My system is not a profitable in all markets

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
Posts: 1
Joined: Fri Aug 13, 2010 8:59 pm

My system is not a profitable in all markets

Post by thuesand » Sat Aug 14, 2010 8:13 am

My system is only a marginal profitable in many markets, and are unprofitable in few, but makes a killing in some. Overall the performance seems descent.

My question is if you leave out bad performering markets in your portfolio when you go live with your system?

Moto moto
Roundtable Knight
Roundtable Knight
Posts: 427
Joined: Mon Jun 01, 2009 4:12 am
Location: once again in the UK

Post by Moto moto » Sat Aug 14, 2010 9:51 am

I would say include everything for diversification and smoothing of the total PL over a continually compounded long timeframe - not just one year say.
Unless of course you have great foresight and know that some markets will never contribute. I for one would love to have a massively concentrated portfolio in only winning positions.

which would you choose - the returns of market A, market B or a combination of the markets.

Code: Select all

	         Mkt A		                  Mkt B	   TOTAL
return %	 100	         return %	     100	   200
10%	      110	            0%	        100	   210
-10%	      99	            0%	        100	   199
10%	      108.9             0%	        100	   208.9
-10%	      98.01           10%	        110	   208.01
10%	      107.811           0%	        110	   217.811
-10%	      97.0299         10%	        121	   218.0299

Roundtable Knight
Roundtable Knight
Posts: 513
Joined: Wed Apr 16, 2003 1:18 am
Location: Nowhere near

Post by Kiwi » Sat Aug 14, 2010 12:41 pm

Before you leave any out consider:
- does an unprofitable market have profitable periods that smooth your equity curve?
- do you know why it is unprofitable (and can thus say that it is never likely to be profitable).

Roundtable Knight
Roundtable Knight
Posts: 218
Joined: Wed Aug 09, 2006 4:01 pm

Post by nodoodahs » Sat Aug 14, 2010 9:22 pm

Along the lines of what Kiwi asked ... of the unprofitable markets, what do they have in common? Are they all physicals, currencies, stock indices, bonds? How is the profitability in the other markets of that type? I would say that if the system had nothing but unprofitable or breakeven performances in one of the four major classes of futures contracts, or in one of the subclasses of the physicals (softs, energy, hards), then perhaps this is a property of these contract types that your particular system is exposing? If this is the case, I would consider excluding that entire class.

Assuming there's not a particular class/subclass that your system "dislikes" then stop to consider how many trades were taken in the unprofitable markets, and how do those trades compare to the trades taken in the profitable ones? If it looks like a random distribution, then perhaps it IS random and I would include those contracts, because if it were just random that they underperformed in that particular backtest, then I have no guarantee or reason to assume that they they will necessarily underperform in the future.

Bud Fox
Contributing Member
Contributing Member
Posts: 9
Joined: Sat Aug 07, 2010 12:14 pm
Location: .de

Re: My system is not a profitable in all markets

Post by Bud Fox » Sun Aug 15, 2010 4:16 pm

thuesand wrote:...makes a killing in some.
Depends on what "some" means for you.

The question could also be: how close am I allowed/able/willing to get to curve fitting while designing a system? This is where emotions (like greed) may come into play. Try to leave away the emotions. Don't ask yourself what are the pros of this system. Ask yourself what are the cons instead. Do it know instead of later. Doing it know doesn't cost you money, but learning from mistakes always does, in this business.

Post Reply