Unfavourable Risk:Reward Ratio

[simcom]

I remember seeing a conversation along these lines as part of another thread a few days ago, however despite much searching I couldn't find it again so decided to start a new thread.

My question is this: What are peoples' views on trading a system that has an unfavourable risk:reward ratio (terrible - something to the tune of, for instance, 6:1)?

The system in question does have a positive expectancy, with historical results being 66% winning trades and the average winner being 0.9 times the average loser. As you can imagine, the largest percentage drawdown is big (32%), but is this necessarily a problem provided that appropriate money management is used? Or is this system too difficult to trade on a practical/psychological level?

Thanks for your views.

 

[shadowninja]

I'm not mentioning the person's name in case Spanish89 rips my head off when he sees I've mentioned his threads again.










Oopsie.

Read the replies to his ideas.

Check this out, too: Random Equity Curve Simulator of a trading system. Learn it before you trade

 

[simcom]

Ah, yes thanks shadowninja - that might very well be where I saw the discussion. Cheers!

 

[TheBramble]

Ditch R:R. It's the last thing you need to be worrying about.

 

[simcom]

Ditch R:R. It's the last thing you need to be worrying about.

Thanks for your comment, TheBramble. I have been reading your 'Trading 201' thread with interest so expected your stance to be along these lines!

In terms of the maximum drawdown (32%) - does this pose a serious problem or is it acceptable with good money management?

 

[TheBramble]

In terms of the maximum drawdown (32%) - does this pose a serious problem or is it acceptable with good money management?

Of course it poses a serious problem. How many max drawdowns in what space of time/number of trades would cause you to reconsider your committment to this system?

 

[simcom]

Of course it poses a serious problem. How many max drawdowns in what space of time/number of trades would cause you to reconsider your committment to this system?

Thanks for your reply.

Is this so, even though this drawdown figure (actually considerably less than stated above - please see below amendment) assumes that constant trade sizes are being used, when in reality the trade size would be decreased as bank depleted?

EDIT:
Shortly after posting the initial message I realised that the percentage drawdown figure is wrongly based upon a level of equity which was started from zero at the beginning of the test period (I calculated it quickly on a spreadsheet). I have recalculated this using actual bank size and the figure is 11%. Is this still large enough to pose a problem, bearing in mind this is from a very large sample size?

 

[TheBramble]

Thanks for your comment, TheBramble. I have been reading your 'Trading 201' thread with interest so expected your stance to be along these lines!

I'm glad you've found it interesting. The style (and length) of the posts in that thread don’t make for ‘easy’ reading as has been pointed out so, so I’ll let that one die a natural and I’ll keep my suggestions to a more comfortable word-bite size for the hard-of-thinking that seem to comprise the majority on this site (not you mate, but the others….LOL) and dot my comments in here and there as appropriate. Here is one such place…

The reason I’m so vehement about this stuff is that as part of my trading education, I’ve been paying attention to stuff like this, but to the detriment of my trading.

As always, it’s a personal style thing, but why treat drawdown as an inevitable expense? Commissions are an unavoidable expense. Spreads are an unavoidable expense. Drawdown is a statistician’s excuse for lack of performance.

I don’t mind, I really don’t mind, taking a number of small (really very small) consecutive losses while I remain convinced of the more likely movement of the instrument I’m trading. It just means my micro-timing is crap, that’s all. When my conviction falters, I remove my bias and wait for a new bias to assert before recommencing my strategy. I’m either right or I’m wrong., And if I’m wrong, I wont wait too long to be convinced of it.

But the thing is, I don’t feel too happy taking large losses, even cumulatively. It isn’t how businesses of any kind flourish. You do something right – you get paid. You do something less than stellar – YOU GET OUT BEFORE YOU GET FOUND OUT. You don’t sit tight in the comfort that your losses don’t yet equal your ‘max drawdown’. Jeeezzzzz…..

Drawdown is a comfort blanket for incompetence. Don’t pretend drawdown is all part of the plan. If you do, you’ll get flucked. Drawdown mustn’t feature as an ‘OK possibility’ in your trading plan or have any absolute or relative value or it WILL become a reality for you at some point….you do make these things happen. Trust me.

If you’re not right more than wrong before your max risk gets taken out – however many trades you are prepared to test feed that into – then you’re wrong. Plain and simple. Come back another day when the market’s right or you are.

Say NO to drawdown….LOL.

 

[TheBramble]

EDIT:
Shortly after posting the initial message I realised that the percentage drawdown figure is wrongly based upon a level of equity which was started from zero at the beginning of the test period (I calculated it quickly on a spreadsheet). I have recalculated this using actual bank size and the figure is 11%. Is this still large enough to pose a problem, bearing in mind this is from a very large sample size?

Man, saw this after I posted my post above.

This is the thing. You're working with numbers like they don't matter too much. Ooops, sorry, I meant 11% not 32%.

That's the whole point. While it's an academic exercise a few percentage points don't matter that much....

When it's your own real money on the table - a quarter percentage point makes a GREAT deal of difference.

Forget the academic formula and measures. Concentrate on taking money the best way you can.

 

[simcom]

Man, saw this after I posted my post above.

This is the thing. You're working with numbers like they don't matter too much. Ooops, sorry, I meant 11% not 32%.

That's the whole point. While it's an academic exercise a few percentage points don't matter that much....

When it's your own real money on the table - a quarter percentage point makes a GREAT deal of difference.

Forget the academic formula and measures. Concentrate on taking money the best way you can.

Lol, yes I suspected that it might appear that way. Although I must stress that this is a hypothetical strategy that I have been giving figures for - rather than one I am currently trading or intend to start trading - so that I could put the discussion to the table. Hope that explains my (normally) rare carelessness with the figures!

I see what you're saying about issues like this being perfect examples of why it's important not to get too bogged-down in the statistics. I drew similar conclusions from your other thread, where I believe one of the points you cited was the fact that statistics represent the performance only up until the present, and stats vary with each new trade placed.

The problem from my point of view is that I trade purely mechanical strategies. I have conflicting schools of thought which on the one hand, say that I should continue trading through periods of loss because I know the system has a positive expectancy overall, and on the other hand that these periods show that something is fundamentally wrong.

I suppose it is inevitable that some elements of discretion, even at a more generalised level, need to be excercised at some point when trading a mechanical system.