Frost,
It's a puzzling question, but personally (in my discretionary trading) I've found it almost impossible to achieve the kind of accuracy (>50%) that would be necessary for R:R ratios of 1 or less than 1, whereas it's very possible to take 10 losing trades and then see one huge winner that puts you up for the day.
Possible explanations I can think of:
- Psychology. One's natural tendency is to let losers run forever while cutting winners early to take profits, and this results in losses overall (if it didn't, then learning to trade would be MUCH easier). Enforcing a high R:R requires you to essentially do the reverse. The market is designed to take profits from the first group, so it should also be designed to give them to the second group.
- High R:Rs are "easier." Maybe finding and taking dead-on accurate entries demands a much higher level of skill and experience than riding that occasional huge winner. Therefore, it wouldn't make sense to suggest that newbies go for accuracy right off the bat. I've found this to be the case in my own experience.
- Nature of price movement. Perhaps high R:R setups are strongly associated with non-random periods of price movement, making a high R:R/low accuracy system more stable over time, with a smoother equity curve, than one demanding high accuracy.
I've observed exactly the same thing you have in my backtesting of automated systems: there's a pretty linear trade-off between accuracy and R:R ratios. However, all the mechanical systems I've ever tested have come back with results that were essentially random, once curve-fitting was eliminated. The trade-off could be a feature of edgeless (random) systems, while profitable systems with a true edge tend to cluster in the region of high R:R. I don't know enough to say, but if you'd like to send me a bunch of profitable systems to analyze, be my guest
Obviously, psychology and "easyness" don't really apply to purely mechanical systems.
It's a puzzling question, but personally (in my discretionary trading) I've found it almost impossible to achieve the kind of accuracy (>50%) that would be necessary for R:R ratios of 1 or less than 1, whereas it's very possible to take 10 losing trades and then see one huge winner that puts you up for the day.
Possible explanations I can think of:
- Psychology. One's natural tendency is to let losers run forever while cutting winners early to take profits, and this results in losses overall (if it didn't, then learning to trade would be MUCH easier). Enforcing a high R:R requires you to essentially do the reverse. The market is designed to take profits from the first group, so it should also be designed to give them to the second group.
- High R:Rs are "easier." Maybe finding and taking dead-on accurate entries demands a much higher level of skill and experience than riding that occasional huge winner. Therefore, it wouldn't make sense to suggest that newbies go for accuracy right off the bat. I've found this to be the case in my own experience.
- Nature of price movement. Perhaps high R:R setups are strongly associated with non-random periods of price movement, making a high R:R/low accuracy system more stable over time, with a smoother equity curve, than one demanding high accuracy.
I've observed exactly the same thing you have in my backtesting of automated systems: there's a pretty linear trade-off between accuracy and R:R ratios. However, all the mechanical systems I've ever tested have come back with results that were essentially random, once curve-fitting was eliminated. The trade-off could be a feature of edgeless (random) systems, while profitable systems with a true edge tend to cluster in the region of high R:R. I don't know enough to say, but if you'd like to send me a bunch of profitable systems to analyze, be my guest
Obviously, psychology and "easyness" don't really apply to purely mechanical systems.
