Vector,
This is not really my cup of tea, however this is an example of a hedge fund that has developed their own software for trading the FX markets. Sounds similar to your approach. The following are the posts that were posted by the managing partner of the hedge fund.
I think it is important firstly to define expectancy, just to be sure you aren't confusing it with something else.
Expectancy is the average amount of money you'll make in your system over many trades, per dollar risked.
Expectancy is controlled by the exit strategy you employ and it is one reason, perhaps the chief reason why I do not like to see traders use technical analysis techniques to exit from their trades. Using TA to enter a trade is one thing, but using a 'reversal' of a TA signal to exit the trade has a tendency I believe to disregard a key objective in successful trading which is to make money. Instead, it tends to play to the traders desire to be proved 'right' and this can and should be avoided.
Back to 'expectancy'...
You suggested in your post that 'expectancy' could come at the expense of opportunity. That is to say, that if you have more opportunities, your expectancy reduces. I just want to be sure you are not confusing 'expectancy' with 'reliability' where 'reliability' is the percentage of time your trades make money. For example, if you trade 10 times and you make money 6 times, your reliability is 60%.
To follow your statement:
"...if you have more opportunities to trade, then your expectancy could be much lower than if you have less opportunities."
This could be right, but only where the sample selected differs significantly. If you trade three times and get two of three right, you'll have a 66% reliability, but if you trade 100 times you may get a completely different picture far removed from 66%.
In any case, 'expectancy' cannot be dealt in isolation (as many people do). There are six elements (if Van Tharp is to be believed) and these include:
1. reliability - the number of winners divided by the number of unprofitable trades.
2. relative size of profits versus losses
3. the cost of trading
4. the opportunity to trade
5. size of investment capital (stake)
6. position sizing model or algorithm
The first four together deal with 'expectancy' and the last two are integral parts of the 'risk management' issue.
A famous analogy of Van Tharp is the "snow ball wall" where he likens trading to your attendance at a snow ball fight. You turn up to the snow field with your trading capital and this dictates the size of the wall you build to protect you from snowballs as they are tossed at you. The more capital you have (Variable 5), the larger/thicker the wall you can build to protect yourself. Makes sense, right?
He asks the reader to imagine the market as tossing two types of snowball at your wall...one is the garden variety made of snow and the other is a 'hot' variety. In the first instance, the snow hits the wall and sticks, strengthening the wall - these are the profitable trades that accumulate to increase the size of your wall (trading account). The second snowball is 'hot' and melts the snow that makes up the wall. These are the losing trades.
Clearly, you want more sticky snow to hit the wall than hot, melting snow. This is 'reliability' and I trust you can see how (as 'expectancy' goes), it is not enough on its own.
Some snowballs are bigger/smaller than others. If you cop a direct hit from a giant hot snowball, it will melt more than the accumulation a number of sticky snowballs can repair. Variable 2 suggests that if you can't achieve a great deal of reliability (and most traders can't) then the relative size of your profits must be monitored and managed. If you don't, your wall (trading capital) could find itself vastly reduced and the chance of you getting hit (wiped out) will increase. I trust you can see the 'risk management/position sizing' necessity here without me going on.
The issue of cost (Variable 3) concerns itself with the slight impact each snowball has on the integrity and strength of the wall. Even with sticky snowballs hitting the wall...each time they impact, a little of the original strength is sapped from the wall...much like brokerage does to your account.
In short...you need to determine a 'net effect' of the whole of your trading activity and this is the true guide to 'expectancy'.
In my business our 'snow wall' gets peppered with seven golf ball sized 'hot' snowballs for every three basketball sized 'sticky' snowballs that come our way.
Our 'reliability' is not that high and many people confuse this with a method that has a low 'expectancy'. I trust you can see how and why we would disagree because we believe 'expectancy' is not a single variable, but the accumulation of at least four (as listed above).
One other thing I'll say about the wall (staying with the analogy). It moves! The position it sat in the field last month is different to the position it is in today. Profits (sticky snowballs) don't land evenly...they tend to spread out or accumulate in the one spot and we find we shuffle left or right to stay behind our wall. In other words, the market evolves, conditions change and today's successful strategy may need adjustment if it is to stay successful in the future.
Whatever trading method you devise, it is important to ensure that it is the fundamental principles that are making it successful NOT the entry strategy (which will be harder to adjust). If your trading methodology is fundamentally strong (i.e for start, it deals with the principles listed above and does not rely solely on the entry strategy) then you've got a transportable chance of long-term survival as a trader.
