I see why sharpe of 3 is so desirable

[rallymode]

That and beating the costs. There are so many things out there where you can make a few hundred K a year with a Sharpe of 3, but scaling it to something meaningful like 5-10 bucks never seem to work.

Sharpe is pretty misleading, all things considered. You can have all sorts of artifacts due to the fact that it penalizes any discontinuity, so big positive days can easily lower it.

OPM aside, sharpe is just a dick measuring tool for traders with small accounts. No one i know who trades for a living even tracks it.

Sharpe of 3 you say? That's barely profitable. Here is something i dug up from the archives. This is what selling vol in 2009 looked like.

 

[traider]

How many variables do you need to optimize for sharpe ratio of 4? Are you then set for life or do these system have a short shelf life?

Also, what are the instruments in your trading universe? Do you run the same strategy across a huge diversified set of instruments? How complicated is your trading strategy?

 

[Lee-]

I could never get the need of 1M account with the 250K at risk.Why the extra 750K?

I'm a bit short of $1M, but I target 20-30% margin utilization (equity options) and would do this even if I had $10M. While I generally only have about 1-3% of total account at risk with respect to each underlying, I keep positions in 10-20 underlyings open at a time. My reasoning is that if things go horribly wrong, I'll be able to ride it out, or at worst if I actually manage to hit max loss on every position, I'd still retain enough of my account to be able to rebuild. If I lost 50%+ of my account value due to an overnight gap, that would be substantially more difficult to come back from.

In some ways I may be a bit too conservative in some ways while in others a bit risky, but everyone has their own methods and reasons. Granted if I were trading stocks, I'd likely be targeting > 30% capital utilization.

 

[Palindrome]

OPM aside, sharpe is just a dick measuring tool for traders with small accounts. No one i know who trades for a living even tracks it.

Sharpe of 3 you say? That's barely profitable. Here is something i dug up from the archives. This is what selling vol in 2009 looked like.

Rallymode, those are incredible results. So you were selling vol on what? What was the strategy? My results are a joke compared to yours. My system is trading futures directionally.

 

[SabreMan]

Rallymode, those are incredible results. So you were selling vol on what? What was the strategy? My results are a joke compared to yours. My system is trading futures directionally.

If you had been around in 2009, you would have killed it as well.
If you can do a Sharpe of almost 4 in 2017/2018 by directionally trading futures, your ratio would have been over 10 in 2009.

Markets were very directional in the short term in 2008 and most of 2009, so would of suited your strategy quite well i think.

Anyway i am very impressed with you guys for doing these insanely high Sharpes.

My trading is 100% automated, so i am happy to be able to hit a Sharpe of 2.0

 

[cruisecontrol]

How many variables do you need to optimize for sharpe ratio of 4? Are you then set for life or do these system have a short shelf life?

There is no linear relationship between number of variables and sharpe but I use a couple dozen inputs to my return predictors. The stronger edges contribute much more to sharpe than weaker ones. Shelf life is really unknowable in advance. I'm still trading some of the same edges that I found 7 years ago when I started researching the market, others have come and gone in a much shorter time (maybe those ones weren't really "edges" to begin with but just some kind temporary trends, problem is you can't always tell the difference)

Also, what are the instruments in your trading universe? Do you run the same strategy across a huge diversified set of instruments? How complicated is your trading strategy?

Instruments are all stocks traded on US exchanges (just common stocks, no ETFs, preferred or other weird stuff). Execution is super simple (daily frequency, auctions), while hedging and prediction are fairly complex. About 150k lines of code for the whole system. I'm looking to get into intraday stuff in the future in order to transcend the limitations of daily trading.. being able to use stop loss, supplying liquidity, etc.

 

[rallymode]

Rallymode, those are incredible results. So you were selling vol on what? What was the strategy? My results are a joke compared to yours. My system is trading futures directionally.

I am into relative risk premia, vol replication type of stuff. In '09 there were a lot of plays on microstructure and straight up arbs if you can believe it. Vol futures were very inefficient, the machines had not stepped in yet. I couldn't trade direction to save my life. My hat's off to guys who can.

 

[Palindrome]

I am into relative risk premia, vol replication type of stuff. In '09 there were a lot of plays on microstructure and straight up arbs if you can believe it. Vol futures were very inefficient, the machines had not stepped in yet. I couldn't trade direction to save my life. My hat's off to guys who can.

Cool, I see. I've played the selling premium game in a very basic form. I could never stomach the risk reward. And then when you hedge, the returns are so small. At least my amateur way of looking at option premium.

What you were doing was much more complex, above my head.

 

[sle]

OPM aside, sharpe is just a dick measuring tool for traders with small accounts. No one i know who trades for a living even tracks it.

Well, it's a good single number to throw out - "oh, my SR is 5, I sleep well"

Couple multi-manager funds/props actually base their payouts on Sharpe (sometimes combined with a return on capital as a 2-d metric, which makes for some really strange strategy incentives).

There is no linear relationship between number of variables and sharpe but I use a couple dozen inputs to my return predictors.

Out of curiosity, are you running it as a single factor-based book or as a portfolio of independent strategies with some netting at the execution level?

 

[ironchef]

I only have a very basic understanding of stats but I think i am understanding why a Sharpe ratio of 3 and above is so desirable.

My current system i am developing (day trading system) has a sharpe ratio of 2 .
It makes 50% a year with a standard deviation of 25% (50:25= sharpe ratio of 2, im ignoring the risk free rate)
As i said my understanding of stats is pretty basic but i think this means that 99.73% of my annual returns should fall between -25% and +125% (50%+/- 75%). As 3 standard deviations of 75% each side of the mean.

Now if my system had a Sharpe ratio of 3, the same 50% a year return but with only 16% stddev. Then the 99.73% range for the yearly returns would be 0% to 100% a year. That would mean no losing years, ever!

On the other hand a sharpe ratio of 1.0 would be quite bad eg 50% a year with 50% stdev.
This means my returns could fall anywhere between -100% and +200% a year.
So you would have to use a really low risk size with a sharpe of 1.0

Excellent OP sir. You gave a very good practical consideration and explanation of Sharpe and the relation to risk and return.

I do have a question for you, consider two strategies: Both have the exact same Sharpe. One has a 50% return and the other 25%. Professional investors would say they have identical "risk adjusted returns" so are the same. For me, a retail investor, the 50% return is intuitively better even though it has the same Sharpe as the 25%. Where do I go wrong if I pick the 50% strategy?

The reason I ask is: I think the Sharpe for small cap is not as good as SPY's but if you buy and hold small cap index for 30-40 yrs, the absolute return is so much better than buy and hold SPY, a puzzle I don't quite understand.

Regards,