Automated trading system developer looking for a contract.

elabunsky

elabunsky

Hello,

I'm looking hedge fund or individual who will like start automated trading on futures market.
Currently I'm developing system for futures market.
I tried delete curve fitting (uses Walk Forward Test only and no backtest) and simulates market (I'm using send order simulation and see when order executed after price over/down of order price ... depends on position's side). For simulation I'm using ticks.
I have enclosed excel file with the 2 month simulation FDAX sample.
Before you reply I like notice that I need small support only before creating complete solution (I estimate that I will able complete it during 4 months) and then I like get resonable % from profit.

Regards,
Eugene.

PS. Please, contact me via private message.
 

Attachments

I have calculated E-Mini ES & NQ samples also.
The same way - no curve fitting and market execution simulation.
Review the results below.
 

Attachments


I tried delete curve fitting (uses Walk Forward Test only and no backtest)


I do not understand the above, would you mind to explain in detail?
 
I checked the Excel sheet and the trades.

This system is not tested properly and dillusional. Realistic execution (FIFO... slippage) will kill the system.

You may be expecting someone to provide you with Low comission or low latency networks to overcome FIFO but what you provided will not work under:

REAL MARKETS

It's what we call the Magic Fill phenomenon. You're always first in line with all the orders sent.

Yes. This model doesn't require "BACK-TESTING". It's a testing flaw. Also, I'd like to note that any kind of manual entry or discretion will make things worse.

Been there, done that.
 
Quote from TSGannGalt:

I checked the Excel sheet and the trades.

This system is not tested properly and dillusional. Realistic execution (FIFO... slippage) will kill the system.

You may be expecting someone to provide you with Low comission or low latency networks to overcome FIFO but what you provided will not work under:

REAL MARKETS

It's what we call the Magic Fill phenomenon. You're always first in line with all the orders sent.

Yes. This model doesn't require "BACK-TESTING". It's a testing flaw. Also, I'd like to note that any kind of manual entry or discretion will make things worse.

Been there, done that.
More...

Why you think that my system not sim. market execution? Because trades have not SS? Or profits similar? Can you undestand that when price over or less than order price it mean that market execute it? Or it's so hard for you?
 
1. You use say you use tick compression for the tests but the test results are under a 1-min. increment. In a tick chart, you should be providing us with a sub-second result. Even more you have:

20080403 07:13:00 B 1 6812.5 1 -170312.5 555.00
20080403 07:13:00 S 1 6815.5 -1 170387.5 1 75.00 2 72.50 627.50 3.00 2.90

....

There's a large possibility that your testing engine iterates market data bar-by-bar (under a 1-min. bar compression), and the trade above gets filled at the same iteration count (open/close at the same time).

Ex. You have a short term system that places both a buy and sell trade. The high price of the bar is 100 and low price is 90. The Buy limit is 91 and Sell limit is 99. The program runs a buy and a sell at the same time (Sell 99 and Buy 91):

if (new bar)
{
if (selltrigger == true)
{
OpenTradeSell(); // This will initiate the Sell trade
}
if (buytrigger == true)
{
CloseTradeBuy(); // Because the Sell trade was initiated above, the program will assume that the position is short, regardless of when it went short.
}
}
nextbar();

This is only an example...

Want more? I got a few more but I'll start here... Please post an Excel file with a better indication of the system being tested more "realistically".
 
Quote from TSGannGalt:

1. You use say you use tick compression for the tests but the test results are under a 1-min. increment. In a tick chart, you should be providing us with a sub-second result. Even more you have:

20080403 07:13:00 B 1 6812.5 1 -170312.5 555.00
20080403 07:13:00 S 1 6815.5 -1 170387.5 1 75.00 2 72.50 627.50 3.00 2.90

....

There's a large possibility that your testing engine iterates market data bar-by-bar (under a 1-min. bar compression), and the trade above gets filled at the same time (open/close at the same time).

This is only an examples...

Want more? I got a few more but I'll start here... Please post an Excel file with a better indication of the system being tested more "realistically".
More...

I got ticks from TradeStation. TradeStation do not provide SS info for ticks. This is why in trader SS info is 00.
During simulation I'm reading tick by tick and I push trades in format YYYYMMDD HH:MM:00.
I have attached part of tick file and you can review it by yourself.
If you have more questions - send me private message, please.
 

Attachments

Quote from 88888888:


I tried delete curve fitting (uses Walk Forward Test only and no backtest)


I do not understand the above, would you mind to explain in detail?
More...

Sorry I missed your post.
What is backtest (curve fitting)? When you calculate system on some period and then analyze system. You are looking better params, rules etc... In other words you are doing equity curve better on KNOWN data and can suppose how it will work for future only. My way is Walk Forward Test.
I defined rules and do a lot of tests (for example, for each trading day) each time apply calculated system on next trading day. As result I have some paper trades. When I review trades given each time as result of apply system on UNKNOW data I can see how system will work for future.
 
After comparing the excel sheet with tick data of the DAX. the exits look plausible.
They are just fixed brackets around the entry.

But the entries only look possible if they are limit orders that are placed well in advance.
Eugene, can you tell us if this is the case?
 
You must log in or register to reply here.
Back
Top