Duly noticed.. I'm off to the sidelines as well, this is making me crazy.
1% a day consistently: possible?
- Thread starter stephencrowley
- Start date
I am having some difficulty with your response. The implication, based on earning 4 or 5% per day is that a trader can earn about 10 to 12 times his original trading capital per annum even without any compounding whatsoever. To be able to do so with any consistency is completely mind boggling to me. To do so with minimal drawdown enters the realm of fantasy from my perspective. Perhaps I am merely projecting my own limitations here. But I don't think so.
Quote from stephencrowley:
Well, hopefully (tomorrow) wed. morning it'll be ready.. I've mainly fixed the quote filtering issue, well enough to start running live anyway. I'm running on about 3 hrs sleep so hopefully I can make it.
I'm cobbling together the code for the main realtime loop. In testing it is very easy because you just fly through the data as fast as it can be read, but in realltime the even main decision-making loop needs to be called. It's not a big deal, just something that has to be done.
The first week is going to be very-very rough, and then only very rough for some undetermined time thereafter.
Stephen,
How did it go? Hope everything is working out for you.
Let us know...
ps: Did you code the latency handler before you go live? If you did, you're one heck of a hack, man. u da maan!!!
I had a few ephiphanys at the last minute and decided to do some more hacking.. one of the helpful posters here noted that mean reversion in the prices themselves can be really handy.. plus some of the coeffecients on my regressions were wacked out and some outliers in the data still.. just too much crap to deal with on so little sleep. I'm a design freak too so I try to balance the hacking/speed with maintanable design so that adds some time as well. Maybe I'll trade this week, maybe not.. I need to overcome my perfectionist instinct and just get it going.
The latency handler wasnt coded into the execution simulator.. I just used the estimates to generate confidence intervals around the entre and exit points.. I've found that even seconds of latency dont affect the overall performance that much.
The latency handler wasnt coded into the execution simulator.. I just used the estimates to generate confidence intervals around the entre and exit points.. I've found that even seconds of latency dont affect the overall performance that much.
Quote from newfool:
Stephen,
How did it go? Hope everything is working out for you.
Let us know...
ps: Did you code the latency handler before you go live? If you did, you're one heck of a hack, man. u da maan!!!
Don't waver, there are always ways to make the the system better. There is nothing like real trading experience, since even in an automated environment, there are temptations after every trade to just make it a "little" better. Also, resist any temptation to shut your system down at anytime when it is trading, unless something drastic like a crash, you need that whole day experience from open to close. Just play with the smallest order size that yet can give you useful data.
I think with two trading days left, you need at least one trading day. Since start trading on a friday is almost perfect, I always bring up new systems on a friday, since you will have the entire weekend (actually 2 days and *three* nights) to fix any problems you may have. And there are no major economic numbers this friday (only durable goods at 8:30, *before* the open). You couldn't ask for better timing.
Good luck.
I am not sure why so many people were so negative on this thread.... In all honesty, if you can pull this off I will just be plain jelous Sometimes I think that is why some people on ET can get a bit too aggressively critical and negative. I think it is entirely possible that you will be able to generate returns averaging around 1%/day. Naturally, if you compound those returns for a while...you will hit a liquidity wall, and slippage, etc will degrade the returns. But that does not mean you won't be sitting on $1 million or even $10 million in accumulated profits by that point
Stephen, you sound like an intelligent fellow with a well thought out and from what i can gather of your posts, thoroughly tested strategy. I would tend to agree with Rufus, try to get in some real trading, even with just a fraction of the size you had intended. It will go a very long ways towards building up your confidence in the system as well.
Rufus and yourself seem to have some good familiarity with these HF systems...I would love to see more threads of this sort on ET. It is nice to log in a learn a bit. I am very curious about HF, but having just a regular commerce (finance) degree...i don't have the robust math/programming background that some on here do, and these sorts of threads can be quite helpful.
When you do start, it would be great to see a journal with some sort of summary/analysis of your trades.
I am skeptical to be sure, but I wish you the best of luck...in a good sense...hahaha...the Renaissance Tech sort of luck!
Sometimes I think that is why some people on ET can get a bit too aggressively critical and negative. I think it is entirely possible that you will be able to generate returns averaging around 1%/day. Naturally, if you compound those returns for a while...you will hit a liquidity wall, and slippage, etc will degrade the returns. But that does not mean you won't be sitting on $1 million or even $10 million in accumulated profits by that point Stephen, you sound like an intelligent fellow with a well thought out and from what i can gather of your posts, thoroughly tested strategy. I would tend to agree with Rufus, try to get in some real trading, even with just a fraction of the size you had intended. It will go a very long ways towards building up your confidence in the system as well.
Rufus and yourself seem to have some good familiarity with these HF systems...I would love to see more threads of this sort on ET. It is nice to log in a learn a bit. I am very curious about HF, but having just a regular commerce (finance) degree...i don't have the robust math/programming background that some on here do, and these sorts of threads can be quite helpful.
When you do start, it would be great to see a journal with some sort of summary/analysis of your trades.
I am skeptical to be sure, but I wish you the best of luck...in a good sense...hahaha...the Renaissance Tech sort of luck!
Very wise advice.. I suspect that no matter how long I work on this there will always be something else to do.. always adapting. Even if it is rewriting this thing in some new languge 10 years from now.
Good to know others are staying up all night squinting at the monitor too
Rufus definately is the expert here and he actually does this stuff for a living so I don't think its fair to group us together.
I'm relatively new (been into this for several years now.. working very intensely on it for the last 6 months)I'll definately start a journal with summaries and random notes.. hopefully the noise will be minimal.
Quote from stephencrowley:
Rufus definately is the expert here and he actually does this stuff for a living so I don't think its fair to group us together.
I'll definately start a journal with summaries and random notes.. hopefully the noise will be minimal.
Expert, heh, I wish I can just be an "armchair expert" and then the profits will just magically flow my way. But the market doesn't work like that, anybody (a discretionary trader or an automated system) have to make trades, make hard decisions and take the necessary risks to pull out the profits.
For me, coming to ET is almost an accident. After I went off on my own, I missed being on the large institutional desks since I would hear everything, everything little rumors about which fund, which ibank is doing, they are poaching so-and-so, etc. So I was hoping that ET would give me some of those "street rumors", but while that is not the case, this is probably the first time I have ever interacted with people outside of large institutional setting, so the new persective is refreshing at times.
Stephen,
I'll be going on a biz trip for a few days and won't have access to the net, so if you go live this Fri as Rufus suggested, I wish you well. If anything, I think you should size yourself on the small side at first. See how it goes. The "market" is still around. No need to rush in big.
Good luck.
-fool
I'll be going on a biz trip for a few days and won't have access to the net, so if you go live this Fri as Rufus suggested, I wish you well. If anything, I think you should size yourself on the small side at first. See how it goes. The "market" is still around. No need to rush in big.
Good luck.
-fool