Traditional economic theory has a well defined position regarding the possible effectiveness of news to determine the future behaviour of equity markets.
One of the basic assumptions, that equity prices essentially perform a random walk around the prevailing trend, would appear to foreclose any likelihood of outperforming the market for any significant period of time. It is claimed that the sheer number and interconnectedness of the variables determining the outcome leads to an intricate interdependent system that can only effectively be modelled using the mathematical concept of a non-linear dynamic system. The essence of a non-linear dynamic system is that minute – practically undetectable – changes in entry conditions may well lead to very significant differences in the outcome. Thus, effectively preventing any meaningful predictions about the future outcome, even if all entry conditions were known.
This leads to a situation of uncertainty which is graphically illustrated by prof. John Allen Paulos (1). He state that all equity market assessments essentially state that “Things will continue roughly as they have been” with the added clause “Until something changes”.
On the other hand , it is intuitively clear that information about specific events can strongly indicate the future value of a company –if this weren’t the case, insider trading would not need to be illegal. Accepting the fact that the modern news industry effectively captures a wide range of socio-economical events on a day by day basis, these events are likely to be reported on in the deluge of news messages.
However, a second popular economic theory: The Efficient Market Hypotheses, states that any available information is already reflected in the equity prices, again, making it impossible to consistently outperform the market by using information known to the market.
The key to this situation may well be the clause “information known to the market”.
In our modern times, the various news agencies function to such elevated level of efficiency that a normal day results in over 4000 news messages. This number is the typical output of just one major news agency and in the English language only. The fact is that many possible relevant news articles are published at a rate that is humanly impossible to digest. This information overload situation would seem to provide an opening in the theoretical obstacles expressed by the random walk and efficient market hypotheses. Due to the sheer number of news, it would seem unlikely that at any given point in time all market traders would be equally informed about all events affecting the value of a specific stock. Relevant news can be lost or simply overlooked in the magnitude of messages.
This is precisely the situation that automatic news analysis is designed to exploit. By using high speed text processing software , incoming news messages are automatically read and intelligent semantic analysis detects the mentioning of specific events and the way they would most logically affect the value of any relevant equity. These systems alert traders about events they should be aware of, or send their findings directly to specially designed algorithmic trading systems that automatically take the news events into account when executing their orders.
Going back to Professor Paulos quote, it would therefore be very likely that the elusive “something” that causes the change, is hidden in the daily stream of news messages, and with a correctly designed news analysis system, it should – in theory – be possible to automatically detect it and execute on it.
Reference
1. Paulos, John Allen. Recession forecast if steps not taken. A Mathematician reads the newspaper. 1995.