In my last article entitled, "Third Wave Investment Management" I introduced the term "Financial Cybernetics" to describe the process of enhancing financial decision making by introducing AI technologies. The term "cybernetics" was originally coined in 1948 by Norbert Wiener to depict the science of modifying or enhancing human systems with artificial electronic systems. In the financial field we might want to think in terms of enhancing the human decision maker with intelligent computer systems. To understand and appreciate the importance of financial cybernetics, you must first understand the importance of money and money management in the free enterprise system.
A basic tenant of the free enterprise system is that innovative ideas and activities (entrepreneuralship) should be rewarded. The proper allocation of these rewards is an essential element in stimulating development of technology and future economic growth. Currently this allocation task is accomplished by financial analysts that spend their days pouring over company annual reports, financials, and SEC filings (10K's, 10Q's, 13D's, etc.). The process is very labor intensive and still uses techniques developed by Charles H. Dow, Samuel A. Nelson, William P. Hamilton, Benjamin Graham and others in the late 1800's and early 1900's. More recent contributions to the process have suggested that the market place is the only efficient allocator of these needed rewards and that analyst add little to the process. Despite this academic dogma most brokerage firms and large money management organizations maintain large research departments to evaluate corporate performance and make recommendations regarding the allocation of financial rewards. The current system is very inefficient and incorporates little of the computer decision theory and support technology that has been developed by the scientific and engineering community.
Just like the generals of today are accused of using the military strategies of wars past, so too are the financial analyst of today using techniques of yesteryear. Most analyst assume the world they know will last indefinitely. They find it difficult to imagine a truly different way of analysis for themselves, let alone one for the entire industry. The Financial Analysis Journal which is suppose to be on the cutting edge of new analytical techniques for the industry has published about three or four articles over the past five years on AI topics. For a financial analyst to find information on topics such as Expert Systems, Fuzzy Logic, Neural Networks, and Genetic Algorithms they must go to the scientific and engineering literature or now AI in Finance, or attend meeting such as the recent IEEE World Congress on Computational Intelligence or the World Congress On Neural Networks. In partial defense of AIMR, (Association for Investment Management and Research) they did recently hold a very good meeting in Boston entitled "Blending Quantitative and Traditional Equity Analysis". Many of the presentations delt with advance nonlinear analysis techniques. A meeting that was attended by two or three times as many people as they expected. This should be an indication of the level of interest in the financial community.
In most fields of endeavor, as experience and knowledge increases so does productivity and performance. This rule or natural law does not seem to hold true in the area of financial management. The excess returns that money managers have been able to generate for their clients has not changed in the last 100 years. Either we are dealing with a totally random process that is not to be improved upon or the "science" of money management is not advancing.
There are a few money managers that have consistently been able to outperform the general market and generate excess returns for their clients, but they have not been successful in communicating or educating the financial community on their techniques. Of course, this may have something to do with competitive advantage.
The use of computers in financial analysis is in its infancy. We are still thinking of P/L statements and balance sheet figures entered into a Lotus or Excel spread sheet as financial analysis. Analyst's use of computers today is like owning a Ferrari and driving it on the San Diego Freeway in down town Los Angeles at rush hour, You rarely get out of first or second gear. The problem is financial analysis is not the totaling of the columns in your spread sheet or calculating the financial ratios, it is identifying meaningful events and seeking out significant relationships. Some of these relationships are so obscure that they are beyond human comprehension and require the assistance of a computer assistant with high level computational intelligence.
Humans are limited in their ability to comprehend complex events. We deal with one, two, and maybe even three dimensional linear space relatively well. But, if we venture much beyond that into K dimensional space we rapidly become overwhelmed. Although we like to talk about investing in low P/E stocks, sophisticated investors actually adjust their concept of low P/E for various industries, for different EPS growth rates, and for various rates of inflation. Attempting to take these adjustments much beyond two or three dimensions simultaneously, however, taxes the human analytical capability to its limits. This limitation is not true, however, for some of the new advanced AI systems that are being developed and used. These systems operate with ease in financial hyperspace far more complex than anything every attempted by human analysts.
If you were impressed with the sight of Cruise missiles flying down the main street of Bagdad seeking out and destroying specific targets, or the fireworks in the sky as Patriot missiles intercepted and destroyed incoming Scud missiles, than you are going to be awestruck by the future uses of smart computer systems. The same type of computer Artificial Intelligence that guided military weapons in the Persian Gulf are now going to control and guide your investment portfolios to new heights. The potential strength of these systems is so great that their future use could drastically change the power structure of world financial markets.
