Note that the 223% peak shrinks to 173%. Secondly, the system returns have more stability (reliability). Thus is due to the number of holdings in the portfolio.
NUMBER OF TOP & BOTTOM RANKED HOLDINGS
vs GAINS AND DRAWDOWNS
September 5, 2008--May 5, 2009
24 weeks out of 40, the trend is down
|
buy |
sshrt |
%gains |
|
% |
% |
peak |
|
top n |
btm n |
buy |
sshrt |
both |
annlzd |
dd |
% |
|
2 |
7 |
15 |
12 |
30 |
40 |
-49 |
223 |
|
12 |
12 |
6 |
19 |
25 |
34 |
-41 |
192 |
|
18 |
18 |
7 |
19 |
27 |
37 |
-34 |
173 |
The third table row summarizes the results of the chart above. The number of stocks held is symmetrical, i.e., balanced, identical. The portfolio comprises 36 stocks. One half that number is 18. The system is continuously invested, long the top-ranked 18 stocks or sold short the bottom-ranked 18.
The portfolio starts out as a hedge fund, sorting all candidate holdings into the strongest and weakest stocks. But instead of taking all positions, half-long and half-short simultaneously, it separates the groups into alternative, timed holding periods. Voilą, a timed hedge fund.
The benefits are evident. The worst %drawdown is diminished by 30% from the %dd of the first row, (top-2, bottom-7 rankings) . The peak is lower but remains astonishing--and is no less tractable to our newly discovered rule of ignoring the rules when you arrive at a profit level over 60%, take it and wait for th next buy signal.
Surprisingly, the third row returns show even better than those of second row. Row 2 shows results using the top and bottom thirds of the portfolio, 12 stocks each, alternatively long and short.
Timing. Timing is exquisitely important. Ranking is easy, but it will never give you absolute returns. Since the market itself is a time-variant phenomenon, to obtain positive returns a trading system must time and counter time its profile relative to the market.
This system depicted in the chart above extracts intrinsic timing information from its own holdings in aggregate. It is, if you will, a micro-image of the entire global stock market--as is every system on this website. Site systems do not use the total market for timing. Each system uses its own portfolio.
Caution. The limited number of observations severely constricts the attractive conclusions this study may suggest. Further historical testing is necessary to achieve actionable validation for use. Nevertheless, the study illustrates that more positions held, rather than fewer, give better results. And it clearly reveals the usefulness of setting a fixed-target profit for every set of trades any manager, myself included, may enter into.
Posted
6/21/2009 5:08 p.m. EDT
|
Back-Archives
| Subscribe
|
