One of the most famous quotes to come out of the Vietnam war was McNamara’s classic observation about “the fog of war” (while he used it extensively, it originated back with Von Clauswitz). The basics behind this observation were that there are so many variables that it is impossible to clearly ascertain at any given point in time exactly what is occurring throughout a battlefield. Similarly the same “fog” applies to the investment process. With 6 billion inhabitants of the planet, each making individual daily economic decisions, influenced by who knows how many local events, it is even more impossible to determine, even with Nvidia’s most powerful chips just what the future holds for individual stocks and markets.
One of the occurrences I’ve used in the past to illustrate this point was “the Arab Spring” (boy, if ever there was a misnomer!). Here you had a small merchant who became so angry at a local bureaucratic that he immolated himself in protest. This one small, local incident led to local riots which led to contagion in other countries in the region and eventually the downfall of governments far from the simple local incident. Now multiply these daily local (economic) occurrences on a global scale and you can see how difficult it is to accurately and consistently predict stock and market prices.
Going back about 20 years since computers have been able to beat world champion humans at chess. The machines use a brute force strategy where they analyze millions of possible outcomes for each move on the board. Even the cleverest person can't keep track of that many options, let alone run through all those scenarios in the time allotted for a match.
And those chess programs haven't been using Artificial Intelligence. The current versions of AI now offer even more advanced analytical capabilities, which should make them even better suited to predict future scenarios. The operative word here being “should,” not “will!”
As could be expected, investors are now trying to use AI to to help them make market-beating investments. Many believe they can just ask ChatGPT which stocks are about to go up, and then buy them ahead of time. As noted above, the stock market is simply too complex to predict, even for AI. Thus, “the fog of investing!”
As discussed, at any given moment, millions of investors are making trades based on an incalculable number of factors. On top of this are the hard-to-measure motivations that drive economic decision making. In all, you have vastly more inputs than the latest supercomputer could possibly analyze.
David Booth (the founder of the prime investment company we use), writing in the Financial Times, observes that that the information available to AI is a subset of all available information. In other words, AI models have less to go on than human analysts. Therefore, it's not reasonable to expect an AI bot to attain greater insight.
One of the problems with trying to predict the market using the brute-force method that works for chess is that there's no way to predict what level of influence any given piece of data will have. Booth says, "The market is fantastically complex. So much so that no one knows how much a particular piece of information impacts a price, because there are so many other simultaneous inputs."
He also points out that even if someone did have an AI capable of predicting the market, they would not make it available for public use. In other words, if I had the simple sauce, why would I share this information for a couple of bucks when I could make billions or more keeping the information to myself?
But just because AI can't function as a crystal ball for the market doesn't mean it's not useful for aspects of financial planning—things like analyzing portfolios and exploring possible retirement scenarios. This of course, is where an “investor coach” comes in. AI doesn’t and can’t replace the human mind and it’s experience or more importantly wisdom. It can’t sit across from an individual and determine what makes that individual tick, what is cared about, how one feels about the uses the investment is intended for – all those human characteristics that go into determining just how and why portfolios are to be constructed.
Some may recall a few years ago I queried AI for investment advice for a couple at retirement age. I gave it all the requisite information and the response that came back was for them so speak with a qualified advisor. I breathed a sigh of relief now having the knowledge that I had not been obsoleted – yet!
An old friend, the late psychology Professor Edwin Timmons (voted the top professor at LSU multiple times) once explained why it is so difficult for even humans to understand each other and their motivations (much less in this case, AI). The human brain has 9 billion synapses. The possible variables in just two human interactions is 9 billion factorum. That’s 9 billion times 8.9… times 8.9… and so on. Is it any wonder?