Poll for your Profit Factor | Page 2 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

New Expert Advisor PROFIT FACTOR 19.39 47 replies

Profit Factor Phase Space 15 replies

Profit Factor OR Expected Payoff...Which is more important?? 5 replies

Profit Factor Target 27 replies

Profit factor : 11.64 2 replies

Attachments: Poll for your Profit Factor
Exit Attachments

Poll for your Profit Factor

  • #21
  • Apr 11, 2010 5:40am Apr 11, 2010 5:40am
Profit Factor = (¸Win% x Avg Profit) / (Loss% x Avg Loss)

Expectancy = Win% x Avg Profit - Loss% x Avg Loss
  • #22
  • Apr 11, 2010 5:47am Apr 11, 2010 5:47am
Cheers,

If they are the same, positive expectency is not meaningful. Expectency has to be greater than 1 to be profitable.

Quoting hanover
Unless I'm missing something, they are the same thing.

To put the poll results in some kind of perspective, a PF > 1.3 is good; above 1.5 is outstanding. Over a minimum of several hundred trades, however, for it to have any real significance.
  • #23
  • Apr 11, 2010 5:48am Apr 11, 2010 5:48am
In this version, Expectancy = Gross Profit - Gross Loss = Net Profit???

Quoting ForexQuant
Profit Factor = (¸Win% x Avg Profit) / (Loss% x Avg Loss)

Expectancy = Win% x Avg Profit - Loss% x Avg Loss
  • #24
  • Apr 11, 2010 6:14am Apr 11, 2010 6:14am
Quoting kk007
Cheers,

If they are the same, positive expectency is not meaningful. Expectency has to be greater than 1 to be profitable.
In order to make money, Expectancy must be greater than 0 and Profit factor must be greater than 1

Expectancy is how much (In average) you expect to make in every trade.

Net profit = Expectancy x Number of trades
  • #25
  • Apr 11, 2010 6:39am Apr 11, 2010 6:39am
Cheers,

This definition makes sense. Under this definition,

Net Profit = Expectancy x Number of trades
Expectancy = Net Profit / Number of trades
Expectancy = Expected Net Profit per trade (unit is in dollar then)

Profit Factor = Gross Profit / Gross Loss
Profit Factor = Profit amount per dollar lost (no unit then)

Two concepts may not be directly related then.

Quoting ForexQuant
In order to make money, Expectancy must be greater than 0 and Profit factor must be greater than 1

Expectancy is how much (In average) you expect to make in every trade.

Net profit = Expectancy x Number of trades
  • #26
  • Apr 11, 2010 6:42am Apr 11, 2010 6:42am
OK Ok.

I understand now, your Avg is the average per trade.

cheers,

Quoting ForexQuant
Profit Factor = (¸Win% x Avg Profit) / (Loss% x Avg Loss)

Expectancy = Win% x Avg Profit - Loss% x Avg Loss
  • #27
  • Apr 11, 2010 11:35am Apr 11, 2010 11:35am
Quoting ForexQuant
Profit Factor = (Win% x Avg Profit) / (Loss% x Avg Loss)

Expectancy = Win% x Avg Profit - Loss% x Avg Loss
OK, agreed, but in that case I believe that PF is the true measure of the efficacy of a 'raw' trading method, because Expectancy factors in position size.

For example:

-- I size my positions at 1 lot, my wins are $600 and my losses $400 over 10 trades. PF = 1.5; expectancy = $20 per trade

-- Now, by making the same trades, but at 10 lots each, my wins will be $6,000 and my losses $4,000. PF = 1.5; expectancy = $200 per trade

Both trading methods are equally efficient, as reflected by the PF, but the expectancy is 10x greater, i.e. all it does is reflect the higher position size.
  • #28
  • Apr 11, 2010 11:57am Apr 11, 2010 11:57am
Quoting hanover
Both trading methods are equally efficient, as reflected by the PF, but the expectancy is 10x greater, i.e. all it does is reflect the higher position size.
IIRC Tharp suggests normalising the expectancy value by dividing both sides of the equation by the size of the average losing trade, otherwise as you point out, a larger poition size leads to a higher expectancy !
  • #29
  • Apr 11, 2010 6:55pm Apr 11, 2010 6:55pm
Agree. PF is the best measure for evaluating performance, provided that the no. of trades is big enough.