I do agree....but the question you ask is another question I get asked all the time. I don't have a statistically hard answer for you and i'm afraid it is something you will have to test and determine yourself. This is mostly because most markets are different. For example, the AUD/USD does not have the same general "range" as the USD/JPY or EUR/USD. A 50-point move in the Aussie occurs far less frequently than a 150-point move in the Dollar-Yen for example. If we are in the money 80-points in the USD/JPY, we might move the stop to break-even (give or take) yet we might move the AUD/USD stop to break even after just 30-points.
We run some "surface maps" from proprietary software that gives us a guide to probabilistic moves in the market and this is used generally as a guide to where exits should be placed. These maps look like synoptic weather charts...and they move across a matrix in much the same way as high/low pressure systems do. They're very interesting. A hard target of 200-points is set with every trade, but after consultation with the surface map a high probability area might lie between 150 and 170 points so the target is adjusted to this space. This might change again the next day to 160 to 190 points etc. We need these because Adam and Matt (who manage our hedge fund/institutional trading) also take half the position off, once half target is reached (or therabouts) and again...after consultation with the surface map. The research that goes into this is considerable and is dependent upon monitoring general movements and volatility in the market. For example, how many times a week might you see a 30-point move (many) and how many times a week might you see a 200-point move (few)? If a market runs 80-points, what is the probability it will extend to 130-points? Our surface map will guide us...if it's a high probability, we hold on past 80. Rough example.
Along the way...our basic approach and 'expectancy' principles don't change. We know the general target, the win/loss ratio, the reliability and we move the stop to break-even and no further. This other stuff...just helps us keep it smooth while seeking above-average performance.
FWIW...our win/scratch/loss record across the currencies we trade is currently:
EUR
Win 35%
Loss 58%
Scr 7%
CHF
Win 43%
Loss 46%
Scr 11%
JPY
Win 25%
Loss 69%
Scr 6%
GBP
Win 31%
Loss 62%
Scr 7%
EuJy
Win 33%
Loss 49%
Scr 18%
Our USD average is approx Win 33%, Lose 57%, Scratch 10%.
I've yet to see any serious long-term trader truly deviate from these numbers. It doesn't surprise me in the least that you've gravitated to the same.
Cheers d998
This is not really my cup of tea, however this is an example of a hedge fund that has developed their own software for trading the FX markets. Sounds similar to your approach. The following are the posts that were posted by the managing partner of the hedge fund.
I think it is important firstly to define expectancy, just to be sure you aren't confusing it with something else.
Expectancy is the average amount of money you'll make in your system over many trades, per dollar risked.
Expectancy is controlled by the exit strategy you employ and it is one reason, perhaps the chief reason why I do not like to see traders use technical analysis techniques to exit from their trades. Using TA to enter a trade is one thing, but using a 'reversal' of a TA signal to exit the trade has a tendency I believe to disregard a key objective in successful trading which is to make money. Instead, it tends to play to the traders desire to be proved 'right' and this can and should be avoided.
Back to 'expectancy'...
You suggested in your post that 'expectancy' could come at the expense of opportunity. That is to say, that if you have more opportunities, your expectancy reduces. I just want to be sure you are not confusing 'expectancy' with 'reliability' where 'reliability' is the percentage of time your trades make money. For example, if you trade 10 times and you make money 6 times, your reliability is 60%.
To follow your statement:
"...if you have more opportunities to trade, then your expectancy could be much lower than if you have less opportunities."
This could be right, but only where the sample selected differs significantly. If you trade three times and get two of three right, you'll have a 66% reliability, but if you trade 100 times you may get a completely different picture far removed from 66%.
In any case, 'expectancy' cannot be dealt in isolation (as many people do). There are six elements (if Van Tharp is to be believed) and these include:
1. reliability - the number of winners divided by the number of unprofitable trades.
2. relative size of profits versus losses
3. the cost of trading
4. the opportunity to trade
5. size of investment capital (stake)
6. position sizing model or algorithm
The first four together deal with 'expectancy' and the last two are integral parts of the 'risk management' issue.
A famous analogy of Van Tharp is the "snow ball wall" where he likens trading to your attendance at a snow ball fight. You turn up to the snow field with your trading capital and this dictates the size of the wall you build to protect you from snowballs as they are tossed at you. The more capital you have (Variable 5), the larger/thicker the wall you can build to protect yourself. Makes sense, right?
