Blaming Paul Samuelson and a Few Others...

Yes, I blame Paul Samuelson for changing the face of finance and economics. I use the word, blame,  tongue in cheek, but Nobel laureate Samuelson was the prime driver for what we are witnessing today in the world of capital markets. Samuelson, I will contend, started a process almost seventy years ago that is ushering in the greatest wave of automation we will ever see in Finance and Economics. 

A bit of history...

The Second World War brought together some of the greatest minds of the twenty century in a struggle to free Europe and Asia. It also witnessed some of the largest leaps in scientific discovery and tools creation. Computers (really calculating machines), operations research and the process of taking theoretical physics into the real world with applied techniques, the Atom Bomb, would forever mark the next generation of thinkers.  Out of this time period, Samuelson would write a landmark text, Foundations of Economic Analysis in 1947.  In a review from 1950 in the American Mathematical Society by Merrill Flood, here is the key change in economic thought that Samuelson embodies.

The exposition is thoroughly mathematical. It is likely that most economists will find it very hard reading even though, as is noted in the Introduction: ".. . The pure mathematician will recognize all too readily the essentially elementary character of the tools used." Nevertheless, the book will undoubtedly lead to a greatly increased use of mathematics by economists in their future work. It sets a new high standard for the mathematical development of unified economic theory.

For Samuelson the fusing of Economics and Mathematics meant meaningful theories with testable hypothesis. Economic analysis would change exactly how economists did their jobs. Rightly or wrongly it would challenge economists to be more like the hard sciences in thought and practice. This change in economics is similar to the change physics underwent with the introduction of Statistical Mechanics by Josiah Gibbs. I mark 1947 as seminal year. It is also around this time that the term "automation' starts entering popular culture, thanks to the modernization of Ford Motor Company's manufacturing plants.

"Give us some more of that automatic business," a Ford vice president reportedly said in a meeting. "Some more of that--that--'automation.' "

(For more on Automation you can read Nicholas Carr's book, "The Glass Cage", where the quote comes from)

Samuelson would also influence and spark the efficient markets hypothesis (EMH) by rediscovering a paper from the 1900's by Louis Bachelier on stock market pricing. Eugene Fama would later take this thread and fully bake out the idea that market prices reflect all information and it's "impossible to beat the market" (you may argue at length with this as most people do) 

Enter mainframe computers and software in the 1960's and you can see how it all starts to come together. Modeling economic ideas on asset prices and using it to predict future movements or the creation of efficient portfolios based on risk and return utility functions, became the discussion in finance over the next few decades. But mainframes were too limited and constrained by the priesthood of job running to be diffusive enough in the industry. Those with emerging knowledge could not get enough access to computing time and processing power to test ideas. 

Fast forward to the microcomputer and the mid 1980's along with network connectivity to market centers and we see a sea of change. I was at Morgan Stanley in the midst of that wave and watched as physicists entered Wall Street and the era of UNIX workstations exploded in financial firms. Computational finance for applied uses in trading began in the mid 80's.  It picked up steam as Sun Microsystems sold sparc workstations by the truck load to every investment bank with a proprietary trading strategy and a dream(s). Numerical data was the coin of the realm along with fast processing and creative analytics.  We've reached full diffusion within the industry on most of these ideas and it was a generational shift in how finance matured.

What does this all have to do with today?

The next wave is coming. It's being ushered in by two forces. The use of both structured/unstructured data and machine learning techniques.  Unstructured data usually implies text and meta-data that is mostly qualitative. Many parts of finance, which influence information the market "knows", are reports, opinions or news. The first wave of machine learning is looking to help categorize that unstructured data for information processing. What in essence use to be human endeavor to synthesize information is being turned into a computer automation process of learning. It is the next generational shift in finance and will have major impacts across all of fintech. But it won't be the only one....