really fascinating.I have done the same work before but not as delicate and thoughtful as you did. I did it to evaluate tech factors. By the way I Have a blog in China and have many many Quant subscribers, I’m going to translate this article to Chinese and post it on my blog, of course I’ll address the origin. Thank your for sharing it.
I have not. Just searched it up and it looks somewhat similar to what I'm solving with the Surefire Ratio.
But, I use an integral to sum up all of the drawdown area and upside area which captures more information in terms of how long you are in a drawdown than using simple statistics like the Serenity Ratio and Ulcer Index do.
Here's a not-so-insignificant issue I see with this calculation: the numbers absolutely blow up. I.E. what does a surefire ratio of 114.11 or 180.48 mean?
I.E. a Sharpe ratio can be thought of as a financialized T-test. A Sharpe Ratio above 1 means you're making money 67% of the time, assuming the future is like the past, give or take. A Sharpe above 2 means 95% of the time, etc. Also, the smaller numbers have a general sense of understanding. Same idea with a Calmar ratio. 0.5, 1, 2, etc.
But a surefire ratio of 120? What does that even mean? What's supposed to be a good value for this ratio, and how should one intuitively think about it in a vacuum, aside from "equity curve A has a higher surefire ratio than equity curve B"?
It’s a comparative study between two equity curves of the same length. I go through it pretty rigorously in the post so it’s better if you just read the math behind it to see what it’s capturing. Any financial ratio has upsides and downsides.
I don’t look at the market in terms of probability like you mention with Sharpe. 67% of the time does not take into consideration long drawdowns where you are underwater for years and then every other time period where you’re not. Again, I describe this in the post so give it another read and see if it makes sense.
I did read the post. I understand it's a comparative statistic--just that what I'm essentially saying is that unlike some, such as the Sharpe ratio, it doesn't have what one might call an "in a vacuum" anchoring point. And I agree with you on the pitfalls of the Sharpe ratio.
really fascinating.I have done the same work before but not as delicate and thoughtful as you did. I did it to evaluate tech factors. By the way I Have a blog in China and have many many Quant subscribers, I’m going to translate this article to Chinese and post it on my blog, of course I’ll address the origin. Thank your for sharing it.
Thanks for the great words! I would love to see the translation when you post it, so send me your blog!
Have you looked at the serenity ratio?
I have not. Just searched it up and it looks somewhat similar to what I'm solving with the Surefire Ratio.
But, I use an integral to sum up all of the drawdown area and upside area which captures more information in terms of how long you are in a drawdown than using simple statistics like the Serenity Ratio and Ulcer Index do.
How high would the ratio be to find that it is a good/profitable system? Sharpe ratio e.g. > 2 is very good for example
Have you backtested whether your ratio predicts future returns?
Ratios are not used to predict returns.
Here's a not-so-insignificant issue I see with this calculation: the numbers absolutely blow up. I.E. what does a surefire ratio of 114.11 or 180.48 mean?
I.E. a Sharpe ratio can be thought of as a financialized T-test. A Sharpe Ratio above 1 means you're making money 67% of the time, assuming the future is like the past, give or take. A Sharpe above 2 means 95% of the time, etc. Also, the smaller numbers have a general sense of understanding. Same idea with a Calmar ratio. 0.5, 1, 2, etc.
But a surefire ratio of 120? What does that even mean? What's supposed to be a good value for this ratio, and how should one intuitively think about it in a vacuum, aside from "equity curve A has a higher surefire ratio than equity curve B"?
It’s a comparative study between two equity curves of the same length. I go through it pretty rigorously in the post so it’s better if you just read the math behind it to see what it’s capturing. Any financial ratio has upsides and downsides.
I don’t look at the market in terms of probability like you mention with Sharpe. 67% of the time does not take into consideration long drawdowns where you are underwater for years and then every other time period where you’re not. Again, I describe this in the post so give it another read and see if it makes sense.
I did read the post. I understand it's a comparative statistic--just that what I'm essentially saying is that unlike some, such as the Sharpe ratio, it doesn't have what one might call an "in a vacuum" anchoring point. And I agree with you on the pitfalls of the Sharpe ratio.
Are you actively using this ratio instead of e.g. the Sharpe ratio when comparing different strategies?
Yes, especially as a loss / reward function. I also use the integrated drawdown metric a lot to compare two assets and which is riskier.