What are real returns achieved by retail algo traders?

A

alex314159

Hello everyone,

So the title says it all really - what returns are people achieving in live trading? I'm trying to benchmark my efforts to what others are achieving and can't find any good data.

On my side I've been running a low frequency (end of day, several trades a month on several stock index futures) algorithm for a year. It's returned 17% with about 20% volatility - the return was basically in line with my backtesting but the volatility was much worse. Clearly not enough to stop my day job even if I was well enough capitalised!

Thanks,
 
alex314159 said:
Hello everyone,

So the title says it all really - what returns are people achieving in live trading? I'm trying to benchmark my efforts to what others are achieving and can't find any good data.

On my side I've been running a low frequency (end of day, several trades a month on several stock index futures) algorithm for a year. It's returned 17% with about 20% volatility - the return was basically in line with my backtesting but the volatility was much worse. Clearly not enough to stop my day job even if I was well enough capitalised!

Thanks,
More...

That is extremely good! That kind of Sharpe Ratio would put you solidly in the middle of the pack in the CTA space over the last year. And its been quite a good year for CTA's compared to recent history.

Here is my track record so far. It's a sample of one.

http://www.elitetrader.com/et/index.php?threads/fully-automated-futures-trading.289589/

I personally expect to get a long average of a SR of 0.5 from my system. To get that you need to trade many more markets - I have 45 markets spread over all the main asset classes. Having 'several' stock index futures isn't enough. You'd be better trading one future from each asset class, or if you had the capital.

Okay the last year was a lot better than that, but as I said its been an exceptional period for CTA's. I'm also a conservative, pessimistic, kind of guy. As you need to be to make your living from this.

If a SR of 0.5 that isn't enough to give up your day job then yes you need more capital.

You can't expect high returns from low frequency trading (law of active management). On the upside your costs, leverage, and hassle are lower. Its also possible to continue doing it part time.
 
alex314159 said:
Hello everyone,

So the title says it all really - what returns are people achieving in live trading? I'm trying to benchmark my efforts to what others are achieving and can't find any good data.

On my side I've been running a low frequency (end of day, several trades a month on several stock index futures) algorithm for a year. It's returned 17% with about 20% volatility - the return was basically in line with my backtesting but the volatility was much worse. Clearly not enough to stop my day job even if I was well enough capitalised!

Thanks,
More...

Just to add: this is very worrying "the return was basically in line with my backtesting but the volatility was much worse." It could mean you have made a major error in implementing your system. What volatility were you targeting? How does the volatility look compared to the range seen in historical periods in your backtest?
 
are you calculating return from notional or minimum margin requirements ?. if from minimum margin my return been about 50 percent per month for the last 6 month.
assuming ib margins
 
Last edited:
ScottColeFTA, thanks for your link, was very useful!

globalarbtrader, thank you, your results are extremely impressive! On the volatility point: my backtesting was indicating returns in the 20% range with vol in the 10% range. As it happened my vol was 2x as bad. The implementation looks OK, it seems to me like it's just slippage between real life and simulated trades - I'm very careful in keeping lots of out of sample data whenever I test strategies but some bias is, I think, inevitable.

AdrianHagh81: interesting point. I never thought about doing it this way. I'm low frequency so I sometimes experience decent drawdowns, at least intra-day/week, so I need to have much more than the minimum margin on my account. My position size also changes. Based on my maximum exposure, minimum margin requirement, I made about 40% over a year, a far cry from your results!
 
alex314159 said:
globalarbtrader, thank you, your results are extremely impressive! On the volatility point: my backtesting was indicating returns in the 20% range with vol in the 10% range. As it happened my vol was 2x as bad. The implementation looks OK, it seems to me like it's just slippage between real life and simulated trades - I'm very careful in keeping lots of out of sample data whenever I test strategies but some bias is, I think, inevitable.
More...

Sorry to keep banging on about this but this is a very important detail. What do you mean by slippage? If it's what I think you mean it would lead to lower returns but not higher vol. It's key that you know if you're vol targeting is correct
 
alex314159 said:
ScottColeFTA, thanks for your link, was very useful!

globalarbtrader, thank you, your results are extremely impressive! On the volatility point: my backtesting was indicating returns in the 20% range with vol in the 10% range. As it happened my vol was 2x as bad. The implementation looks OK, it seems to me like it's just slippage between real life and simulated trades - I'm very careful in keeping lots of out of sample data whenever I test strategies but some bias is, I think, inevitable.

AdrianHagh81: interesting point. I never thought about doing it this way. I'm low frequency so I sometimes experience decent drawdowns, at least intra-day/week, so I need to have much more than the minimum margin on my account. My position size also changes. Based on my maximum exposure, minimum margin requirement, I made about 40% over a year, a far cry from your results!
More...

I might just be on a lucky streak from a curve fit, but hey so far so good.
I trade intraday only, no overnight.
I gave up ony any complicated position sizing.
So been trading with a fixed contract amount for the past 3 month.
As soon as I find a system that works for other markets I'll deploy it, won't try to
compound the current one.
 
globalarbtrader - I don't have a good answer for you. When I went live I expected similar vol as my backtesting and lower returns, but the opposite happened, mainly because of the November/December vol in markets with my system trading a lot (traded almost every day). It does worry me a bit, but I'm always very conservative on leverage to account for this.
 
I like the following analysis from Meb Faber:

http://mebfaber.com/2009/12/03/where-have-all-the-sharpe-ratios-over-1-gone/

All professional fund managers have problem to achieve Sharpe over 1 in a long term. You may have nice numbers short-term (3-5 years). But it is very hard to achieve it long term. So I would start with that. I do not believe retail traders have better numbers that professionals.

Here is also fine list of managers, you will have problem to find somebody with Sharpe over 1.

http://www.managedfutures.com/top_cta_rankings.aspx
 
Last edited:
You must log in or register to reply here.
Back
Top