This post was contributed by a guest author, and does not necessarily reflect the views of Richard or MovetheMarkets.com
November was a good month for me profit-wise. I let my winners run to their targets, and that made all the difference in my profit/loss column. More importantly, I learned some valuable lessons that I’ll apply from here on out. First, I will continue to cut losses, but I won’t hastily exit trades. My “tape reading” was more about fear of loss, and it caused me to exit trades prematurely. I will, for the most part, stick to letting the candle close before I decide whether or not a trade is proving correct. Secondly, I will not “revenge trade” thanks to some new rules taped to my monitor and forehead. That nasty habit cost me -5.4 R over only two trading days. I traded .67% of my account on each trade. Here are my results:
| Total P/L: | 11.85 R Net |
|---|---|
| Trades Taken: | 36 |
| Winners: | 9 (25%) |
| Break Even Trades: | 15 |
| Losses: | 12 |
| Expectancy: | 0.33 R |
| Biggest Winner: | 4.7 R |
| Biggest Loser: | -1.5 R |
This post was contributed by a guest author, and does not necessarily reflect the views of Richard or MovetheMarkets.com
December 1st, 2006 at 8:28 am
Zoomie, what are your criteria to separate a break-even trade from a winner or a loser, and how do those play in to your expectancy calculations? My own “win/loss” rate is pretty poor, but that’s also including some trades that were really break-even “losses” of a $1 or two.
December 1st, 2006 at 8:36 am
ExEng:
My break even trades are -0.2R or greater. My smallest winner was R 1.5. I will have to check on that though. I basically hold till my target, end of day, or when I am stopped near break-even. Exceptions are obvious pivots the stock may create on the way to my target.
December 1st, 2006 at 8:48 am
Edit: Smallest winner was 0.85 R. Most of my “break-even” trades were between a tiny loss and -0.2 R.
December 1st, 2006 at 9:13 am
Ex Eng: You will have to elaborate on “how do those play in to your expectancy calculations?” My expectancy is derived from my net profit divided by my average risk. Then that R return is divided by my TOTAL number of trades. But hey, I am not a Rocket Surgeon. I mean scientist ;).
December 1st, 2006 at 9:15 am
So you categorize winners as any trade that has a profit > zero, and your break-even range is from zero down to -0.2R, with losers falling below the -0.2R line?
I was surprised by how many break-even trades you had this month out of the overall number. I guess this reflects proactively declaring a trade a “loser” and getting out before -1R?
If you’re willing, I’d be interested to see your P/L broken out like so:
Net of All Winners: + X.X R
Average Winner: + X.X R
Net of All Breakeven: +/-X.X R
Average Breakeven: +/-X.X R
Net of All Losers: - X.X R
Average Loser: - X.X R
Thanks, Zoomie!
.exe
December 1st, 2006 at 9:30 am
I calculate expectancy differently:
Expectancy = [ (Average Win R)*(Win Rate) ] - [ (Average Loss R)*(Loss Rate) ]
where
Win Rate = (# Wins)/(# Trades) and Loss Rate = (# Losses)/(# Trades)
I have to complicate things! I’m a ‘Rocket Scientist’, not a ‘Rocket Payload’ like yourself. ;) But in all seriousness, I’d trade places with you in a second. I’ve always wanted to fly jets :( If this trading thing works out, I’ll buy a Javelin: http://www.avtechgroup.com/index.asp
December 1st, 2006 at 10:32 am
ExEng: LOL. Rocket payload, I like that. I will give your computations a try. Gimmee some time though, I am running low on brain-power today. Well, especially today.
December 1st, 2006 at 10:49 am
Zoomie,
11R in a month is amazing, well done! I figure that if you can risk a mere 1% of your stake on each trade and can maintain these figures, you’ll pull a 250% year. And it looks like you’re doing this feat with relatively little risk.
Congrats.
December 1st, 2006 at 11:06 am
@ExEngineer: I think you’ll find that your expectancy calculation is equal to the average R p/l of all your trades. You are just taking the average win and weighting it by the percent of wins, then the average loss weighted by the percent of losses. The simple mean of all your R results should give the same value.
