To be a good investor, there needs to be a strong sense of history. To understand financial panic and crashes, the past crises have to be studied. To understand why opportunities sometimes persist or disappear, there needs to be an appreciation of past behavior and market structure. Nevertheless, history is not linear. The same investment mistakes are made as past lesson are often never learned. There is not really an arc of progress that can be bent or even followed. Progress in finance and in particular valuation can lurch forward or it can fall back based on the latest behavior of the crowd.

There can be progress with technology and tools such as more data, faster computers, and new statistical techniques, but there is not a linear progress with our understanding of markets and human nature. History is not progressive with investors always growing smarter, and we cannot take our current view and impose it on the thinking of even a decade ago. The great historian Herbert Butterfield argued for this interpretation of history in the mid-20th century. He influenced the writings of Thomas Kuhn and his ideas concerning the evolution of science.

This ebb and flow of ideas and interpretation of facts applies to finance just as it does with science in general. Of course, change sometimes occurs slowly. Even within finance there can be a dominant idea for decades only to see it fail and replaced with a new model to describe markets when there is a market crisis. This Kuhnian worldview of paradigm shifts is helpful with understanding how to make money in a world where alpha is elusive.

This shifting foundation in knowledge is foundational to the concept of adaptive markets. In a linear progressive history, markets should only get more efficient as we uncover more knowledge. In reality, ideas and views are often proven to be false after hey cannot explain current behavior. Investors make mistakes with their interpretation of facts, and progress is both made and lost. There will always be new inefficiencies and market noise will always be in a state of flux. This is why research is necessary. New models will replace the systematic models of today tomorrow as we learn from our mistakes.