I will perhaps never get a 100% day, but I would like to see something above 90%. It is difficult to say though - missing one or two signals can make a big difference. Given the difference this can make, there may be more bias in the EOD then I have assumed. Time will tell.

Trade well.

Details:

No blow by blow today, but here are the charts:

Again, I hesitated on a couple of key decisions, meaning a late entry, improper reversal, or no trade at all. This is the only difference between the live and the EOD. Granted EOD was accomplished knowing what the charts were going to look like, but I chose what I did live today because I did not consider market sentiment in the decision. For example, the reversal on the 10:40 (long to short): the candle wicked towards the day's open and I debated the reverse. I recall thinking that the resistance had been broken once already and thought that perhaps this did not make it valid any longer. But - considering that it was the open for the day and the market was 'tentative' in reaching up to the open, it was below it - this should have been a sign of weakness. This happened again on the 11:05: the long signal was debatable, especially with the weak market sentiment.

The 13:30 short was just signal at all, which I realized after the order was filled, and I completely missed the fact that the 13:15 stopped on prior S/R. I was aware of the 15:45, but hesitated and went with the no.

A few hesitations and it changed the course of the day substantially.

I calculated the percent in plan by listing times and actions taken for both the live and EOD applications. If the actions matched over a particular time slot, the time slot received a score of 1. If they did not match, the time slot received no score. I then removed the blanks from the EOD application and divided the score by the number of actions in te EOD column:

Complete data set:

With blanks removed: