Intraday Mean Reversion Return Expectations

F

flapjackery

Is it reasonable to expect 2% returns per day with automated intraday mean reversion strategies with holding periods of a few seconds to a few minutes? I'm referring to intraday stat arb-type strategies that for example use cointegration (e.g. trading a basket of stocks against an ETF).

I'm just starting to look at these types of strategies and want to get a realistic feel for expected returns. Starting capital I want to put into these types of trades is probably $100-$150k tops for now, so is it reasonable to expect that I would avoid a lot of slippage and transactional cost due to the bid/ask spread unlike the larger players with >$10m in capital to invest?
 
trendmomentum said:
More like 0.35% with spotless execution and a good short list universe.
More...

Dead nuts fucking on 100 percent. Buying bids, selling offers, and fading moves is a tough way to go. I'm sure that there are traders who do it successfully, but they are 1 %' ers that's for sure.

Best possibility would be for some sort of super high positive correlation arbitrage - like between physical ( cash ) and futures... which would also be next to impossible because the banks and commercials and bigger HFs would be in it.

Maybe trade smaller and think hours or possibly days instead of seconds... take a step back from the turbulence if you will.
 
I see this as being difficult to do for such a short amount of time, I have run back tests of this through the years, between simply not getting filled and commissions which will be high, one healthy loss plus slippage will eat into bottom line. Longer the timeframe makes more sense and having very low losing percentage can overcome the breakeven trades, which are losers cause costs to get in and losing trades. Going for very short profits seem to work best with huge sums of shares/contracts but that said, much more at risk. I have not seen any 2% methods per day last more than couple months of high volatility and my tests shows losses further testing like three years out. It is possible of having 1-2 years of making over 400% return in a year, but that has usually occurred by luck of picking one instrument and holding it. To shoot for 2% return per day and get everyday to me means you will be risking a great deal and often losing days will be huge. This certainly would not be considered investing. Often times I have tried, simply can't get filled in time to get difference in underlying.

Good luck.
 
bone said:
Dead nuts fucking on 100 percent. Buying bids, selling offers, and fading moves is a tough way to go. I'm sure that there are traders who do it successfully, but they are 1 %' ers that's for sure.

Best possibility would be for some sort of super high positive correlation arbitrage - like between physical ( cash ) and futures... which would also be next to impossible because the banks and commercials and bigger HFs would be in it.

Maybe trade smaller and think hours or possibly days instead of seconds... take a step back from the turbulence if you will.
More...

I'll take 0.35% as a benchmark then. What is the average sharpe ratio on these types of trades? The reason I was shooting for shorter holding periods was to decrease daily return variances (i.e. Law of large numbers). I think hourly holdings are reasonable too. In your experiences, have sharpe ratios over a fixed interval typically increased with the number of trades?
 
flapjackery said:
I'll take 0.35% as a benchmark then. What is the average sharpe ratio on these types of trades? The reason I was shooting for shorter holding periods was to decrease daily return variances (i.e. Law of large numbers). I think hourly holdings are reasonable too. In your experiences, have sharpe ratios over a fixed interval typically increased with the number of trades?
More...

I think the better consideration is your selection of names/ instruments to be traded, a viable strategy in terms of entry/exit/risk/position management that fits those names/instruments, and trade execution abilities and what kind of platform and ECN infrastructure you realistically have at your disposal in order to accomplish the trade in the holding timeframes your strategy has been optimized for. Talk about Sharpe Ratios is meaningless until you solve for these considerations.

And that's just my opinion. YMMV
 
flapjackery said:
Is it reasonable to expect 2% returns per day with automated intraday mean reversion strategies with holding periods of a few seconds to a few minutes? I'm referring to intraday stat arb-type strategies that for example use cointegration (e.g. trading a basket of stocks against an ETF).

I'm just starting to look at these types of strategies and want to get a realistic feel for expected returns. Starting capital I want to put into these types of trades is probably $100-$150k tops for now, so is it reasonable to expect that I would avoid a lot of slippage and transactional cost due to the bid/ask spread unlike the larger players with >$10m in capital to invest?
More...


Don't expect to receive guidance on your expectations for profit or loss on any trading system or investment, from anyone. Especially here. There are hundreds of variables that can impact performance - many of which will never be under your control - and not a single trader can project their own future, much less yours.
 
lindq said:
Don't expect to receive guidance on your expectations for profit or loss on any trading system or investment, from anyone. Especially here. There are hundreds of variables that can impact performance - many of which will never be under your control - and not a single trader can project their own future, much less yours.
More...

Then your reply is probably the best guidance I've received in this thread. Thanks for the heads up. Time to start coding.
 
You must log in or register to reply here.
Back
Top