Pete's Place

Quote from peterfigliozzi:

All of the entries this week were one of the three methods:
Trend Touch, S/R Reversal, or S/R continuation.

I compared the distribution of MFE's for each method to that of 24 random entries. Histograms attached!

Both Trend Touch and S/R Continuation showed no appreciable difference from random entry! Actually, they were a little worse.

I also looked at win probabilities and some other numerical metrics. Same results: not much different than random, maybe a little worse.

In contrast S/R Reversal looks really good compared to random. However, there were only six trades on the week, so i'll need to trade it for a while longer to get enough statistics.
More...

Supporting the idea that it's not the entry that matters but position management and the exit ;)

http://www.wealth-lab.com/cgi-bin/WealthLab.DLL/editsystem?id=18907 Tom Basso's Random Entry - Wealth Lab

http://www.wealth-lab.com/cgi-bin/WealthLab.DLL/editsystem?id=18919 Dr. Koch's Random Entry - Wealth Lab
 
Well, I'm not really sure what to do at this point, so i'll continue with the same three entry methods on the simulator. M, Tues, Weds; then taking a week off. Plenty of time to think about possible changes. In the meantime, more data.

Also, I need to think about characteristics of entries that the statistics and numbers don't show. For example, the speed at which you can recognize the trade isn't working.
 
With all due respect to Igor's links, I've got a couple of problems with them.

First, I believe they are premature to Peter's ongoing inexorable quest towards profitability.

Peter needs to be profitable before getting into heavy-duty money management. I can't speak for anyone else but I find getting a better entry not only allows me to reach my target but reduces the stress level inherent in every trade. I miss a number of trades because I put in limit orders a few ticks above (if looking to short) the close of the bar giving me the signal.

In regards to the two links, while I have tremendous respect for Van Tharp, I don't find his money management methods lend themselves well to day-trading. They're for after we get tired of day-trading and have enough capital to generate income based on systems we've been tucking away for that day when we don't need to spend more than an hour a day reviewing past, current and potential trades.

Best regards,
jd
PS I have to admit my mind shuts off when I see methodologies explained with rampant misspellings (not checking one's output on the internet in trying to promote one's system isn't reassuring, but then I have an exaggerated fondness for the written word) and the word "random" to say nothing of "my system is in the market all the time."
 
I'm with Murphy when he says, "Things we do not understand appear random."

I know that it's possible to be profitable with a random entry, however I believe one can do better than random entry.

I'm using random entry as a yardstick against my own entry methods, nothing more.

So far "S/R Reversal" is better than random. Will it continue to outperform? We'll see.
 
Thanks Peter:
I have been reading your journal entries (always trying to learn something new), and noticed your comments about the colors of your charts ("puke green"). I have been having some eye strain problems and I thought I would try a setup that favors that color scheme. Well I have to tell you, it makes all the difference. Nice calming effect too by the way. Anyway, I know this just little stuff but I feel obliged to say thanks for the idea, and I hope you find success with your approach. Best Regards, Steve46
 
Let me tell you, finding out most of my entries were no better than random entries was a SEVERE blow, however...

... drum roll please ...

I screwed up the random entry calculation, making it seem a lot better than it actually was...

now that I've done it correctly, (a) the numbers the random entries are giving me make sense, i.e., you can't make money setting a limit order to exit at "X" points and having your stop at "X" points; but more importantly (b)

--> Trend Touch is significantly better than random entry <--

I knew it was. The entries almost all gave a chance for a graceful exit... not that I took that exit each time (gotta work on it), but there didn't seem to be anything wrong with the entries.

Also, the numbers show that "S/R Continuation" is indeed a crappy entry method, about equal to random entry.
That's where I would enter after a S/R was broken, and then backed up into, in the original direction, without waiting for a candle shadow. Apparently that's not the brightest thing to do.

Actually today I traded "S/R Continuation" and waited for a shadow... got me long at 1402.50.

Here's the updated spreadsheet where I compared the methods, with the random entries done correctly. To fix it I went both long and short from each random entry and counted it as a separate trade. For truly random entries it shouldn't change the results; it did; so I took the average of the results and it gave numbers that made sense.
 

Attachments

Quote from steve46:

Thanks Peter:
I have been reading your journal entries (always trying to learn something new), and noticed your comments about the colors of your charts ("puke green"). I have been having some eye strain problems and I thought I would try a setup that favors that color scheme. Well I have to tell you, it makes all the difference. Nice calming effect too by the way. Anyway, I know this just little stuff but I feel obliged to say thanks for the idea, and I hope you find success with your approach. Best Regards, Steve46
More...

Hey there Steve, glad you like the "puke green"!
 
Note to self:

I'm finding that the aggressiveness of my exits are based on the P/L of the day, and that leads to problems. If I am down, i'll tend to go for home runs. I'm i'm up, i'll tend to go for singles, or even less.

Seems like the right thing to do is (1) decide on exits before opening the trade (2) based on market action, not P/L of the day.
 
Quote from JDRower:

Pete
Do you make decisions based on perceived trend? If so, what do you use and what standard size/length chart do you use?
JD
More...

Yes I do-- the time frame varies... for "Trend Touch" I look for a clean looking chart. I usually cycle through 100 tick to 2000 ticks. One chart in the window, and I flip through like a slide show. The main thing is to draw a trendline, because there are all kinds of trends and I have to be concrete about which one i'm going with.

I have also been using the "Longer Time Frame Wins" rule, when two time frames disagree.
 
You must log in or register to reply here.
Back
Top