There are a few votes for the option >50 here, but I doubt if their sample sizes are very small and whether this PF is sustainable at all, or even just backtesting result. Since no details are given, I cannot make my accurate evaluation.

Quoting hanover
OK, agreed, but in that case I believe that PF is the true measure of the efficacy of a 'raw' trading method, because Expectancy factors in position size.

For example:

-- I size my positions at 1 lot, my wins are $600 and my losses $400 over 10 trades. PF = 1.5; expectancy = $20 per trade

-- Now, by making the same trades, but at 10 lots each, my wins will be $6,000 and my losses $4,000. PF = 1.5; expectancy = $200 per trade

Both trading methods are equally efficient, as reflected by the PF, but the expectancy is 10x greater, i.e. all it...
  • #30
  • Apr 11, 2010 6:59pm Apr 11, 2010 6:59pm
50+ Screenshot
Attached Image (click to enlarge)
Click to Enlarge Name: Snap19.png Size: 30 KB
  • #31
  • Apr 11, 2010 7:35pm Apr 11, 2010 7:35pm
I have taken the average PF of the profitable traders, did not take the outliners (PF>10) into the calculation, and taken those PF between 1.5 to 10.

The average PF = 2.727 (33 out of 52 Votes)
.
  • #32
  • Apr 12, 2010 3:01am Apr 12, 2010 3:01am
In one of his interview Paul Tudor Jones said he very rarely saw traders with a profit factor better than 2...
  • #33
  • Apr 12, 2010 3:44am Apr 12, 2010 3:44am
Quoting Beman
In one of his interview Paul Tudor Jones said he very rarely saw traders with a profit factor better than 2...
That's in keeping with what I've read, and it makes obvious mathematical sense. Anybody managing to maintain a PF of 2 over several years, and with a basic knowledge of MM, would be astronomically wealthy.

Recently there was a thread telling us that no more than 1% of traders were profitable at all. Now we find a poll stating that the average PF of those polled (supposedly a representative cross-section of FF members) is a whopping 2.7. Clearly one of the two statements must be miles off the mark.
  • #34
  • Apr 12, 2010 5:08am Apr 12, 2010 5:08am
I have done a number of pollings on forexfactory.com. Although from time to time, people ask me not to believe the polling result. Nevertheless, the answers to my different polling questions and those from other pollings give me consistently feeling that most of the respondents do not intentionally fake the response (although there may be exception occasionally). And I believe under the management of the administrator, different responses mostly likely come from different people.
  • #35
  • Apr 12, 2010 6:05am Apr 12, 2010 6:05am
Remove the outliners who has PF higher than 5 then u may get a more realistic results.

Although I never conduct such a study but I believe individual performance statistic is not normal distributed, I think it could be lognormal distributed, exponential distributed or Pareto distributed.
  • #36
  • Apr 12, 2010 8:26am Apr 12, 2010 8:26am
Quoting ForexQuant
Remove the outliners who has PF higher than 5 then u may get a more realistic results.

Although I never conduct such a study but I believe individual performance statistic is not normal distributed, I think it could be lognormal distributed, exponential distributed or Pareto distributed.
You are right that PF are more a lognormal disturbution. I was lazy and just took the simple average of the PFs rather than taking average of the log values. hehe....

About what is the highest realistic PF, I really have no idea. I really want the persons who voted for >50 to post here about how they obtained their values.
  • #37
  • Apr 12, 2010 1:24pm Apr 12, 2010 1:24pm
Quoting kk007
About what is the highest realistic PF, I really have no idea. I really want the persons who voted for >50 to post here about how they obtained their values.
I would want anybody with a PF > 2.5 to post their number of trades, and also how they obtained their values. And any Martingale type systems should be excluded from the poll, because their current PF will not have factored in the effect of the yet-to-occur 'death trade'.

