Realized expectancy of your discretionary system

[Zyker]

What kind of an average realized per-trade expectancy does your discretionary system yield in the long run?

Of course, this depends on your time frame, and is far from being stationary in time, but it would seem to be interesting to gauge the extent of "as-good-as-it-gets" in this context.

For reference, based on a backtesting campaign of over 10000 trades, my system (ideally) provides an average long-term realized per-trade expectancy of about 0.2 on the daily time frame. Yes, these are simulated trades, but this result arises from several markets, with varying market conditions, and many, many years of data. Further, I did my absolute best to get rid of the notorious hindsight bias, only taking those trades that I'd taken in live conditions - and treating missed opportunities as such.

On the live trading side, this year I am likely to fall between 0 and 0.1 net of all expenses. Even at 265 trades executed year-to-date, this deviation from the mean is normal for the system; there is significant serial correlation in the trade outcomes.

EDIT: For the sake of clarity, let me point out that expectancy here is defined as relative to the amount risked per trade: E.g. risk X, expect to gain 0.2*X

 

[toolazy]

0.22

average loss 0.1, average win 1.5, 4 out of 5 are losers.

1 is unit of risk where disaster stop is.

this year around 0 as of now.

 

[Zyker]

Quote from toolazy:

0.22

average loss 0.1, average win 1.5, 4 out of 5 are losers.

1 is unit of risk where disaster stop is.

this year around 0 as of now.

Mine's 1.3 for an average win and -0.6 for an average loss. Obviously I too have some more losses than wins.

Very good average loss size, btw. It seems that you do not hesitate to move the disaster stop quickly to breakeven.

 

[toolazy]

Quote from Zyker:

Mine's 1.3 for an average win and -0.6 for an average loss. Obviously I too have some more losses than wins.

Very good average loss size, btw. It seems that you do not hesitate to move the disaster stop quickly to breakeven.

your frequency is much higher.
time stops keep me out of trouble most of time. exit even a small win if no passion behind the move. My max loss was 0.45 risk so far. this is based on 50 trades, swing only.

But I am right now running series of 13 trades with no win. Current balance of last 13 is around 0 as current trade working so far. Not too much fun but I am comfortable waiting as my time will come.

 

[Sergio77]

Quote from toolazy:

0.22

average loss 0.1, average win 1.5, 4 out of 5 are losers.

1 is unit of risk where disaster stop is.

this year around 0 as of now.

0.22 is the expected return per trade, not the expectancy which is 2.2. Actually good, 2.2 units for every unit risked.

My real trading expectancy is about 1.2. I was never able to get any higher than that.

 

[Zyker]

Quote from Sergio77:

0.22 is the expected return per trade, not the expectancy which is 2.2. Actually good, 2.2 units for every unit risked.

My real trading expectancy is about 1.2. I was never able to get any higher than that.

You're right about the terminology; I was being careless in the initial post.

What is your typical trade duration?

 

[kut2k2]

Quote from Zyker:

What kind of an average realized per-trade expectancy does your discretionary system yield in the long run?

Of course, this depends on your time frame, and is far from being stationary in time, but it would seem to be interesting to gauge the extent of "as-good-as-it-gets" in this context.

For reference, based on a backtesting campaign of over 10000 trades, my system (ideally) provides an average long-term realized per-trade expectancy of about 0.2 on the daily time frame. Yes, these are simulated trades, but this result arises from several markets, with varying market conditions, and many, many years of data. Further, I did my absolute best to get rid of the notorious hindsight bias, only taking those trades that I'd taken in live conditions - and treating missed opportunities as such.

On the live trading side, this year I am likely to fall between 0 and 0.1 net of all expenses. Even at 265 trades executed year-to-date, this deviation from the mean is normal for the system; there is significant serial correlation in the trade outcomes.

EDIT: For the sake of clarity, let me point out that expectancy here is defined as relative to the amount risked per trade: E.g. risk X, expect to gain 0.2*X

Are you saying you're a discretionary trader who manually backtested over 10000 trades? :eek:

 

[jack hershey]

Quote from kut2k2:

Are you saying you're a discretionary trader who manually backtested over 10000 trades? :eek:

His post says that.

This thread topic seems to be: how much per trade does a trade make?

Trades take various lengths of time and trades make varying amounts of money.

This profit taking has to be tempered with lost time and lost money caused when trades are not profitable.

I can't tab up what he wants to find out, since market conditions affect the opportunities.

I have a team learning to trade right now.

As they learn, they find out what is important that contributes to profits. I find that they automatically tab the market's offer as the day goes by. They dislike not taking the full offer.

I always give them rope.

as they go through learning facility, they then focus on carving.

I see I have scheduled "anticipation" to further refine their work.

As I read the OP and his posts, it seems that he is not doing too well. So what I read is a person skirting issues and not making much money.

I believe he wants to know what making money looks like in real trading.

Our team is very aware of BP and that doubling BP takes 21 points net profit per contract to double capital. (ES mini context)

We know there are four types of trends and that trading is done by staying on the correct side of the market at all times.

I did an edit; I confused this guy with another person temporarily.

 

[Zyker]

Quote from kut2k2:

Are you saying you're a discretionary trader who manually backtested over 10000 trades? :eek:

Yes. The process started in early 2012, as I wanted to find out the roots of my performance (or lack thereof).

While this work has been tedious, it has also been quite revealing in terms of finding out the "normal" temporal fluctuation in the trade outcomes.

 

[Zyker]

Quote from jack hershey:

His post says that.

This thread topic seems to be: how much per trade does a trade make?

Trades take various lengths of time and trades make varying amounts of money.

This profit taking has to be tempered with lost time and lost money caused when trades are not profitable.

I can't tab up what he wants to find out, since market conditions affect the opportunities.

I have a team learning to trade right now.

As they learn, they find out what is important that contributes to profits. I find that they automatically tab the market's offer as the day goes by. They dislike not taking the full offer.

I always give them rope.

as they go through learning facility, they then focus on carving.

I see I have scheduled "anticipation" to further refine their work.

As I read the OP and his posts, it seems that he is not doing too well. So what I read is a person skirting issues and not making much money.

I believe he wants to know what making money looks like in real trading.

Our team is very aware of BP and that doubling BP takes 21 points net profit per contract to double capital. (ES mini context)

We know there are four types of trends and that trading is done by staying on the correct side of the market at all times.

I did an edit; I confused this guy with another person temporarily.

So should I ignore this post? I have nothing to prove. And I am not interested in the specifics of your - or anyone else's - strategy/strategies.

What I am interested in is evaluating how much there is to improve in my own methodology, given my time frame. This should be interesting reference information to anyone for that matter.

Care to share your expected per-trade outcome, as per the original topic?