|
PureBytes Links
Trading Reference Links
|
I was asked this privately. I think it was meant for the list; apologies
otherwise.
> Dennis, then why use the standard deviation of winning trades at all? Why
> not just use the losers in your sample, or a function of the losers, a la
> Sortino Ratio/MAR? Why is upside volatility being used to denote risk at
> all?
Why is upside volatility important, given equal returns at the end of
the year? Imagine you're going out looking for a real job and you think
you're worth about $100K a year. You find two companies that want to
hire you and they appear identical and equally desireable in every way
except for how they pay. Company A says the pay is $104K/year and you
get paid $2K every Friday, no exceptions. Company B says the pay is
$104K/year but they can't tell you what your pay schedule will be. They
will pay you at irregular intervals when the money comes in from jobs
completed. Me, I'd choose company A. A's high Sharpe ratio translates
into peace of mind.
> Are we saying that upside trades could have been equivalent downside trades,
> so we should incorporate both measures of volatility in our calculation of
> risk? That would mean that no active trade management was taking place and
> that the system passively allowed equal movement in winning and losing
> trades. Not a system that I personally would want to invest in...
Money management is great (and necessary) but it's after the fact. You
don't reduce your size until you've already had that big loser. Then
you're trading smaller, trying to earn it back. Better to have a system
that doesn't have any big losers.
--
Dennis
|