Friday, September 18, 2009

Peirce and abduction

It sometimes seems as if the only philosophers traders have ever heard of are Karl Popper and Ayn Rand. I assume that more than anything else it’s because of the power of association. George Soros lauded Karl Popper and he’s rich; Alan Greenspan was part of Ayn Rand’s inner circle and he’s influential. Logic like that can leave you both undereducated and broke.

Popper’s doctrine of falsification (that a theory is scientific only insofar as it is falsifiable and that scientists can’t prove hypotheses to be true but can only identify which hypotheses are false) is incomplete but nonetheless important. Ayn Rand I leave to her fans.

I’ve always been intrigued by how people formulate hypotheses and how they fix broken ideas. Both are central to investing and trading. Today I want to deal with the first and introduce Charles Sanders Peirce’s notion of abduction, outlined in papers written just after the turn of the twentieth century. [His name, by the way, for those unfamiliar with his work is pronounced “purse.”]

Peirce recognized three elementary kinds of reasoning: deduction, induction, and abduction. Deduction and induction are familiar logical concepts. Abduction is not; we normally understand abduction to mean kidnapping. But for Peirce abduction describes “all the operations by which theories and conceptions are engendered.” It “consists in studying facts and devising a theory to explain them.” It offers suggestions; it “suggests that something may be.” It’s something like a conjecture or a guess.

The form of inference is this:
“The surprising fact, C, is observed;
But if A were true, C would be a matter of course,
Hence, there is reason to suspect that A is true.”

Or, as further elaborated, “A mass of facts is before us. We go through them. We examine them. We find them a confused snarl, an impenetrable jungle. We are unable to hold them in our minds. We endeavor to set them down upon paper, but they seem so multiplex intricate that we can neither satisfy ourselves that what we have set down represents the facts, nor can we get any clear idea of what it is that we have set down. But suddenly, while we are poring over our digest of the facts and are endeavoring to set them into order, it occurs to us that if we were to assume something to be true that we do not know to be true, these facts would arrange themselves luminously. That is abduction.” We cannot say that the hypothesis is true or even probable (that is, not probable in a way that underwriters could safely make it the basis of business). But it’s likely, “in the sense of being some sort of approach to the truth, in an indefinite sense. The conclusion is drawn in the interrogative mood.”

The value of these tentative hypotheses is economic—“economy of money, time, thought, and energy.” They allow us to regulate our future conduct rationally; induction from past experience gives us reason to believe that an abductive inference will be successful in the future.

So what value does Peirce’s hypothesis about abduction hold for the trader or investor? First, it does not restrict hypotheses to those that can be falsified. These are not scientific hypotheses in Popper’s sense. They make sense out of data but make no claim to truth themselves. They offer a way of moving forward, of trying to make things work. They may eventually give way to other hypotheses that are a better fit with the facts. But in the meantime they have pragmatic value.

Trading and investing do not lend themselves to true scientific hypotheses. No matter how sophisticated the mathematics, for instance, financial hypotheses are in the end mere conjectures with only a vague claim to truth. Does that make them worthless? Of course not. If they rise to the level of abduction (that is, if they are the result of inference, not merely a shot in the dark), they provide a way of looking at surprising or snarled data and making sense of it. Abductive inferences, it should be stressed, don’t have to be mathematical. That may be the flavor of the day, but there are many ways to hypothesize about financial data. It’s all up to the abductive powers of the trader!


  1. Hi Brenda,

    Excellent post, thanks for the insight. Could you please provide an example (just as an ilustration, it doesn't have to be a groundbreaking revelation)? Also, how do you avoid falling into the correlation/causation error when abducting?

    Thank you for the blog, very original. Best trading,


  2. Thanks for the compliment, Jorge.

    Let’s take Elliott wave theory as one example. Stock charts appear to be a tangled mess, but the Elliott wave theorist purports to describe price movement by appealing to a few general principles. And he claims predictive power for these principles. Of course, as critics point out, there’s no clear way to apply the principles. Without the equivalent of an eraser (oops, it’s not wave 3 but wave a, etc.) Elliott wave theory can’t move forward, so it’s not a particularly good hypothesis.

    A better example, in the sense that it more or less works and was Nobel Prize worthy, is the Black and Scholes method of options volatility trading. That is, using a pricing model (and some skill) a trader can reap the amount by which an option is mispriced by a continuous delta-neutral hedging process. (I’m going to write more about this next week in a book review.)

    The very thorny problem of causality vs. correlation doesn’t really plague hypothesis formulation in Peirce’s sense. The abduction is meant to be an explanation of events that may be causally linked, may be correlated, or may even be close to random.

    By the way, Peirce did a lot of seminal work on statistics (using regression and correlation, for instance) because he believed that the fruits of science were at best only statistical probabilities.