Page 201 - DCOM507_STOCK_MARKET_OPERATIONS
P. 201

Stock Market Operations




                    Notes              chaos theory may eventually provide some potential answers. If the apparent randomness
                                       of security price changes, can be shown to be non-random, much of the theory of finance
                                       would need revision.

                                   Self Assessment

                                   Fill in the blanks:
                                   6.  ………………..................... numbers have intrigued mathematicians  and scientists  for
                                       hundred of years.

                                   7.  Nikolay Kondratev was a ………………..................... economist and statistician born in 1892.

                                   10.4 Neutral Networks

                                   A neutral network is a trading system in which a forecasting model is trained to find desired
                                   output from past trading data. By repeatedly cycling through the data, the neutral network
                                   eventually learns the pattern that produces the desired output. If the desired output remains
                                   elusive, more data is included until a pattern is found. Neutral networks may also include a
                                   feedback mechanism whereby experience gained from past errors.
                                      This topic is a hot one in  the investment community. National  conferences have  been
                                       organized dealing exclusively  with this  topic, and the trade literature publishes many
                                       articles upon this. A problem with concept of a neutral network is that the stock market is
                                       seldom deterministic. Situations constantly change, and what may have been true a few
                                       years ago will not necessarily prevail tomorrow. Financial academics are especially leery
                                       of back-tests, or research that tests a hypotheses using past data. Mining the data  will
                                       almost always result in some apparent cause and effect between past events and stock
                                       market performance. Research that tests a hypothesis using subsequent data is much more
                                       useful. An article in the popular press describes Wall Street’s response to this criticism.
                                      One way to get around this hazard is to build something called a genetic algorithm into
                                       your neutral network. A sexy term that currently causes Wall Street rocket scientists to
                                       swoon, genetic algorithms enable neutral nets to adapt to the future buy spawning schools
                                       of baby nets, each of which is sent to swim against the changing flow of data, where only
                                       the fittest survive to take over the role of the mother.

                                      No matter what someone’s field of study, they are interested in the search for a better
                                       mousetrap. Essentially, what all security analysts seek to do is to find improvements in
                                       their methodology for security selection.

                                   10.5 Tools of Technical Analysis

                                   The technician must (1) identify the trend, (2) recognize when one trend comes to an end and
                                   prices set off in the opposite direction. His central problem is to distinguish between reversals
                                   within a trend and real changes in the trend itself. This problem of sorting out price changes is
                                   critical, since prices do not change in a smooth, uninterrupted fashion.
                                   The two variables concerning groups of stocks or individual stocks are:
                                   1.  Behaviour of prices, and
                                   2.  Volume of trading contributing to and influenced by changing prices.

                                   The use of technical ‘indicators’ to measure the direction of overall market should precede any
                                   technical analysis of individual stocks, because of systematic influence of the general market on
                                   stock prices. In addition, some technicians feel that forecasting aggregates a more reliable, since
                                   individual errors can be filtered out.


          196                               LOVELY PROFESSIONAL UNIVERSITY
   196   197   198   199   200   201   202   203   204   205   206