Quote from mind:
i am starting some analysis on hang seng futures on a daily basis. data source is csi. i look at the the time from 1.4.1993 until 20.4.2004. i have no experience in this market, therefore i would like to hear about peoples' thoughts.
some market statistics
dailyGain mean: +0,01%
dailGain median: +0.03%
dailyRange mean: 2.09%
dailyrange median: 1.76%
an interesting finding:
if the market closes in the upper 10% of the bar of the day and is below a slow exponential moving average, the figures for the next day are:
dailyGain mean: -0,72%
dailGain median: -0.54%
if you just sold the next open and bought it back on close you would end up with 110 trades, a hitratio of 70%, a payOffRatio of 1.50, an annual return of 14.9% (given a leverage of 2), an annualised volatility of 10.8%, and a modified sharpe ratio of 1.32.
yearly results
1994 40%
1995 21%
1996 00%
1997 13%
1998 24%
1999 05%
2000 05%
2001 22%
2002 12%
2003 02%
2004 00%
these numbers are basically good. given that you can trade it, it would be a fine strategy. i used 0.02% as a rough estimation for roundturn (when we really start trading something, we do these things in more accuracy. here it is just an indication thing). but there are assumptions inherent that can easily jeopardise it: i assume i can trade the open and the close, which is not necessary true. i assume the data is reliable. and so forth.
maybe this post can induce some discussion.
