실: Risk to reward vs Expected Value

Most of us calculate RR = potential profitpotential loss

For example, if our TP gives $200 and SL lose $100 then the RR = 2 right?

However, most of the time we don't take our pseudo edge into account.

Before calculating risk to reward ratio... a trader must know what is the odds of success.

For example lets say your pseudo edge (i'm calling it pseudo for personal belief) is 30% chance.

Then if you TP gives $200 and SL lose $100, the expected value will be

Expected Value = Odds x Payout

= 0.3 x 200 = $60 which is less than potential loss of $100

This trade should be passed on that means take a hike. But if you know the odds of success to be lets say, 70% then

= 0.7 x 200 = 140, which is more than $100 that means its a good bet and your true RR is 140/100 = 1.4 pretty good.

원래에 의해 게시 됨 ;

Most of us calculate RR = potential profitpotential loss......

Yes, you are talking about what has been variously termed 'expectancy', 'edge', or 'profit factor', i.e. the product of win rate and average win/loss size (RR). This ultimately equates to 'dollars won divided by dollars lost', which I believe is the only fair way to compare trader/system performance. RR is dynamic and changes as a trade progresses. For example, if you reach the point where you move the stoploss to breakeven, the risk is now zero. Hence I believe that the best way to measure performance is in hindsight.

Yes it is dynamic depending as the trade progress but the way to calculate RR does not take the system's edge into account. Anyone can simply say I'm risking 10 pips for a 12 pips target... well what is the odd of that happening? The point I'm trying to make is that first find the success rate of your edge and then calculate. Not simply, I'm risking 20 pips for 10 pips target so RR is 0.5 (below 1) and its bad or I'm risking 10 pips for 15 pips so the RR is 1.5 now so its good. First and foremost a trader must know the odds. Backtest/forward test... 200 years of data (whatever)... first get the statistics right :-) Its not my idea, I got it from
http://en.wikipedia.org/wiki/Daniel_Bernoulli

원래에 의해 게시 됨 ;

Yes it is dynamic depending as the trade progress but the way to calculate RR does not take the system's edge into account. Anyone can simply say I'm risking 10 pips for a 12 pips target... well what is the odd of that happening? The point I'm trying to make is that first find the success rate of your edge and then calculate. Not simply, I'm risking 20 pips for 10 pips target so RR is 0.5 (below 1) and its bad or I'm risking 10 pips for 15 pips so the RR is 1.5 now so its good. First and foremost a trader must know the odds. Backtest/forward...

Yes, we are in agreement. Expectancy is the product of RR and win rate.