Thinking about Thinking: Part II

Last week, we examined the three types of statements deemed true. This week we will discuss the appropriate assignment of these statements and the dangers in their inappropriate use. We will conclude with how investors can use this analysis. As an aside, these last two WGRs have examined a broad topic outside the usual scope of this report, some “summertime reading,” if you will. Next week, we will return to our usual analysis.

The Proper Roles: Self-Evident Truths
There appears to be a human need for certainty. At weddings, we say “until death do you part” even though casual observation suggests that many marriages end in divorce. We want to know what people stand for. We want to know what is true. These sentiments usually boil down to a priori synthetic statements, which are statements of belief.

A priori synthetic statements that are derived from sensory observation are fraught with risk. That’s because this source of knowledge is always conditional. By induction, it is reasonable to believe the sun will rise in the morning. However, by some geologic or astronomical event, it might not. The scientific method may give us comfort that a relationship between cause and effect is always true, but that is usually only under laboratory conditions.

A priori synthetic statements are most useful in matters of religion, morals and ethics. These core beliefs, to a great extent, define the key characteristics of human beings. Where do these core beliefs come from? Often, core beliefs come from revelation, tradition and sentiment. At the same time, although we may individually treat these statements as self-evident, they are not universally held, which means that they are not self-evident.

After Kant, the general trend in philosophy has been to postulate that there is no logical way to evaluate moral or ethical claims. In other words, a priori synthetic statements about the “oughts” of human behavior cannot be tested. Although we understand philosophers’ reluctance to delve into these areas, the reality is that not testing them makes all of them acceptable. It’s difficult to see how one could build a functioning society where there are no generally acceptable guidelines to behavior.

So, here are a couple of tests that I have found useful.

First, it probably makes sense to test an a priori synthetic statement on morals and ethics by the outcomes it produces. For example, Peter Singer, an Australian moral philosopher, is a strict utilitarian. The goodness or badness of an outcome is in the pleasure or pain it produces. In other words, something is good if it makes more people happy than makes them sad. This has led him to postulate that a chimpanzee, which may be more sensate than a severely disabled human, has a greater claim on life and resources because the former can feel more pleasure and pain. This is a logical conclusion from his self-evident truth. I would argue that this outcome is so repulsive that it undermines the veracity of his a priori synthetic starting point and thus should prompt us to find a different starting premise.

Second, a priori synthetic statements about morals and ethics should be tested by Kant’s categorical imperative. Essentially, it states that if everyone held one’s a priori synthetic statement as the highest good, how would society exist if that statement became how all acted?1 For example, if a society held that infanticide was its highest value such a society would probably die out in a few decades. Thus, it could not be the highest value for human behavior because if it were, there would be no more humans!

Still, even with these two tests, it is hard to discuss a person’s closely held beliefs. It is not the purpose of this report to fully develop an ethical stance. Instead, my goal is to point out that religion, ethics and morals are the proper place for a priori synthetic statements of knowledge. Given how deeply these positions are held, there is a natural tendency to associate with people who hold similar views. The internet has probably accentuated this tendency as one can screen information sources to best match one’s self-evident truths. While this may make people feel comfortable, it also tends to separate people into tribes and can reduce tolerance toward people who hold differing beliefs.

Given the power of a priori synthetic statements, it is important to carefully consider them throughout life. Most of us adopt the positions of our parents; often, these are categorically rejected in adolescence.2 Still, because they define us, such statements should be the product of serious reflection. It is probably important to realize that tolerating the different self-evident truths of others doesn’t necessarily mean agreeing with them. It just means that one is cognizant of their importance.

Using a posteriori statements for ethical and moral values is probably inappropriate because it suggests that a person would hold a belief until new information is presented. Such potential vacillation in morals and ethics isn’t something a person should do lightly. In other words, the risk of not using a priori synthetic statements in the area of morals and ethics runs the risk of devolving into nihilism and solipsism.

The Proper Roles: Scientifically-Derived Truths

On the other hand, in most other matters, conditional statements of knowledge are probably appropriate. For investors, this is a key insight.

One of the most insipid quotes one hears is, “You are entitled to your own opinion but not your own facts.” Why is this quote meaningless? Because we rarely quarrel over facts as facts are simply data. The key isn’t facts, it’s how the facts relate to each other. In other words, assume there is a fact, such as a market correction. The fact that the correction occurred isn’t much of a debate. The debate is “why” it occurred. Outside the laboratory, it is usually impossible to control all the factors that can influence a cause and effect relationship and thus, everything we can say about such relationships is tentative.

Throughout my career, there have been market relationships that appeared to be strongly correlated but eventually faded. That doesn’t mean the information isn’t valuable. It just means that one should realize that (a) what appears to be causal now may not remain so indefinitely, and (b) it pays to be aware of other factors surrounding the relationship as well. For example, a classic linear regression equation is:

Dependent variable = Constant + independent variable + error term

The error term contains all the other variables that can affect the relationship between the dependent and independent variables. In creating models, the analyst looks for residuals that are stable. If they are not, it suggests that there is a missing variable that should be included in the model.

