more money, worst mar

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
Post Reply
marriot
Roundtable Knight
Roundtable Knight
Posts: 365
Joined: Thu Nov 20, 2008 3:02 am

more money, worst mar

Post by marriot »

Running tests with growing "starting equity", mar keep going down.
I cannot undestand the reason.
Ps not only with over optimized suites.
Attachments
starting eqt.jpg
starting eqt.jpg (57.73 KiB) Viewed 4514 times
sluggo
Roundtable Knight
Roundtable Knight
Posts: 2987
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo »

Do you understand why the numbers in the column "Trades" are different from one another?

26315
27407
27731
27874
27946
marriot
Roundtable Knight
Roundtable Knight
Posts: 365
Joined: Thu Nov 20, 2008 3:02 am

Post by marriot »

Ok, this test runs as up to 100 millions start trading equity, at a certain point trades does not chenge, but mar keep on going down.
Attachments
mar.jpg
mar.jpg (61.2 KiB) Viewed 4446 times
DeanoT
Roundtable Fellow
Roundtable Fellow
Posts: 52
Joined: Sat Apr 15, 2006 7:37 am
Location: Melbourne, Australia

Post by DeanoT »

It seems that you have run the starting equity from $8m to $100m in increments of $1m. How many trades are there at the $8m level ? If the number is less than 28145 then this means that you would have had insufficient equity in your account to take all entry signals at the $8m level.

The fact that performance appears to deteriorate with increasing starting equity is likely due to those missed trades being poor performing trades. Since the higher starting equity values of closer to $100m would have accepted all trades, including the filtered non-performing trades at lower starting equity levels, performance is lower the higher your starting equity is.

Which is why it is good to test all your systems with a very high starting equity ($100m plus) to ensure that your theoretical returns are not being adversely impacted by an issue such as insufficient starting equity.
sluggo
Roundtable Knight
Roundtable Knight
Posts: 2987
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo »

It definitely is surprising and I'm not sure what is the cause.

I experimented with the settings on the Global Parameters tab for about 20 minutes, and was able to get a slightly similar behavior: picture below. Perhaps you might find the cause of your anomaly if you ran a simulation with Global Parameter #1 set at "zero" and another simulation with Global Parameter #1 set at "(your value)".

Then set GP #1 back to (your value) and run a simulation with Global Parameter #2 set at "zero", followed by another simulation with Global Parameter #2 set at "(your value)".

Repeat for each of the Global Parameters.

That's what I did, to get the picture below.
Attachments
Experiments with Global Parameter setting
Experiments with Global Parameter setting
getsworse.png (33.99 KiB) Viewed 4402 times
marriot
Roundtable Knight
Roundtable Knight
Posts: 365
Joined: Thu Nov 20, 2008 3:02 am

Post by marriot »

Yes!
Earn interest set to true/false.
thank you
Post Reply