I have covered this method in my previous posting. I am repeating it here for readers that have missed out on this simple indicator that can be used for trading index ETF like STI ETF.
If today’s close is greater than close x days ago, score = 1
If today’s close is less than the close x+1 days ago, score = -1
Tuschar Chande suggested comparing the close to the closes from 11 to 20 days ago.
Therefore,
Trendscore = 10 day sum of the scores from days 11 to 20.
The trend score for this case will oscillate between +10 to -10.
The metastock formula is given below:
If(C>=Ref(C,-11),1,-1)+
If(C>=Ref(C,-12),1,-1)+
If(C>=Ref(C,-13),1,-1)+
If(C>=Ref(C,-14),1,-1)+
If(C>=Ref(C,-15),1,-1)+
If(C>=Ref(C,-16),1,-1)+
If(C>=Ref(C,-17),1,-1)+
If(C>=Ref(C,-18),1,-1)+
If(C>=Ref(C,-19),1,-1)+
If(C>=Ref(C,-20),1,-1)
Let’s say that we buy 1 lot of Straits Times Index (STI) when the Trendscore reaches +10 and exit the trade when Trendscore becomes negative.
The trade record is given below:
| Position | Entry Date | Entry Price | Exit Date | Exit Price | % Change |
|---|---|---|---|---|---|
| Long | 12-Jan-07 | 2995.35 | 1-Mar-07 | 3133.88 | 4.62 |
| Long | 15-Feb-07 | 3222.98 | 1-Mar-07 | 3133.88 | -2.76 |
| Long | 28-Mar-07 | 3224.55 | 11-Jun-07 | 3530.38 | 9.48 |
| Long | 4-May-07 | 3477.9 | 11-Jun-07 | 3530.38 | 1.51 |
| Long | 13-Jun-07 | 3530.12 | 27-Jun-07 | 3513.5 | -0.47 |
| Long | 15-Jun-07 | 3601.05 | 27-Jun-07 | 3513.5 | -2.43 |
| Long | 16-Jul-07 | 3679.97 | 30-Jul-07 | 3467.36 | -5.78 |
| Long | 23-Jul-07 | 3627.97 | 30-Jul-07 | 3467.36 | -4.43 |
| Long | 6-Sep-07 | 3374.01 | 23-Oct-07 | 3670.21 | 8.78 |
| Long | 11-Dec-07 | 3583.32 | 18-Dec-07 | 3301.42 | -7.87 |
| Long | 2-Apr-08 | 3046.54 | 23-May-08 | 3160.86 | 3.75 |
| Long | 23-Apr-08 | 3187.23 | 23-May-08 | 3160.86 | -0.83 |
| Long | 28-Apr-08 | 3189.2 | 23-May-08 | 3160.86 | -0.89 |
| Long | 5-May-08 | 3236.1 | 23-May-08 | 3160.86 | -2.33 |
| Long | 14-May-08 | 3203.42 | 23-May-08 | 3160.86 | -1.33 |
| Long | 16-May-08 | 3207.43 | 23-May-08 | 3160.86 | -1.45 |
| Long | 11-Dec-08 | 1821.7 | 13-Jan-09 | 1776.25 | -2.49 |
| Long | 17-Dec-08 | 1782.09 | 13-Jan-09 | 1776.25 | -0.33 |
| Long | 26-Dec-08 | 1736.99 | 13-Jan-09 | 1776.25 | 2.26 |
| Long | 5-Jan-09 | 1829.71 | 13-Jan-09 | 1776.25 | -2.92 |
| Long | 24-Mar-09 | 1664.08 | 19-Jun-09 | 2237.2 | 34.44 |
| Long | 24-Apr-09 | 1859.98 | 19-Jun-09 | 2237.2 | 20.28 |
| Long | 1-May-09 | 1920.28 | 19-Jun-09 | 2237.2 | 16.5 |
| Long | 28-May-09 | 2306.08 | 19-Jun-09 | 2237.2 | -2.99 |
| Long | 16-Jul-09 | 2389.42 | 19-Aug-09 | 2567.72 | 7.46 |
| Long | 9-Sep-09 | 2660.91 | 5-Oct-09 | 2604.53 | -2.12 |
| Long | 17-Sep-09 | 2674.42 | 5-Oct-09 | 2604.53 | -2.61 |
| Long | 15-Oct-09 | 2708.48 | 29-Oct-09 | 2648.98 | -2.2 |
| Long | 12-Nov-09 | 2740.43 | 22-Jan-10 | 2850.98 | 4.03 |
| Long | 3-Dec-09 | 2796.34 | 22-Jan-10 | 2850.98 | 1.95 |
| Long | 14-Dec-09 | 2800.75 | 22-Jan-10 | 2850.98 | 1.79 |
| Long | 23-Dec-09 | 2823.82 | 22-Jan-10 | 2850.98 | 0.96 |
| Long | 24-Feb-10 | 2782.55 | 29-Apr-10 | 2932.04 | 5.37 |
| Long | 2-Mar-10 | 2774.06 | 29-Apr-10 | 2932.04 | 5.69 |
| Long | 9-Mar-10 | 2834.57 | 29-Apr-10 | 2932.04 | 3.44 |
| Long | 5-Apr-10 | 2943.02 | 29-Apr-10 | 2932.04 | -0.37 |
| Long | 26-Apr-10 | 2988.49 | 29-Apr-10 | 2932.04 | -1.89 |
| Long | 17-Jun-10 | 2846.94 | 13-Aug-10 | 2927.04 | 2.81 |
| Long | 9-Jul-10 | 2897.15 | 13-Aug-10 | 2927.04 | 1.03 |
| Long | 6-Sep-10 | 3011.03 | 28-Oct-10 | 3124.38 | 3.76 |
| Long | 3-Nov-10 | 3205.28 | 24-Nov-10 | 3126.3 | -2.46 |
The net gain from 2007 to 2010 is around 88% (excluding transaction costs).
2 comments:
If you try it for the HSI which is a much more trade-able index because of highly liquid mini-futures and CFD linked to it. Wonder what results you get?
Based on the results fo Garth based models, I expect the trading approach based on Chande trend indicator to perform better than when used on STI.
Paper on Garth approach for HSI:
http://www.springerlink.com/content
/678178735774h313/
Just to add. From the Garth model the non-normal behavior of the HSI shows that the intermediate trends (intra-year) to be larger than expected from the normal distribution.
So I would expect the Chade indicator to generate more profitable trades for HSI than the STI.
Post a Comment