How humans use cognitive salience to guide mutually recursive decision making

Try the following experiment:

In a classroom, give every student a slip of paper. The challenge is that they must devise a way to all meet each other at some location in France at some time during some arbitrary day in the future, say October 14 2012--only they are not allowed to communicate with each other. Each student must write down what is essentially a guess on the slip of paper, completely incommunicado.

As you can probably guess, the students are for the most part successful: most will write down "Eiffel Tower, noon" on their slip of paper.

What happens, of course, is that each student goes through a process of mutually recursive decision making: the student must guess what the others will guess, knowing that those students' guesses are dependent on his own, and so on ad infinitum. Realizing that the guess will be hopelessly arbitrary, the student supposes that, all things being equal, he would meet the others at the most canonical place and the most canonical time of day. But, knowing that the others will also gravitate to this same canonical time and place, the student can now know with some certainty that they will choose the same time and location. And so it is the very rational arbitrariness of the decision that reveals the particular salience of one time and place (in this case, Eiffel Tower at noon). This salience is then used as an empirical datapoint to rationally arrive at a now-suddenly-non-arbitrary choice, "Eiffel tower at noon".

My thinking is that this process is precisely the same one that governs markets. Stock price, for example, though naively a reflection of the "value of a company", is strictly speaking a reflection of other investors' demand for the stock. But since all investors are using the same decision function, and the function takes as its inputs the decisions of all other investors, rational decision making stalls on an infinite regress, rendering the decision arbitrary. But the arbitrariness of the decision, now placing all choices on an equal footing, reveals that some choices are more salient than others--in particular, the ones that correspond to the "naive" understanding of stock price as a reflection of the "value of a company". Because if all theories are equally arbitrary, what other theory would everyone pick? Surely, they will pick the naive theory, since that is the most salient one--because canonically, one buys the stock of a company that is doing well.

However, if there is no most-salient theory to fall back on when the initial mutually recursive decision process fails in infinite regress, then chaotic instability ensues, and bubbles and sell-offs will non-linearly continue until a sufficiently salient theory once again takes hold, and investors bind their decisions once again to independent empirical data. (Note that it is entirely possible that the chaotic movements of the stock will themselves form the basis of a new theory about the company--for example, if the stock chaotically plummets in a sell-off, this will be interpreted as an "adjustment" in response to "new information" that reveals that the company was "overvalued". But if the stock had chaotically risen in a bubble, the community might just as easily have interpreted this as "renewed investor confidence" and selected some good news about the company to use as cognitively plausible "evidence" for the company's good prospects.)

Given all this, you have to wonder how it is that computers can model the movements in financial markets. Unless your algorithm somehow is able to take into account the salience of various explanatory theories, you cannot make headway into the problem, unless you already assume some empirical theory, and use the empirical data from the theory as inputs for the computer program. But even with this, it would be the human's task to monitor the investor zeitgeist to determine when it shifts to a different most-salient theory, or when there is a theory vacuum and the market goes into chaotic instability.

This whole post was instigated, by the way, by a blog post by Zachary Karabell, where he concludes:

That is where trust becomes even more essential: we have to know that executives are behaving responsibly, in their own self-interest, and that regulators are ensuring that leverage isn’t excessive and capital is. We have to believe that ratings agencies are diligent in affirming strength, especially if we then give them credence when they announce weakness à la downgrading the United States. And we have to imagine that the media report things that have a tangible relationship to something called the truth. But we do not live in that world, and that is a headwind pushing against currents of balance, growth, and repair.

Here, I take him to be saying, in so many words, that the most-salient theories need to be reinforced credibly by institutions if we are to avoid the dreaded state of theory-less chaotic instability in the markets. However, I wonder if what he advises next makes sense:

In that world of trust deficit, we’d do well to repeat the following mantra: just because it happened last time doesn’t mean it is happening again. Being skeptical is healthy; being cynical, not so much. And the only way to judge the present is on the present, not on false application of the lessons of the past, and not on irrational fears of what the future might hold.

My problem with what he says here is that he seems to be placing the blame--or at least, expecting the fix to come entirely from--the investors who have lost "trust" in the prevailing salient theories, rather than the institutions that are charged with instilling and maintaining trust in the theories. It's hardly "irrational" of investors to withdraw from the market when suddenly there's a chance the US Government will stop paying its bills and the German government is mulling over whether it should just let Greece, Ireland, and Spain default on huge amounts of government debt--these remarkable and historical economic events threaten to bring into being a "new normal" that wipes away the old, long-agreed upon most-salient theories
that kept the markets from spiraling off into non-linear chaos.

In light of this, I would think the right "call to action" is to get Wall Street and the rest of the financial world to put extreme political pressure on the US and German governments to guarantee government debt and double-down on the prevailing global financial world order.