Saturday, May 2, 2009

Using Trade Statistics to Refine Your Edge

In my last post, I talked about how I suspected that I don't trade the preleaders/prelosers as well early in the day as I do the day rangers. However, as I mentioned before, sometimes what we perceive and what is true are two different things. Being a hard-evidence-kind-of-guy, I wanted to see if I truly performed better on the day rangers or if that was just my perception.

The great thing about the TradePerformance software is I can break down my performance by ticker or by sets of tickers. Thus, I tediously went through all my trades over the past 3 weeks, identifying which tickers were preleaders/prelosers and which ones were day rangers. I then ran the stats on both sets.

Here's the stats for the day rangers. Pay attention to only the long trades in the middle column.




Trading Statistics - Selected Trades



All TradesLong TradesShort Trades
Net Profit1763.0881430.273332.815
Total Gains3036.1832541.558494.625
Total Losses(1273.095)(1111.285)(161.81)
Ratio of Gains to Losses2.38492.28703.0568
Number of Trades14913514
Number of Winning Trades91838
Number of Losing Trades58526
Percentage of Winning Trades61.074%61.481%57.143%
Percentage of Losing Trades38.926%38.519%42.857%
Largest Winning Trade195.00195.0095.66
Largest Losing Trade(65.70)(65.70)(48.00)
Average Winning Trade33.364630.621261.8281
Average Losing Trade(21.9499)(21.3709)(26.97)
Average Trade11.832810.594623.7725
Win/Loss Ratio1.52001.43282.2926
Average Winning Trade Length0.000.000.00
Average Losing Trade Length0.000.000.00
Maximum Consecutive Winners1094
Maximum Consecutive Losers7104
Expectancy11.832810.594623.7725
Expectancy Percent53.9082%49.5750%88.1497%
Mathematical Expectation0.53910.49580.8815
Kelly Maximum Bet Size0.44750.44640.4312
Kelly Maximum Reduced 50%0.22380.22320.2156

Report Generated by TradePerformance™ Software - http://www.TradePerformance.com
Copyright (c) 2000-2008 SCG Associates, LLC. All Rights Reserved
TradePerformance is a Trademark or Registered Trademark of SCG Associates, LLC in the United States and other countries



And here's my stats for going long on the preleaders/prelosers. Again, only look at the stats for the long trades column.




Trading Statistics - Selected Trades



All TradesLong TradesShort Trades
Net Profit536.5394703.1994(166.66)
Total Gains1542.59941542.59940.00
Total Losses(1006.06)(839.40)(166.66)
Ratio of Gains to Losses1.53331.83770.00
Number of Trades93858
Number of Winning Trades51510
Number of Losing Trades42348
Percentage of Winning Trades54.839%60.00%0%
Percentage of Losing Trades45.161%40.00%100%
Largest Winning Trade98.32598.3250.00
Largest Losing Trade(62.53)(62.53)(56.145)
Average Winning Trade30.247030.24700
Average Losing Trade(23.954)(24.688)(20.832)
Average Trade5.76928.2729(20.832)
Win/Loss Ratio1.26271.22520.00
Average Winning Trade Length0.000.000
Average Losing Trade Length0.000.000.00
Maximum Consecutive Winners960
Maximum Consecutive Losers638
Expectancy5.76928.2729(20.8325)
Expectancy Percent24.0849%33.5096%(100.00)%
Mathematical Expectation0.24080.33510
Kelly Maximum Bet Size0.25390.38230
Kelly Maximum Reduced 50%0.12690.19120.00

Report Generated by TradePerformance™ Software - http://www.TradePerformance.com
Copyright (c) 2000-2008 SCG Associates, LLC. All Rights Reserved
TradePerformance is a Trademark or Registered Trademark of SCG Associates, LLC in the United States and other countries



Thus, it is evident that I perform better longing the day rangers versus the preleaders. While my winning % was only 1.5% higher, my ratio of gains to losses was 24% higher, and my winning trade/losing trade ratio was 16% higher. I also averaged over $2 more per trade.

