POLL: Your choice of a positive expectancy system

C

candletrader

Let me assume that (irrespective of your trading methodology) you are profitable and are primarily using one of the following types of positive expectancy approaches:

1) A high probability (65% and upwards) methodology with a relatively mediocre avg. win size: avg. loss size ratio (anything from 1 to perhaps 1.5)
2) A low probability (40% and below) methodology with a relatively high avg. win size: avg. loss size ratio (upwards of 3 would be a reasonable assumption)

Ceteris paribus (putting opportunity factors and other considerations to one side), I would like to take a poll of profitable members from within our community...

As always, feel free to not only vote but to add any pertinent commentary as you see fit...
 
i voted for your second option, although my real numbers are lower on the % and higher on the r:r

still, i think a high prob/low r:r system, with the % > 75, and r:r around 1:1, is way superior to low % and high r:r

for example, i would far prefer to trade a system with 80% winners and 1:1 reward to risk, rather than 40% winners and 3:1 r/r.....even though they both have the same expectancy


why?

because i think the position sizing possibilities are far superior with the former.

you're making some money? great! go ahead and increase those positions sizes without worrying about lengthy drawdowns! hell, you could probably even get away with doing the unthinkable -- use a martingale strategy and come out shining!
 
Quote from candletrader:

Let me assume that (irrespective of your trading methodology) you are profitable and are primarily using one of the following types of positive expectancy approaches:

1) A high probability (65% and upwards) methodology with a relatively mediocre avg. win size: avg. loss size ratio (anything from 1 to perhaps 1.5)
2) A low probability (40% and below) methodology with a relatively high avg. win size: avg. loss size ratio (upwards of 3 would be a reasonable assumption)
More...

Can we really separate the expectancy from the methodology? I mean, successful trading looks like managing expectations. Different methods have different expectations.

OPC
 
OPC,

My use of the word "irrespective" is to avoid specific talk of strategies on this thread... I fully agree with your above comment, which doesn't contradict anything I have said... yes, different strategies have different R:R and win rate% characteristics (and may also have different per trade and / or total opportunity-factor weighted expectancies, which may or may not be known of a priori), but this thread is not about strategy discussion but about the more general choice of R:R and win rate% combination, irrespective of what strategy is used (with the one constraint that the strategy is positive expectancy)... but if you want specific examples in the intraday perspective:
- Option 1 (the high probability approach) may include a channel trading strategy using 1 minute bars and (20,2,2) Bollinger Bands (and will also invariably use other criteria to refine the methodology)... the risk would not differ significantly from the reward on a per trade basis, but you would expect to win many more times than you lose...
- Option 2 (the low probability approach) may include a 5 minute candlestick, 20 period moving average trend following strategy with relatively large volatility based stops (and will also invariably use other criteria to refine the methodology)... the reward could be a major multiple of risk when this strategy pays off but you would expect to lose many more times than you win...

Bear in mind that long run total expectancies for option 1 and option 2 strategies can be equivalent, because of the impact of things like R:R in combination with opportunity factors etc... but one will not know, a priori, without specific backtesting of the
various strategies, what the expectancy differentials are...

The point of this thread is not to go into specific strategies, but to discuss things at the more general level, irrespective of the specific strategy employed... it should be pretty apparent to the trader which of the two Options his trading strategies (of which there is an infinite number - hence my desire to avoid strategy discussion) is more closely aligned with... the only requirement, in order to produce a meaningful poll, is that the trader must have a positive expectancy methodology i.e. be profitable on a week by week, or at least on a month by month, basis...

Candle
 
Quote from daniel_m:

i voted for your second option, although my real numbers are lower on the % and higher on the r:r

still, i think a high prob/low r:r system, with the % > 75, and r:r around 1:1, is way superior to low % and high r:r

for example, i would far prefer to trade a system with 80% winners and 1:1 reward to risk, rather than 40% winners and 3:1 r/r.....even though they both have the same expectancy


why?

because i think the position sizing possibilities are far superior with the former.

you're making some money? great! go ahead and increase those positions sizes without worrying about lengthy drawdowns! hell, you could probably even get away with doing the unthinkable -- use a martingale strategy and come out shining!
More...


Theorically, the high % is preferable, because of the much more attractive money management options, but practically, these high % plays are associated with the use of an edge not derivated from the underlying traded. (for example, a discrepancy between and index future contract and an element of the index, or getting price improvement and so on...). Sinces these edges are mostly found on the smaller timeframes, the money management advantage of having a high % strategy is not interesting, because one will reach the maximum size extremely quickly, and then have to go back to a fixed size/fixed share MM type of MM.

So, the trader is basically stuck between high probability strategies with limited size, and high r:r strategies, allowing for much more size but less creative MM.
I see the former type of strategy as paying the bills and the second as long term growth of the business, so, no preference from me :)

OHLC
 
"MU!"

Different terrains demand different approaches. There are periods when it is better to 'hit & run' and others when the 'swing for the fences' approach is by far superior...

I will leave my own preferences out of it and assume the shape of my container...
 
Quote from Fluidity:

"MU!"

Different terrains demand different approaches. There are periods when it is better to 'hit & run' and others when the 'swing for the fences' approach is by far superior...

I will leave my own preferences out of it and assume the shape of my container...
More...

I take your point that there can be evidence indicating the nature of the current environment... but then, of course, you are introducing another factor into your decision process... your approach is totally valid if thats the way you wish to play the game... personally, I would have great difficulty switching between strategies on potentially a half hourly basis (imagine being in scalping mode and to bail for 5% of what you could have made on one trade, and THEN to go into trend trading mode and to have the price pop up and retrace on you)... for me, I prefer a consistency of approach, but your comments do indeed have merit and perhaps the moderator can introduce your comments as a third option in the Poll... I will private message our moderator to introduce a third option:-
(3) "I assume the shape of my container": I dynamically assess the environment to decide on the appropriate R:R and win% characteristics, and align my strategy with my assessment of these characteristics...
 
Quote from daniel_m:

i voted for your second option, although my real numbers are lower on the % and higher on the r:r

still, i think a high prob/low r:r system, with the % > 75, and r:r around 1:1, is way superior to low % and high r:r

for example, i would far prefer to trade a system with 80% winners and 1:1 reward to risk, rather than 40% winners and 3:1 r/r.....even though they both have the same expectancy


why?

because i think the position sizing possibilities are far superior with the former.

you're making some money? great! go ahead and increase those positions sizes without worrying about lengthy drawdowns! hell, you could probably even get away with doing the unthinkable -- use a martingale strategy and come out shining!
More...

OH BOY :eek:

Dannyboy,

Are you so sure that the higher win % will produce the higher DD to Net Ratio? Perhaps it is the overall expectancy or profit-factor that holds the most weight?
 
You must log in or register to reply here.
Back
Top