[LeeDanello]

use the quote button?

[LeeDanello]

Julian Cohen wrote:

Sorry Paul, I only just saw your post.

I changed my settings on Fusion to allow 4 threads and 8 GB of RAM to the virtual machine. As long as I don’t try and run anything like Photoshop and leave the MacBook Pro alone to run the MCS, I can get 1000 runs on the Russell 1000 from 2010-2016 on a 25% trade skipping done in an hour and a half. I’m quite happy with that.

I don’t think it’s worth my while to get a Windows machine just to run Monte Carlos faster, so I’ll keep going as is for the moment.

That speed is about par for the course

[LeeDanello]

I can’t remember where I found this, but could be useful albeit a little late

[LeeDanello]

Hi Tony,
My 2c worth. I had no clear performance metrics on how an alternative system to my long term system would perform before I started to design it. I just dived in and thought “let’s see where this takes me”. I think the point is to try and diversify your systems to perform differently to each other so when one is not working as well another might be making you the money. I have a long term system that trades in sync with the index. I thought that having a shorter term system that traded when the longer term system was off might improve my portfolio returns and improve the robustness of the portfolio. The mean reversion system l designed doesn’t use an index filter so it’s always active, it has a high win rate so it should keep me happy and testing shows that it’s profitable which hopefully will compliment the returns of the long term portfolio. I had no specific objectives for it other than it should be reasonably profitable, have a low drawdown and it allows me to diversify my trading system/style to compliment my bottom line. My idea of diversifying my portfolio would be to have a short term system ie one that holds in days, a medium term system one that holds in weeks and a long term system one that holds on for months to years. So I’m testing the short term system in a simulated account and when I’m happy that it performs as tested, I’ll start trading that and then I will start designing a mid term or momentum system.

[LeeDanello]

Why not use the MAR ratio and pick the highest one

[LeeDanello]

I’m assuming you’re using the API. IF you have 5 slots available in your portfolio and you have 50 pending orders. Do they get filled on a 1st in basis or alphabetically?
On a down day like today you’d expect most if not all to be filled. Do you set limits on your practice portfolio equity from the start. I know the notional starting equity for the IB practice portfolio is $1M. Just curious of the machinations.

[LeeDanello]

Depends on your capital. If you don’t have much to start with ie less than 50k, I think it’s pretty pointless having 20 positions. I think with a high win rate, 10 positions is ok. Your system will make more with a higher drawdown but if you accept the testing as being Ok then why not.

[LeeDanello]

Darryl, I love your enhanced profit table

[LeeDanello]

I’ve tried the search function but end up going back to the topic index and click where I think things might be residing and keep clicking until I find it.

[LeeDanello]

Yeah I have it saved. It’s a matter is setting up the interface to remove all the clutter. I don’t generate enough orders to use the API so whatever I have goes in manually. I might have a look at the spreadsheet at the Chartist forum where someone wrote a macro to place orders. See if that suits my needs.

And also Saids spreadsheet.

[LeeDanello]

Finally started some simulated trading on IB. Going to test the system on the ASX and Russell 1000. Navigating around the menu the IB menus can be a bit daunting if you haven’t done in a while so I’ll be sticking to the classic interface.

[LeeDanello]

Julian Cohen wrote:

If Amibroker will attempt to convert when you use a different currency then if you set the AUD/USD exchange rate as 1:1 then you shouldn’t have to worry about changing them when testing different markets. That way it should always be producing a correct result. Does that make sense or am I missing something?

It’s obviously a check that Amibroker does in case you want to use some sort of currency conversion in your backtesting so if you don’t get that right it can screw up your backtest start date.

[LeeDanello]

This is what happened when I had the wrong base currency when I tested on the US

And when I changed it to the right currency the backtest started in the right year

This is what Richard Dale wrote to me “AmiBroker will attempt to perform FX rate conversion using your base currency”

[LeeDanello]

I’ll try and reproduce the issue I had

[LeeDanello]

I would do both. I’ve had issues with the backtester when the base currency was different.

[Author]

Posts