Now, I obviously have an edge in both cases. It's just slightly better for the day rangers. Thus, I'll put more focus on those in the first 15 minutes of trading. I'll still continue to go long on preleaders, but maybe be a bit more selective.

11 comments:

Blz said...

Hi Yngvai,

Great job on the consistency. Good to see your hard work is paying off. I just had a thought on the long trades being more profitable as of late that I would throw out there. Perhaps the long winning percentage is a result of the bulls seemingly being in control as of late. Plus as of late buyers seem to be willing to come in on a dip whereas 4 months ago any pullback would bring on additional selling. Perhaps just an assessment of the intermediate trend would be helpful in determining which types of plays of the pre-leaders/losers will have greater odds of being profitable.

Don't know if that makes any sense in light of your trading rules but after reading your post I thought it was something I'd throw out there. Again great work!

Yngvai said...

Blz,

Thanks for the comment.

Keep in mind these stats are only comparing long preleaders to long day rangers...the purpose isn't to compare long to short performance. Still, you are right that these stats are also potentially a function of market conditions. Things probably will change as time goes on. The nice thing is though that there are many ways to play these things so if, say if longing becomes less profitable, I can start shorting the preleaders/prelosers more. Of course my trading rules automatically take care of that in a sense anyway...since I only tend to buy intraday breakouts near HODS or short intraday breakdowns near LODS. Thus, a shift in market sentiment would automatically cause a shift in the plays I do.

Anonymous said...

does some of that have to do with the fact that you see some of the alerts in the chat. I mean if you went with out chat for a few months how would that affect your performance?

Yngvai said...

I almost never trade the chat alerts anymore. I've been using StrategyDesk for my alerts for a while now. I've probably only traded 2-3 chat alerts in the past 3 weeks.

Anonymous said...

I'm a member of IU and about to join CyGroup...what are the commission fees? Are there any monthly fees?

Yngvai said...

Anon,

IU members pay .007 per share. There is also a monthly data fee for NYSE, NASDAQ, and AMEX data feeds which runs about $100/month for all 3. But you can pick which data feeds you want; if you don't trade stocks on AMEX you don't need the data feed for it.

spaniel said...

Hi yngvai, did you find the investorunderground chat room and web site useful?

What are your experience with it? I am new to your site, so please point me to your entries if you have written about them previously.

I am working full time, and would like to dabble part time into trading. Any recommendations or suggestions? I am not sure if I can afford the time during the all the market hours to trade stocks.

Anonymous said...

Great site you have here, and it seems you have come a long way and have put in lots of effort.

Good luck and all the best!

Anonymous said...

Congrats!

Yngvai said...

Thanks, guys!

Yngvai said...

Spaniel,

I've been a member of the IU chat for a long time...in fact I was a member when they were called the GreenontheScreen chat last year. Much of what I learned about trading has come from Muddy's methods. While I don't trade off the chat alerts much anymore (I use StrategyDesk for my alerts now), my experience there has been very positive.

Muddy's method does take a lot of time to learn, however. It's taken me a year to finally get to where I'm at. You don't need to be able trade full time to take advantage of it, though. Johnny Vento has been a member of the chat for along time too, and he just trades part time using Muddy's methods. But we've both adapted his methods to our own style of training. It takes time to find what works for you, and it does take a lot of work and practice.

They have some very useful seminars in there. You should try the free trial just to get a feel for what goes on in the chat room. Also, take a look at some of Tim Sykes stuff. While I find Tim's marketing and arrogance annoying, he is the guy that I first learned how to trade from, and he's got good stuff when it comes to short selling ridiculous run-ups (like AGEN and HEB last week).

If you are looking to trade, I would start with either paper trading or very small amounts of real money. That way you don't blow a bunch as you try to learn.