Jun

19

The earliest point and figure system I have come across is in the Ticker Magazine of August 1908, in an article entitled The Trader's Convention, by Mr. E.C. Cummings of North Wales, PA. Mr. Cummings noted that stocks often move 10 points in one direction, and gives a workout of American Sugar to prove that with proper stops, even during the 50% decline for the market of 1907, a point and figure system would have made money on the long side. He lists the following rules for his system:

  1. Consider only movements and reactions of one point or more.
  2. After a stock has moved in the desired direction by a point or more, wait for the reaction to that particular movement. After the reaction has terminated, and a second movement has begun one point from the ending of the reaction, buy the stock.
  3. Protect each trade with a two point stop order.
  4. After the second movement of a point or more has reacted, and the third movement has begun, trade again, providing the second reaction did not reach as low as the first reaction.
  5. After two points of profits, move the stop to the purchase price.

Rules for entering shorts are symmetrical to those above.

The system seems much more fairly tested and defined than anything that I have read in recent Point and Figure books, and there is no mumbo jumbo connected with it, what with the system being a clear incarnation of supply and demand.

The Artful Simulator, Mr. Tom Downing, and I tested a variant of the Cummings system in his honor. We bought the Dow whenever there was a rise of one percent from the lowest close, and sold when there was a decline of one percent from the highest close, looking at data from year end 1899 to June 15th, 1907. Over the full period, the average profits were as follows, (in percent):

                                     Long           Short
Mean profit per trade       0.40%         -0.09%
Number of observations   2229            2229
Standard Deviation          2.85%          2.7%
T Score                           6.7              1.6
Percent Profit                   44               38
Duration in Days               7                 5

This makes a most magnificent instantiation for a very skeptical speculator. Of course, that's the good news: that the shorts were almost able to overcome the drift of 0.25% a trade, not counting dividends et. al. was also amazing, along with the fact that such a system actually performed almost as well from 1920-1940, (total of 0.2.% profit per trade on the long side, and just a -0.01% loss on the short side, with 469 trades during that period). From 1920-1940 the trade duration was slightly lower than average, at five days per trade, showing that prices were more variable in that second fifth of the century than they are today.

All in all, this is a magnificent confirmation of the value of point and figure methods for predicting the Dow, but, regretfully, during the last ten years … well that's an exercise for the reader. 


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search