1% a day consistently: possible?

1% a day consistently, no down weeks: possible?

  • Yes

    Votes: 58 47.9%

  • No

    Votes: 63 52.1%
  • Total voters
    121
  • Poll closed .
S

stephencrowley

General question for intraday automated traders.. no manual intervention except in extreme cases.

In your opinion, Is 1% a day return on average possible with 4:1 buying power? This is after accounting for commissions but excluding things like colo fees, exchange fees, etc.

Backtesting my stuff is showing really good results even after conservatively accounting for commissions, spreads, slippage, etc. My out of sample tests sometimes performs even better than the in-sample data I trained my models with..

Basically, I think my results are too good to be true.. have you ever tried something and thought the same, and then have it turn out to really be true, or in most cases has it really been too good to be true?

I'm always fully hedged (according to historical optimal hedge ratios) and always place hard stop-loss orders immediately upon entering the position so I don't think the results are skewed to some long-market bias..

I really have my doubts whether 1% a day is really possible.. this would mean 250% a year and really good fund managers are lucky to even beat the market.. do you know anyone making this kind of money?
 
that means if you take 25k which is the minimum to be able to use 4:1 leverage you will have more then 1m in just three years. Good luck man.
 
Quote from stephencrowley:

General question for intraday automated traders.. no manual intervention except in extreme cases.

In your opinion, Is 1% a day return on average possible with 4:1 buying power? This is after accounting for commissions but excluding things like colo fees, exchange fees, etc.

Backtesting my stuff is showing really good results even after conservatively accounting for commissions, spreads, slippage, etc. My out of sample tests sometimes performs even better than the in-sample data I trained my models with..

Basically, I think my results are too good to be true.. have you ever tried something and thought the same, and then have it turn out to really be true, or in most cases has it really been too good to be true?

I'm always fully hedged (according to historical optimal hedge ratios) and always place hard stop-loss orders immediately upon entering the position so I don't think the results are skewed to some long-market bias..

I really have my doubts whether 1% a day is really possible.. this would mean 250% a year and really good fund managers are lucky to even beat the market.. do you know anyone making this kind of money?
More...

1% a day compounded would actually be over 900% per annum!

((1.01^20)^12)-1

Who knows it may very well be. I believe Hull noted that he made 400% per annum the first two years he started and 100% from then on.

I would however suggest that you look for any errors in your backtesting system. Somethings simple as looking at the next tick and the placing trades on the previous ones can really skew results.

A 100% per annum is phenomenal as it is. Beyond that is simply astonishing (especially if it has a large capital capacity).
 
Quote from stephencrowley:

General question for intraday automated traders.. no manual intervention except in extreme cases.

In your opinion, Is 1% a day return on average possible with 4:1 buying power? This is after accounting for commissions but excluding things like colo fees, exchange fees, etc.

Backtesting my stuff is showing really good results even after conservatively accounting for commissions, spreads, slippage, etc. My out of sample tests sometimes performs even better than the in-sample data I trained my models with..

Basically, I think my results are too good to be true.. have you ever tried something and thought the same, and then have it turn out to really be true, or in most cases has it really been too good to be true?

I'm always fully hedged (according to historical optimal hedge ratios) and always place hard stop-loss orders immediately upon entering the position so I don't think the results are skewed to some long-market bias..

I really have my doubts whether 1% a day is really possible.. this would mean 250% a year and really good fund managers are lucky to even beat the market.. do you know anyone making this kind of money?
More...

Without talking about specific returns, I have seen around half dozen automated trading cases where > 1% daily being done (present tense), without a single down week for 2-3 years (in 1-2 top cases, actually have only 2-3 down *days* per month). Before everyone reaching for calculators saying this is 600-900% a year, most of these type of systems are not compounding that much, since there are clear strategy capacity issue (such as accounting for a noticable percentage of daily volume traded), but 300-400% a year are done in some of the automated systems that I know. So let's say a system starts the year with $1.5-2M in capital, would return profits of $5-7M a year, actually in a few cases, a little higher.