December 1st, 2006 at 11:37 am
Tyro: Thanks. Cool blog you have.
Richard: Are you saying it is not worth it for me to do ExEng’s calculations? Cause it would take me a fair amount of work. But if it is informative, well I can do it. It will also force me to keep better records.
December 1st, 2006 at 12:45 pm
@Zoomie: I used to do it like ex-engineer, but then I proved to myself that it’s mathematically equivalent to the straight average. You will get the exact same number. Here’s an example:
Say I have one 2R winner and one 1R loser, and one break-even 0R trade.
The straight average is: ( 2 - 1 + 0 ) / 3 = 1/3 = 0.33
The ExEngineer calc is: 2*(1/3) - 1*(1/3) = 1/3 = 0.33
The reason to keep good enough records for the ExEngineer calc is that you can do deeper analysis of avgwin avgloss winrate lossrate etc. I would add that the std. deviations on the avgwin avgloss are interesting, too. So, the ExEngineer formula tells you more about why your expectancy is what it is. But, if you just want the expectancy number, add up all your R’s and divide by the number of trades and you should get the same answer.
December 1st, 2006 at 2:43 pm
zoomie: nice job!
December 1st, 2006 at 8:36 pm
Richard,
You are correct about the end result of expectancy being the same, but they only match because you chose 0R; it’s my definition that is incomplete. You do need to change my definition to include the breakeven trades along with the wins and losses. If you replace the 0R in your example with Zoomie’s definition of -0.2R and you get:
The straight average is: ( 2 - 1 - 0.2 ) / 3 = 0.8/3 = 0.267
The ExEngineer calc is still: 2*(1/3) - 1*(1/3) = 1/3 = 0.33
I need to include the weighted average R for breakeven trades to get the same number. Mathematically, then the two methods produce the same result. You are also correct that the only reason to do it “my” way is the greater visibility into the components. If you just average the R’s, you don’t know if your profitability is due to lots of small winners, or a few home runs. Same for losses. But there’s no need for Zoomie to hurt his head today. :)
December 1st, 2006 at 9:42 pm
@Ex: I hope neither you nor zoomie eleminates a whole band of trades when calculating expectancy. I don’t think that would be a very good idea. For expectancy purposes, -0.2R is a loss. It has to be. I have been assuming Zoomie’s count of breakeven trades vs winners vs losers is a seperate thing.
In fact, looking at his numbers above, I see he just divided his net profit 11.85R by total # trades 36. So, that should be the right number, assuming the 11.85 has all the “break-even” trades in it.
December 2nd, 2006 at 7:00 am
Richard:”assuming the 11.85 has all the “break-even” trades in it.”
Yes it does. I factor in all my tiny losses.
December 2nd, 2006 at 10:36 am
@Richard: Zoomie’s calculation is accurate for expectancy. I have historically calculated expectancy with only winners (R>0) and losers (RR>-0.2) are counted as losses. I’ve been wondering what to do about this, and I think a breakeven category would be a smart thing to track, and better reflect my true win/loss rate.
December 2nd, 2006 at 11:12 am
I consider a break even trade a win in my mindset but not in my statistics shown here. A break-even trade for me usually means the trade went my way in an order of magnitude that if it did come back to my entry point the trade was probably not proving correct. Of course I am wrong sometimes, but I feel the way I position my stop-loss to break-even saves me more sanity and money than hoping the trade will come back. However, all trades are different and require different exit criteria.
December 2nd, 2006 at 2:01 pm
@Ex: I’m not saying it’s a bad thing to track a break-even band of trades. I agree that can be very useful information. Leaving any trades out of your expectancy keeps the value from telling you what it’s supposed to tell you, though.
You can easily expand your formula to have your break-even average weighted by the break-even percentage of trades, like I think you already mentioned. Nothing wrong with that!
March 3rd, 2007 at 1:10 pm
[...] go re-read Zoomie’s November review, and I bet it will sound even more impressive than it did the first [...]