|
PureBytes Links
Trading Reference Links
|
> -----Original Message-----
> From: Justin Fanning
> Sent: Monday, January 05, 2004 10:20 AM
>
> As you may have read from my introductory post(s),
> I work with one of the exchanges, and as such have
> access to the clearing house figures.
>
> In my view, most day-traders, as a business,
> are long term unsuccessful. This is mostly true
> for floor traders also, to hear them talk you'd
> think they are all successful, but again the
> clearing house figures do not back this up.
>
> The successful traders, again in my view, are the
> dedicated system traders who hold mid - long term
> positions (days - weeks).
>
Is there an analytical justificaion for a particular timeframe being more
profitable versus another? For example, consider the following experiment:
Using the last 20 years of S&P data (a continuous contract would be most
realistic) collect daily, weekly, and monthly quotes. Then, randomly select
100 samples from each set and caculate:
1) the percentage of times that this sample closed in the same direction
as the previous sample's change. (a rough measure of the effectiveness of
the effectiveness of a trend-following system).
2) Assuming a trading system that always trades in the direction of the
previous sample's change, calculate the ratio of the average win to average
loss,
probability of win, max loss expressed as a perentage, and so on.
Now, maybe trading with the trend of the previous sample's change isn't the
best system in the world, but it is certainly simple, and easy to apply
across
all timeframes. To make things a bit more realistic, it'd be good to add
some
fudge for commission and slippage, because certainly for shorter timeframes
this is a bigger factor in overall profitability.
|