Thursday, September 14, 2006
More on R
The portfolio was down 0.3 percent today. My post on the merits of R has spawned a lively debate in the comments section. First, let me say that I continue to have utmost respect for such traders like Trader Mike and Richard at Move the Markets who attest to the usefulness of R. These guys are talented, take their jobs seriously, and from what I can tell, they are both great traders.
However, at the end of the day, R still fails the “sniff test” for me. Let me put it this way: if a hedge fund manager told you he made 300R last year, would you be more likely or less likely to give him your money compared to a hedge fund manager who told you his fund returned 50 percent last year?
Fine, you say, R is not a great measure of performance, but at least it’s a good measure of risk. Again, I disagree. Suppose you go long a $10 stock. If you devote 50 percent of your capital to this trade and set a stop at $9, then your R is 5% (since at most you will lose 5 percent of your capital on the trade). Note, however, if you devote 10 percent of your capital but set a super wide stop at $5, your R is still 5% (because again, at most you will lose 5 percent of your capital on the trade).
Now I would argue that both trades are qualitatively and quantitatively different in terms of risk management. In particular, most traders who base their decisions on technical analysis would agree that anyone who allows a position to fall 50 percent before getting out has screwed up. Yet, both trades get the same R. For my money, a simple statement like “I risked x percent of my capital on GOOG, and made y percent on the trade” is a lot more informative.
However, at the end of the day, R still fails the “sniff test” for me. Let me put it this way: if a hedge fund manager told you he made 300R last year, would you be more likely or less likely to give him your money compared to a hedge fund manager who told you his fund returned 50 percent last year?
Fine, you say, R is not a great measure of performance, but at least it’s a good measure of risk. Again, I disagree. Suppose you go long a $10 stock. If you devote 50 percent of your capital to this trade and set a stop at $9, then your R is 5% (since at most you will lose 5 percent of your capital on the trade). Note, however, if you devote 10 percent of your capital but set a super wide stop at $5, your R is still 5% (because again, at most you will lose 5 percent of your capital on the trade).
Now I would argue that both trades are qualitatively and quantitatively different in terms of risk management. In particular, most traders who base their decisions on technical analysis would agree that anyone who allows a position to fall 50 percent before getting out has screwed up. Yet, both trades get the same R. For my money, a simple statement like “I risked x percent of my capital on GOOG, and made y percent on the trade” is a lot more informative.
# posted by Stockcoach @ 4:37 PM 
Comments:
<< Home
You said "For my money, a simple statement like “I risked x percent of my capital on GOOG, and made y percent on the trade” is a lot more informative."
Hmm, so Y would be a multiple of X. So you like Ys and not Rs. :-)
But seriously, you're saying exactly the same things that I've been doing. In my "100 R" post I gave my actual percentage gain (contrary to your implication that I was misleading people) and readers of my blog know that for me, 1R = 0.75% of my portfolio. So whenever I say I made so many R the actual percentage is easily calculated. And I often explicitly state it as well.
Bottom line though, if you dont like R-Multiples, fine, don't use them. But there are many of us who find that they're very useful.
Hmm, so Y would be a multiple of X. So you like Ys and not Rs. :-)
But seriously, you're saying exactly the same things that I've been doing. In my "100 R" post I gave my actual percentage gain (contrary to your implication that I was misleading people) and readers of my blog know that for me, 1R = 0.75% of my portfolio. So whenever I say I made so many R the actual percentage is easily calculated. And I often explicitly state it as well.
Bottom line though, if you dont like R-Multiples, fine, don't use them. But there are many of us who find that they're very useful.
Thanks for the kind words. I agree with you on both your points, today. R doesn't measure either of those things very well. People who expect R to measure those things will always consider it a crock!
R-trackers want the same thing you do, but with a twist. Your example is missing the result of the trade. Say both traders got out at $15. Trader 1 made 5R, while Trader 2 made only 1R. That's the rest of the sentence "I risked x and made y" that you'd want to hear. Only, instead of stating it in terms of percent equity, it's in risk-multiple terms.
Both flavors of the sentence have advantages and disadvantages, depending on what you want the sentence to tell you. It so happens that, in my opinion, the most important things to track are what the risk multiples can elucidate. First and formost, is my trading good (regardless of whether my account grew)? If my trades were good, then I can adjust my risk management, or trade frequency, or whatever, and make money. Kinda like your example trader yesterday. If not, then I need to make more fundamental changes in order to be profitable.
R-trackers want the same thing you do, but with a twist. Your example is missing the result of the trade. Say both traders got out at $15. Trader 1 made 5R, while Trader 2 made only 1R. That's the rest of the sentence "I risked x and made y" that you'd want to hear. Only, instead of stating it in terms of percent equity, it's in risk-multiple terms.
Both flavors of the sentence have advantages and disadvantages, depending on what you want the sentence to tell you. It so happens that, in my opinion, the most important things to track are what the risk multiples can elucidate. First and formost, is my trading good (regardless of whether my account grew)? If my trades were good, then I can adjust my risk management, or trade frequency, or whatever, and make money. Kinda like your example trader yesterday. If not, then I need to make more fundamental changes in order to be profitable.
# posted by 
: 7:49 PM, September 14, 2006
: 7:49 PM, September 14, 2006
Changing the subject, when I click on your 'Long' and 'Short' labels on the left, are these your current positions? Thanks.
# posted by : 7:52 PM, September 14, 2006
: 7:52 PM, September 14, 2006
Changing the subject, when I click on your 'Long' and 'Short' labels on the left, are these your current positions? Thanks.
# posted by : 7:53 PM, September 14, 2006
: 7:53 PM, September 14, 2006
From what I see, you are a longer term trader than we daytraders. Controling risk uniformly is utmost important for a daytrader. That is why using R multiples is easy, self explanitory, and effective as long as all definitions are set forth. Trader Mike was very clear on his methods he used to calculate his reward:risk. I plan on increasing my risk size, and I think it will be an easy transition since I trade in terms of R multiples.
# posted by : 8:23 PM, September 14, 2006
: 8:23 PM, September 14, 2006
I think the R concept is useful for day traders who manage each individual trade as a separate "event". But my main criticism of R is that it is limited- e.g. only one dimensional.
It is not that useful in managing a diversified portfolio where you don't necessarily care about the R of single position. You may be holding a long position and hedging it with options, short positions, or other negatively correlated long positions. So you may not use stop losses at all.
It is not that useful in managing a diversified portfolio where you don't necessarily care about the R of single position. You may be holding a long position and hedging it with options, short positions, or other negatively correlated long positions. So you may not use stop losses at all.
# posted by : 9:17 PM, September 14, 2006
: 9:17 PM, September 14, 2006
It's simply another measurement tool.
I was introduced to R multiples from Tharp's book 'TYWTFF'
The way I use R is independant of % of account equity risked on the trade. So if you always risk 2% per trade then your stop will still only be 1R, and if you risk 5% of equity then your stop will still be 1R and so on.
I find it useful enough. Perhaps all measurements can be misleading if used in isolation. Take what works and run with it.
Regards,
Andrew
www.humblemoney.com
Post a Comment
I was introduced to R multiples from Tharp's book 'TYWTFF'
The way I use R is independant of % of account equity risked on the trade. So if you always risk 2% per trade then your stop will still only be 1R, and if you risk 5% of equity then your stop will still be 1R and so on.
I find it useful enough. Perhaps all measurements can be misleading if used in isolation. Take what works and run with it.
Regards,
Andrew
www.humblemoney.com
# posted by : 3:48 AM, September 15, 2006
: 3:48 AM, September 15, 2006 << Home
