Why Consistency is a Sign of The Wrong Strategy

[Antanykki]

I always see lots of discussion on this forum about consistency - whether it be consistent systems, trading methodologies, etc
the simple facts [ not an opinion] are that if you look at the most successful traders and investment managers in history [ i.e. a track record of at least 20 years compounding somewhere around 20% per anum - incidentally only 4 exist above 20%] then the one feature they all have in common is consistency is not achievable for long term success. In fact one could argue the more consistent a strategy is over the short term the less likely it is to work in the long term. The simple reason is that its likely to be optimised and curve fit.
Robustness is far more important than consistent and the 2 are not related

[AnthonyMl]

Maybe you can explain what makes something robust as opposed to consistent and the constituent parts of each.

[Anthonytype]

Robustness is defined as something that will survive when you make underlying errors in its design i.e. even with changed parameters it still works. Also if you make errors in your judgement about the underlying data used in that system - the system still works
so for example over 30 years a robust system maybe a simple break out system of 200 days - it doesn't really matter if its 100 or 300 it still works. Similarly if markets don't trend for a period of time it doesn't blow up either.

A highly consistent system maybe for example all the systems we saw between 2004-2008 that sold volatility - day after day you could print money and erroneously assume the holy grail had been found - this was a consistent way to make money for years - then one day - they all blew up as the underlying market circumstances changed - perhaps anyone who believes in consistency should also have a look at the story of long term capital management ?
so comparing that to a relevant example today - the whole world believes stocks could not possibly collapse as the central banks of the world will always support them / print money - a great consistent system that has worked well for 5 years - but a system almost certain to blow up at some point in the not too distant future ?

[AntonyDowl]

What makes trading so hard is that the vast majority [ 99%+} are lured into trading by the promise of consistent profits through unproven nonsensical stuff like technical analysis [ most indicators] or even worse fundamental analysis - people often then suffer from recency bias [ the highly prominent human condition of thinking that just because something has been "the way it is" for your lifetime / the last few years - that's the way it will always be] a very dangerous condition

The best evidence really is the lack of audited success stories of consistent profits - they simply do not exist in the long run - we all know a mate / trader who for 2 whole years has made x% - but does anyone ever meet one who has done it for 20 years ? Probably the most consistent long term proft maker was Monroe Trout - so one then has to ask what was his edge ?

[admin]

Well I agree that a trend trading approach will work when markets are trending and that a volatility expansion strategy will work when markets are volatile. However, what happens to a trend trading strategy when markets stay volatile for a very long time ? They lose money and a lot of it. I think the problem is that there is no "one size fits all" approach and the key is knowing what phase a market is in and I doubt most long term funds have the ability to switch as they tend to stick with one approach.

What would you suggest as a way forward ?