In real world analysis, there is always the potential for an infinite number of missing variables. As long as these missing variables, which are contained in the error term, offset each other, a model can perform well over time. However, oftentimes a factor will emerge that has a significant impact on the causal relationship for a certain amount of time, only to fade into insignificance in the future. When St. Louis FRB President James Bullard talked about “policy regimes” in a recent paper, 3 he was, in my opinion, suggesting that “timeless” relationships are not always evident and central banks will focus on different factors at different times. This paper caused quite a stir. Here’s why…

Isaiah Berlin once penned an essay called “The Hedgehog and the Fox.”4 He began by quoting a Greek poet named Archilochus who said, “The fox knows many things, but the hedgehog knows one big thing.” Berlin suggested that hedgehogs were thinkers who tried to create a unifying principle for all knowledge. Foxes, in comparison, don’t necessarily focus on any unifying principle but try to observe the world and derive relationships which may or may not reflect a unifying idea.

Analysts who hold that a unifying principle is true in all cases are hedgehogs. They are often on television. Producers like them because they don’t surprise hosts with inconsistent views. Often, a show wants a pundit who supports a position and, for balance, another with views opposite from the first pundit. Producers want two hedgehogs because they fear having two guests on the show who unexpectedly agree, which may be closer to the truth but makes for bad television.

There is a tendency for analysts, investors and academics to elevate an a posteriori synthetic statement to an a priori synthetic statement as they become more invested in the position. In financial markets, it is not hard to find analysts labeled “permabulls” or “permabears” because of their persistently held positions. This occurs in academia as well. Max Planck, the famous physicist, was quoted as saying, “Science advances one funeral at a time.” In other words, as students progress through their studies into graduate work, they tend to join theoretic “tribes” and hold those positions throughout their careers. Dominant theories don’t fall from grace because academicians change their minds but because their proponents pass from this earthly plane.5

The academic world seems to create hedgehogs. However, in a recent book, Philip Tetlock made the case that foxes are superior in forecasting events.6 If an analyst or investor starts with a unifying principle, an a priori synthetic statement, it can blind them to other possible explanations for a causal relationship. Although there is a certain comfort in believing in unchanging principles, there is a strong case to be made that in matters of science, both hard and social, as well as in investing, it is probably prudent to believe that causal relationships between variables are only true until proven otherwise. In other words, a posteriori statements, scientifically-derived truths, should always be treated as conditional and subject to change.

Concluding Thoughts
First, the next few months will likely reveal political, social and emotional divisions in the U.S. that will put people on edge. To a great extent, the upcoming presidential election will lay bare competing a priori synthetic positions that have previously been unexamined.

It is probably worth remembering that once self-evident truths have been accepted, they are rarely changed. It is probably hopeless to argue with someone about such positions. This doesn’t mean they can’t be discussed but, once a self-evident truth is uncovered, it’s probably better to treat it as an insight rather than as a challenge to change a person’s mind. At the same time, the proper use of a priori analytic statements, basic logic, are also important. Sometimes, a person will hold a view or engage in a behavior that is inconsistent with a stated belief. This can open up avenues of discussion.

Second, it is worth reiterating that assigning statements to the proper areas is critical. A priori synthetic statements are appropriate for moral, religious and ethical arenas. Not having them is the path to nihilism and solipsism. If all moral choices are situational and without reference to a self-evident truth, a person might be capable of anything. In other words, lacking a priori synthetic principles could lead a person to act as though the ends justify the means and thus to immoral behaviors. Obviously, each person needs to examine these self-evident truths and make their own moral judgements (after all, that’s what following one’s conscience is all about), but it is in this area that self-evident truths should reside.

In other areas of life, especially in observations of the outside world, a posteriori synthetic statements should dominate. However, one should be acutely aware of the weaknesses of such statements. You never know anything with certainty; you rely on something being true “so far.” This means the term “settled science” is, in a sense, a non sequitur. Instead, we can say something is true based on our analysis of the available evidence, but that position can change as new information emerges.

Bill O’Grady
August 22, 2016

This report was prepared by Bill O’Grady of Confluence Investment Management LLC and reflects the current opinion of the author. It is based upon sources and data believed to be accurate and reliable. Opinions and forward looking statements expressed are subject to change without notice. This information does not constitute a solicitation or an offer to buy or sell any security.

Confluence Investment Management LLC

Confluence Investment Management LLC is an independent, SEC Registered Investment Advisor located in St. Louis, Missouri. The firm provides professional portfolio management and advisory services to institutional and individual clients. Confluence’s investment philosophy is based upon independent, fundamental research that integrates the firm’s evaluation of market cycles, macroeconomics and geopolitical analysis with a value-driven, fundamental company-specific approach. The firm’s portfolio management philosophy begins by assessing risk, and follows through by positioning client portfolios to achieve stated income and growth objectives. The Confluence team is comprised of experienced investment professionals who are dedicated to an exceptional level of client service and communication.

1 In practical terms, this is what most mothers would retort to a child’s plea that “everyone is doing it.” Her response is, “If everyone ran off a cliff, would you do it too?”

2 Which is why raising teenagers is such a struggle.



5 For example, in economics, Keynesians dominated until the late 1960s when monetarists began to overturn the tenets of Keynes. The Rational Expectations school, along with Supply Side economics, emerged in the 1980s. A resurgence of Keynes has occurred in the wake of the 2008 Financial Crisis.

6 Tetlock, P. (2015). Superforecasting: The Art and Science of Prediction. New York, NY: Crown Publishers.

© Confluence Investment Management LLC

© Confluence Investment Management

Read more commentaries by Confluence Investment Management