Now there are many that will read this and say that this is nothing more than hyperbola that will cast dark shadows over the field just as it did in 1956 when the Perceptron was presented at the Dartmouth AI conference. I would feel the same way myself if I had not seen first hand the results that can be obtained with these advanced hybrid AI systems. Watch the returns and the asset gathering power of those companies such as Fidelity, LBS Capital Management, Union Capital Advisers, Deere & Company and RMC Capital Management that are using these systems to manage assets. Money tends to flow fairly rapidly to those organizations that provide the best returns.
Investing is basically a resource allocation problem involving decision making with regard to how to allocate limited dollar resources among the multitude of investment options. These are not trivial decisions. The future success and prosperity of our society depends on the decisions that are made. And these decisions have to be made on something other than the discounted present value of future dividends. There is more to the problem than a simple discounting equation. John Maynard Keynes in his book, The General Theory of Employment, Interest and Money suggested that the investment decision is like the "...newspaper competitions in which the competitors have to pick out the six prettiest faces from a hundred photographs, the prize being awarded to the competitor whose choice most nearly corresponds to the average preferences of the competitors as a whole; so that each competitor has to pick, not those which he himself finds prettiest, but those which he thinks likeliest to catch the fancy of the other competitors, all of whom are looking at the problem from the same point of view."
This is not to suggest that investing is just a beauty contest, but it does point up the fact that this is a very dynamic nonlinear problem that requires special analytical tools. Text books on the subject of investing state that there are two approaches to the problem; technical analysis and fundamental analysis. This is a wild over simplification of the problem solution. In actuality one should be using any and all information that reliably helps with the problem solution and it is more a matter of time frame than any arbitrary classification system. If your time horizon is two or three years you better be looking at "fundamental" financial and performance data. If your time frame is next week or even the next three months you better pay attention to the supply and demand dynamics captured in some "technical measures".
Now that we have expanded the problem domain to include all financial measures how do we deal with a problem of such immense proportion. The answer, of course, lies in the topic of this article, Financial Cybernetics. We must enhance the ability of our decision makers by expanding their capabilities with computerized intelligence.
It is quite apparent that we can not take the Graham/Dodd financial ratios, plug them into a linear equation and hope to solve the investment asset allocation problem. It is equally apparent that the answer to the problem does not lie in chaos theory. As the members of the Santa Fe Institute like to point out, there is something between order and chaos and that something is "complexity." To work in this area of transition one needs the special tools encompassed in the broad area referred to as Artificial or Computational Intelligence.
Most, if not all, of the complex evaluation problems facing the financial community involve dynamic, multiple, interacting, nonlinear, fuzzy variables. To understand the problem we need to address each of these attributes.
Dynamic Variables - The financial markets are constantly changing and the rules that worked last year will not necessarily work this year. Financial analyst have found and reported that computer models built prior to 1987 are just not producing the same results in the post crash era.
Multiple Variables - Few if any financial problems can be solved by only looking at a single variable. The complexity of the problem dictates that multiple variables be addressed to even approximate a solution.
Interacting Variables - Variables do not act in isolation; as they change they have an impact on everything around them. This interrelationship among variables creates a degree of complexity that requires the use sophisticated analysis techniques capable of dealing with the totality of the problem.
Nonlinear Variables - When a solution cannot be expressed as a simple multiple of the evaluation parameters, it is said to be nonlinear or nonalgorithmic. This describes the situation in many financial problems. You cannot take several market variables, plug them into an linear equation, and solve for market direction over the next six months with any consistent degree of accuracy.
Fuzzy Variables - Much of the information processed by the financial community is fuzzy or ambiguous in nature. People tend to think and speak in a fuzzy way. They talk about fast sports cars, loud music, tall men, happy children, accurate watches, and hard work, and with excellent understanding. Most things are not black and white, but are shades of gray and our language reflects this. Fuzzy systems are often useful in handing language inputs for analytical systems - including financial systems.
By combining the concepts of dynamic, multiple, interacting, nonlinear, and fuzzy variables, you start to gain an appreciation for the complexity of the financial analysis problem and the need for Financial Cybernetics.
About the Author: Commander Berghage is the President and Chief Executive Officer of NeuWorld Financial, a San Diego based research and development corporation developing and licensing AI technology to financial institutions. While in the service Commander Berghage was a research psychologist working on human factor problems in advance systems. Since leaving the service in 1983 he has been the Director of Research for two national brokerage firms and a San Diego based money management firm. He is a past president of the Financial Analysis Society of San Diego.