So the upshot is that yes it is possible, but beware of the capacity of your strategy, and properly size the order sizes. However, that being said, since I know derivatives (options, futures, etc) much better than equities, most of the automated systems that I know (probably 5 out of the 6-7, with only 1-2 equity oriented systems) focuses on derivatives trading. And yes, the names of some of the firms that currently have this type of returns have been asked on ET forums for their information.
 
Well, some days it is 0.5%.. some days 1.5% so it doesn't exactly gain 1% a day, I'd have to fit a distribution to the returns and then project it into the future to find upper and lower bounds within some interval..

Who is Hull? that sounds really interesting..

As far as backtesting.. I've ran hundreds of simulations with different parameters and the large majority of them generate profitibility and my average trade duration is 10 to 30 minutes so I don't think I'm *that* sensitive to sub-second tick flucuations..

Also, as far as scaling capital.. I'm sure there is some upper bound on the amount of money I can push around without affecting the market myself, but then I can always expand the strategy to more symbols/markets/etc.

Quote from mahras2:

1% a day compounded would actually be over 900% per annum!

((1.01^20)^12)-1

Who knows it may very well be. I believe Hull noted that he made 400% per annum the first two years he started and 100% from then on.

I would however suggest that you look for any errors in your backtesting system. Somethings simple as looking at the next tick and the placing trades on the previous ones can really skew results.

A 100% per annum is phenomenal as it is. Beyond that is simply astonishing (especially if it has a large capital capacity).
More...
 
600-900% a year is just insane.. like you said, but wow.. 300-400% is incredible. I assume it's all intraday and hedged? Would this thing blow up with another 9/11 or just take it in stride? I'm fairly confident mine would fare just fine but I really need some high frequency data from one of these days to tell.

Quote from rufus_4000:

Without talking about specific returns, I have seen around half dozen automated trading cases where > 1% daily being done (present tense), without a single down week for 2-3 years (in 1-2 top cases, actually have only 2-3 down *days* per month). Before everyone reaching for calculators saying this is 600-900% a year, most of these type of systems are not compounding that much, since there are clear strategy capacity issue (such as accounting for a noticable percentage of daily volume traded), but 300-400% a year are done in some of the automated systems that I know. So let's say a system starts the year with $1.5-2M in capital, would return profits of $5-7M a year, actually in a few cases, a little higher.

So the upshot is that yes it is possible, but beware of the capacity of your strategy, and properly size the order sizes. However, that being said, since I know derivatives (options, futures, etc) much better than equities, most of the automated systems that I know (probably 5 out of the 6-7, with only 1-2 equity oriented systems) focuses on derivatives trading. And yes, the names of some of the firms that currently have this type of returns have been asked on ET forums for their information.
More...
 
1% of 25k is quite easy. try that with 50k, 100k, 300k, then 500k.....

maybe for swing traders, but i find myself (scalping) reaching a ceiling as to how much i can pull out of the market dollars wise but not percentage. make 2k, 5k or 10k is not difficult, but to make a daily goal of 1-2% is plain rediculous.

oops, my bad. i did not read clearly that question was in regard to automatic trading. i guess its time for me to hit the sack. :D
 
Quote from stephencrowley:


.....

Also, as far as scaling capital.. I'm sure there is some upper bound on the amount of money I can push around without affecting the market myself, but then I can always expand the strategy to more symbols/markets/etc.
More...

A system that works well in one market may fail miserably in other markets (other types of instruments or symbols). I find little evidence for the existence of one-system-fits-all strategies but maybe someone else has found system(s) that can work across a broad range of markets with some parameter tweaking
 
My system is general. I use statistics to verify the underlyings are 'tradeable' and then determine the maxium theoretical profit and then optimize rules to get as close to the maximum theorotical amount as possible while also hedging and limiting drawdown as well.

Quote from giggollo:

A system that works well in one market may fail miserably in other markets (other types of instruments or symbols). I find little evidence for the existence of one-system-fits-all strategies but maybe someone else has found system(s) that can work across a broad range of markets with some parameter tweaking
More...
 
You must log in or register to reply here.
Back
Top