Day trading and Sharpe Ratio

[QuasWexExort]

Win rate: 40%
Average winner: 2%
Average loser: 1%

A trading system with the above metrics is not an outstanding trading system, it is pretty mediocre in fact

With stats like that and multiple opportunities PER DAY on average it is hardly mediocre. I'd pay 6 figures for it if it was mechanical, back tested 15 years+, forward tested 3years+ or back the trader if he had a consistent performance like that overf the past few years.

What is an outstanding system to you ? 55% and 3:1 ?

 

[SabreMan]

With stats like that and multiple opportunities PER DAY on average it is hardly mediocre. I'd pay 6 figures for it if it was mechanical, back tested 15 years+, forward tested 3years+ or back the trader if he had a consistent performance like that overf the past few years.

What is an outstanding system to you ? 55% and 3:1 ?

The 'expectancy' of that system is 0.2
Which means it makes a profit of 0.2% on average for every trade taken with 1% risked.
Obviously that is better than a system with an expectancy of 0.1.
But not as a good as one with 0.33 or 0.5 or 1.0...

But as i showed in the opening post even a 0.2 expectancy system is fantastic if you can get 4 trades a day out of it.

 

[truetype]

Consider this theoretical trading system (.. ) As you can see from these results, Sharpe ratios above 2 and 3 are possible when day trading

Yes, everyone had a high Sharpe in backtest.

 

[sle]

The Sharpe Ratio is considered to be laughable by professional traders because it punishes upside volatility as if it was downside volatility.

That is definitely not true, all professional traders I know use Sharpe (or some call it “information”) ratio. Drawbacks are known and all stem from the assumption of normal distribution.

Penalties for upside volatility are somewhat fair unless your returns are skewed by design (eg an option buying strategy). The upside gaps that skew your sharpe could be a result of lucky positioning and next time it might go against you.

 

[schweiz]

Penalties for upside volatility are somewhat fair unless your returns are skewed by design (eg an option buying strategy). The upside gaps that skew your sharpe could be a result of lucky positioning and next time it might go against you.

That is complete nonsense to me. I daytrade over 20 years (even before Emini existed). EVERY WEEK my sharpe ratio would be better if I would remove 1 profitable trade. So it has nothing to do with luck, at least not for me.
I have small losses and big profits, and at the same time very few losing trades. The impact of a big winner is never compensated by a big loss, so Sharpe ratio is not working correctly in that case.

Sharpe Ratio Calculations - Using Daily Returns


Explanation by @nonlineair5:
The Sharpe ratio penalizes the "fat tails", i.e. abnormally high and abnormally low returns. Under certain conditions, it may lead to what's known as a violation of the "first order stochastic dominance" rule. In laymen terms, it means that Sharpe ratio may lead to nonsensical results when evaluating performance.

 

[bln]

Agree Sharpe is a not a good measurement tool of risk vs reward.

A trader with with a 5:1 PL ratio would get lower Sharpe than a 3:1 even if the risk per trade is exactly the same. Due to punishment of up side volatility of positive returns.

That greater returns equals greater risk are not always true.

Shape is easy to calculate youself in a spreadsheet: youtube

 

[sle]

That is complete nonsense to me. I daytrade over 20 years (even before Emini existed). EVERY WEEK my sharpe ratio would be better if I would remove 1 profitable trade. So it has nothing to do with luck, at least not for me.
I have small losses and big profits, and at the same time very few losing trades.

It is possible that you have found the holy grail of intraday strategies. However, my prior will be that you have a strong hindsight bias and remember your big winners better than you remember your big losers.

Explanation by @nonlineair5:
The Sharpe ratio penalizes the "fat tails", i.e. abnormally high and abnormally low returns. Under certain conditions, it may lead to what's known as a violation of the "first order stochastic dominance" rule. In laymen terms, it means that Sharpe ratio may lead to nonsensical results when evaluating performance.

As I said, Sharpe Ratio assumes that your returns follow a normal distribution and that you have sufficient sample. Once that assumption breaks (by design or by chance), you will get a results that makes no sense - either you will get a Sharpe that’s too good or you’ll get a Sharpe that’s too bad. Like any descriptive statistic, you have to understand it’s limitations.

 

[sle]

I’ll add something cynical.

Everyone here seems to have a Sharpe of 3+ and their strategy has a sufficient upside skew to cause distortion in the information content. Also it seems like everyone has traded for 20 years, has predicted every great crash and has never has a losing year.

“They don’t make them like they used to and they never have”

 

[traider]

I’ll add something cynical.

Everyone here seems to have a Sharpe of 3+ and their strategy has a sufficient upside skew to cause distortion in the information content. Also it seems like everyone has traded for 20 years, has predicted every great crash and has never has a losing year.

“They don’t make them like they used to and they never have”

No, I don't have any sharpe 3. My average strategy is probably sharpe 0.3 to 0.7. Run a bunch of them so the overall equity curve is a bit nicer. That's why I keep saying that sharpe works great for most people. Even buffett can be measured by sharpe without being penalised let alone the average guy.

 

[comagnum]

That is definitely not true, all professional traders I know use Sharpe (or some call it “information”) ratio. Drawbacks are known and all stem from the assumption of normal distribution.

Source - Investopedia. (see link below)
The Sharpe ratio is a good measure of risk for large, diversified, liquid investments, but for others, such as hedge funds, it can only be used as one of a number of risk/return measures.

The SR is better suited to investors in something like mutual funds or fixed income or those running something like a big pension fund.

The SR has been sold to the public by an industry of fund managers unable to beat the SP500 in order to justify their fees- "hey, we may not beat the market but check out our SR".

I sure would not center my trading around the Sharpe - to do so I would sharply lower my profits and in no way would it give me better risk mgmt.

https://www.investopedia.com/articles/07/sharperatio.asp