From basketball to random walk theory

I was watching the 76ers game the other day and heard something in the background commentary that caught my attention. Something along the lines of… the probability of an NBA player of making a shot does not depend on whether he made the shot before that. That claim basically said the “hot hand” was irrelevant when betting on whether a player would make their next shot. To me, that claim seemed like total crap. Apparently, upon researching this a little further after the game, I learned that there was a study done on the 76ers regarding this issue which concluded that there was no correlation between past shots’ influence on future shots.

I realized that this theory was also related to the random walk theory, originally proposed Burton Malkiel in 1973 to explain stock price fluctuation in financial markets. This really got me interested. I’m not going to go into much detail but basically what this theory suggested (based on a coin flipping experiment conducted by Malkiel) is that the fluctuation in stock prices are completely random due to the efficiency of the market.

Though some economists continue to believe in this theory to this day, I feel that there are three things fundamentally wrong with this claim:

1. Efficiency of market – there is a subtle claim in Malkiel’s theory which is that stock prices fluctuate randomly “due to the efficiency of the market.” But is the market completely efficient? I don’t believe so. You have government intervention bailing out the banks, time-lags, and significant human involvement (though many things are now becoming algorithmic driven), all three of which contribute to inefficiencies. Thus, Malkiel’s theory has some holes in its assumptions.
2. Market vs. Individual stocks – this is akin to the difference between the entire NBA vs. an individual player on the court. From what I understand, Malkiel’s theory applies to the NBA (the market), and so does the shot percentages research conducted with no correlation. But Statistics 101 says averaging data, and drawing more and more general correlations masks individualized data, which might give a whole new story. Perhaps random walk theory can suggest predicting market movement is impossible as it is random, but it cannot say predicting an individual stock’s movement is random. Similarly, the research on basketball shots combined data from the NBA and hence can apply to the NBA but not the Kobe Bryant’s or Michael Jordan’s of the world.
3. Present Day Wall Street – present day Wall Street makes the money on the individual stock, the individual player. And the trillions of dollars to be made in the financial industry clearly shed light on the fact that trades are not random. Otherwise, people would not be in this industry. There is clearly intelligence involved in using the large amounts of data, sorting out the relevant bits, and drawing conclusions – betting on IPOs, arbitrage trading, momentum trading are relevant examples. Though there is always an element of uncertainty and luck, it doesn’t govern the process but instead just plays a role.
   

The 76ers won that game 98-92.