Just for fun, I went to collective2 (http://collective2.com/cgi-perl/system/grid.mpl) and clicked the 'Forex' button to restrict the search to forex systems only. Then I sorted the list by number of trades descending. Only 20 systems had more than 500 trades, and the highest PF among them was 1.4

Not a very scientific approach, I concede (and 500 trades is too low to be of any real statistical value), but the result is in what I'd call a believable ballpark. And at least there's the credibility of an independent audit.

IMHO a poll structured something like:

< 1
1
1.05
1.1
1.15
1.2
1.25
1.3
1.35
1.4
1.45
1.5
1.6
1.7
1.8
2
2.5
3
4
5 or above

would provide a more meaningful result.
  • #38
  • Edited 9:35pm Apr 12, 2010 7:24pm | Edited 9:35pm
Quoting hanover
I would want anybody with a PF > 2.5 to post their number of trades, and also how they obtained their values. And any Martingale type systems should be excluded from the poll, because their current PF will not have factored in the effect of the yet-to-occur 'death trade'.
Profit factor do not reflect the unrealised drawdown. If a system do not close losing trades, its profit factor would remain infinity. An extreme large PF can be produced by non-closing of lossing trades.

Quoting hanover
Just for fun, I went to collective2 (http://collective2.com/cgi-perl/system/grid.mpl) and clicked the 'Forex' button to restrict the search to forex systems only. Then I sorted the list by number of trades descending. Only 20 systems had more than 500 trades, and the highest PF among them was 1.4.
All are mechanical systems there, and I believe that in general discretionary trading should give a much higher PF than mechanical systems.

Quoting hanover
IMHO a poll structured something like:

< 1
1
1.05
1.1
1.15
1.2
1.25
1.3
1.35
1.4
1.45
1.5
1.6
1.7
1.8
2
2.5
3
4
5 or above

would provide a more meaningful result.
I doubt how many of them has a stable PF down to 0.05 precision. If you don't like the options above 5, you can threshold all above-5 options down to 5 in your analysis.
  • #39
  • Edited 9:49pm Apr 12, 2010 9:37pm | Edited 9:49pm
You see extremely high PF on this polling, one probable reason is:

Quoting kk007
Profit factor do not reflect the unrealised drawdown. If a system do not close losing trades, its profit factor would remain infinity. An extreme large PF can be produced by non-closing of lossing trades.
So, some of the >50 voters may be using a no-stoploss system.
.
  • #40
  • Apr 12, 2010 11:13pm Apr 12, 2010 11:13pm
Quoting kk007
All are mechanical systems there, and I believe that in general discretionary trading should give a much higher PF than mechanical systems.
Quite possibly; if all of those systems are in fact mechanical, that's a very fair point.

Quoting kk007
I doubt how many of them has a stable PF down to 0.05 precision. If you don't like the options above 5, you can threshold all above-5 options down to 5 in your analysis.
All that's required is to divide the total profit by the total loss. Anyway, no problem; like you say, anybody can threshold the results however they wish. Personally, I would liked to have been able to see a greater window of detail in an area of the curve in which (IMO) most of the results 'should' lie.

Quoting kk007
Profit factor do not reflect the unrealised drawdown. If a system do not close losing trades, its profit factor would remain infinity. An extreme large PF can be produced by non-closing of lossing trades.
That's true, but assuming the object of the poll is to give a realistic result of what's possible in forex, shouldn't unrealized P/L be included? If somebody doesn't close any of his losing trades, does that make him a better trader than one who does?

I guess it all comes down to what you're aiming to measure with the poll, and of course that's none of my business. All I'm saying is that when I see a result that might imply that 12.28% of the forex traders in the world make more than 50x as much money as they ever lose, it's difficult for me (and I suspect anybody that's been trading for several years) to take it seriously.

Hey, don't get me wrong, I'm not taking a shot at you. I realize that all you've done is innocently posted a poll on a forum, the response to which is completely out of your control.
0 traders viewing now