He asks the reader to imagine the market as tossing two types of snowball at your wall...one is the garden variety made of snow and the other is a 'hot' variety. In the first instance, the snow hits the wall and sticks, strengthening the wall - these are the profitable trades that accumulate to increase the size of your wall (trading account). The second snowball is 'hot' and melts the snow that makes up the wall. These are the losing trades.
Clearly, you want more sticky snow to hit the wall than hot, melting snow. This is 'reliability' and I trust you can see how (as 'expectancy' goes), it is not enough on its own.
Some snowballs are bigger/smaller than others. If you cop a direct hit from a giant hot snowball, it will melt more than the accumulation a number of sticky snowballs can repair. Variable 2 suggests that if you can't achieve a great deal of reliability (and most traders can't) then the relative size of your profits must be monitored and managed. If you don't, your wall (trading capital) could find itself vastly reduced and the chance of you getting hit (wiped out) will increase. I trust you can see the 'risk management/position sizing' necessity here without me going on.
The issue of cost (Variable 3) concerns itself with the slight impact each snowball has on the integrity and strength of the wall. Even with sticky snowballs hitting the wall...each time they impact, a little of the original strength is sapped from the wall...much like brokerage does to your account.
In short...you need to determine a 'net effect' of the whole of your trading activity and this is the true guide to 'expectancy'.
In my business our 'snow wall' gets peppered with seven golf ball sized 'hot' snowballs for every three basketball sized 'sticky' snowballs that come our way.
Our 'reliability' is not that high and many people confuse this with a method that has a low 'expectancy'. I trust you can see how and why we would disagree because we believe 'expectancy' is not a single variable, but the accumulation of at least four (as listed above).
One other thing I'll say about the wall (staying with the analogy). It moves! The position it sat in the field last month is different to the position it is in today. Profits (sticky snowballs) don't land evenly...they tend to spread out or accumulate in the one spot and we find we shuffle left or right to stay behind our wall. In other words, the market evolves, conditions change and today's successful strategy may need adjustment if it is to stay successful in the future.
Whatever trading method you devise, it is important to ensure that it is the fundamental principles that are making it successful NOT the entry strategy (which will be harder to adjust). If your trading methodology is fundamentally strong (i.e for start, it deals with the principles listed above and does not rely solely on the entry strategy) then you've got a transportable chance of long-term survival as a trader.
I do agree....but the question you ask is another question I get asked all the time. I don't have a statistically hard answer for you and i'm afraid it is something you will have to test and determine yourself. This is mostly because most markets are different. For example, the AUD/USD does not have the same general "range" as the USD/JPY or EUR/USD. A 50-point move in the Aussie occurs far less frequently than a 150-point move in the Dollar-Yen for example. If we are in the money 80-points in the USD/JPY, we might move the stop to break-even (give or take) yet we might move the AUD/USD stop to break even after just 30-points.
We run some "surface maps" from proprietary software that gives us a guide to probabilistic moves in the market and this is used generally as a guide to where exits should be placed. These maps look like synoptic weather charts...and they move across a matrix in much the same way as high/low pressure systems do. They're very interesting. A hard target of 200-points is set with every trade, but after consultation with the surface map a high probability area might lie between 150 and 170 points so the target is adjusted to this space. This might change again the next day to 160 to 190 points etc. We need these because Adam and Matt (who manage our hedge fund/institutional trading) also take half the position off, once half target is reached (or therabouts) and again...after consultation with the surface map. The research that goes into this is considerable and is dependent upon monitoring general movements and volatility in the market. For example, how many times a week might you see a 30-point move (many) and how many times a week might you see a 200-point move (few)? If a market runs 80-points, what is the probability it will extend to 130-points? Our surface map will guide us...if it's a high probability, we hold on past 80. Rough example.
Along the way...our basic approach and 'expectancy' principles don't change. We know the general target, the win/loss ratio, the reliability and we move the stop to break-even and no further. This other stuff...just helps us keep it smooth while seeking above-average performance.
FWIW...our win/scratch/loss record across the currencies we trade is currently:
EUR
Win 35%
Loss 58%
Scr 7%
CHF
Win 43%
Loss 46%
Scr 11%
JPY
Win 25%
Loss 69%
Scr 6%
GBP
Win 31%
Loss 62%
Scr 7%
EuJy
Win 33%
Loss 49%
Scr 18%
Our USD average is approx Win 33%, Lose 57%, Scratch 10%.
I've yet to see any serious long-term trader truly deviate from these numbers. It doesn't surprise me in the least that you've gravitated to the same.
Cheers d998
