Day trading and Sharpe Ratio

[southall]

To know if it is really as impressive as it looks we should know if:

1. the robot received a rubik cube without knowing in front how the different patterns on each side would look
   
2. the robot received a rubik cube knowing in front how the different patterns on each side would look

The first case would not have slowed the robot down in any way.
It is not difficult for a computer to read the faces of the cube without error. They are very simple coloured squares. And optimal software solving algorithms for the cube were also developed decades ago.

 

[ironchef]

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

There are better metrics for the lumpy returns that are inherit with trading such as: profit-factor, closed NAV, Calmar, and gain-to-pain to name a few.

I always wondered about and tried to understand Sharpe. Those who has an edge in trading usually has higher upside than downside, so of those you mentioned, which metric is a better measure?

Thanks.

 

[ironchef]

I keep things simple if I want to measure the performance of my trading:

• return
• drawdown
• time to recover

is all I watch.

Is that sufficient to know whether my system truly has an edge over buy and hold an index like SPY?

A hypothetical question: If my return is 2x SPY, my drawdown is 2x SPY my time to recover is the same, do I have an edge?

 

[schweiz]

Is that sufficient to know whether my system truly has an edge over buy and hold an index like SPY?

A hypothetical question: If my return is 2x SPY, my drawdown is 2x SPY my time to recover is the same, do I have an edge?

You should analyze these figures and find the answer yourself. For me it is not a (good enough) edge. I think you need better ratio's.

My return is a multiple of my drawdown and my recovery time is less than 48 hours. As long as it stays like that I only watch: return, drawdown and recovery time.
Given my real ratio's I don't need any other proof of having an edge.

The first warning signal would be the evolution of my account in$$$. Then I would watch avg win vs avg loss, win/loss ratio, evolution open P/L...

A hypothetical question for you: If my return is 10x SPY, my drawdown is 1x SPY, and my time to recover is less than 3 days, do I have an edge?

 

[ironchef]

A hypothetical question for you: If my return is 10x SPY, my drawdown is 1x SPY, and my time to recover is less than 3 days, do I have an edge?

Then you would become a very wealthy trader in a short time!

I thank you for the response. It is a quick way of keeping score for a short term trader.

On second thought, perhaps not a sufficient gauge for long term investors selecting securities for multi-decade buy and hold as I could select, in backtest, different securities with the same three numbers but very different risk profiles.

 

[schweiz]

On second thought, perhaps not a sufficient gauge for long term investors

Daytrading is a completely different game from LT trading/investing for most people.

Out of curiosity I tested my daytrading system on trading stocks on longer terms in past. It did not perform bad, although the test was not extensive enough to make any serious conclusions.

For me daytrading is at this moment all I am interested in. I never have to watch any positions overnight, and returns are much higher, so that I can skip days, weeks or even months if I don't want to trade.

 

[southall]

My return is a multiple of my drawdown and my recovery time is less than 48 hours.

Is that statement hypothetical or are you really claiming you never have a drawdown longer than 2 or 3 days?

 

[schweiz]

Is that statement hypothetical or are you really claiming you never have a drawdown longer than 2 or 3 days?

I have no drawdowns that last over 48 hours.
I am daytrading and after a losing trade (which is normally less than 3 points ES as that is my stop) I switch automatically to 1-3 point take profits. I did not have 2 consecutive 3 point stops in a row for several years. Most of the time I manage to recover within 24 hours.

 

[southall]

I did not have 2 consecutive 3 point stops in a row for several years. Most of the time I manage to recover within 24 hours.

It sounds like you average down or have a martingale strategy.

 

[Xela]

It sounds like you average down or have a martingale strategy.

Really? It doesn't sound that way at all, to me.

To me it just sounds "normal-to-good" for someone making their living by doing fast-ish intraday trading, with multiple trades per day. (I have no idea whether that's what Schweiz is actually doing - that's just how it sounds, to me.)

I also can't remember the last time I had a drawdown that lasted over 48 hours, and there's nothing particularly clever or difficult about the way I trade: I just have a genuine edge combined with a decently high win-rate (together, in my case, with a significant proportion of trades that either break even or more commonly make a tick or two in profit, which I consider a "break-even" to all intents and purposes), and I trade very frequently, by general standards.