What’s called the “confirmatory bias” is on the mind of James Montier in chapter eight of "The Little Book of Behavioral Investing: How Not to Be Your Own Worst Enemy." Broadly speaking, confirmatory bias refers to the human inclination to select evidence that “confirms” what we already think we know.
To explore this topic, Montier created a simple test: What rule did he use when he created the following sequence of numbers: 2-4-6? You may create another three-number sequence to illustrate your understanding of the underlying rule.
He reported most people who take the test come up with 4-6-8 or 10-12-14, which do fit the original sequence. Generalizing, test subjects also say “Any number that increases in increments of two,” or “Even numbers that increase by increments of two,” but Montier says neither of those represent the rule he used:
“In fact, the rule used was 'Any ascending numbers,' but very, very few people ever manage to uncover the rule. The easiest way of doing so is to identify sequences that generate the response, 'No, that doesn’t fit the rule,' such as a sequence of descending numbers or a mixed order of numbers. But most of us simply don’t think to suggest these kinds of sequences. Again, we are too busy looking for information that confirms our hypothesis.”
Montier argued this happens because we are biased toward finding evidence that confirms our bias, rather than evidence that contradicts it. He pointed to Charles Darwin’s practice of writing down facts that apparently contradicted his theory of evolution, and then tried to think through the implications, including the possibility of ultimately making the fact fit.
He added:
“Time and again, psychologists have found that confidence and biased assimilation perform a strange tango. It appears the more sure people were that they had the correct view, the more they distorted new evidence to suit their existing preference, which in turn made them even more confident.”
Investing without confirmatory bias
Investors need to be sure they are not making decisions based on confirmatory biases. Montier offers a couple of success stories, including one that originated with Julian Robertson (Trades, Portfolio) of Tiger Management. One of his many successful disciples was Robert Williamson, and Montier quoted Williamson as saying:
“Julian Robertson was always adamant about seeking out the opposite point of view and then being completely honest with yourself in deciding whether your analysis overrides that. That’s something we try to practice every day.”
Berkowitz
Montier also quoted Bruce Berkowitz (Trades, Portfolio) of Fairholme Capital Management, who said he and his colleagues always look for ways to kill the companies they are considering for investment. They introduce a number of conditions that could hurt a company, including: “recession, stagflation, zooming interest rates, or a dirty bomb going off.”
Berkowitz also created a list of ways in which “Companies die and how they’re killed.”:
- They don’t generate cash.
- They burn cash.
- They become over-leveraged.
- They play Russian Roulette.
- They have idiots for management.
- They put together bad boards.
- They de-worsify.
- They buy back stock when it is overvalued.
- They lie with GAAP accounting measures.
Macpherson
GuruFocus contributor Thomas Macpherson also expressed interest of this kind by attempting to break the investing case for the companies he was considering. In his book, “Seeking Wisdom: Thoughts on Value Investing,” he explained a process called “Getting to Zero”:
“A key component in my investment selection process is creating models that get my investment to exactly that place – broken, impaired, bankrupt. As a means of achieving this, I will test each model, estimate, and assumption and find a way to reach a zero valuation. Put more simply, I want to know how this investment opportunity’s capital can be permanently impaired.”
In his book, Macpherson "deconstructs" and "pressure tests" his business case in five ways:
- Revenue: He started by asking what would happen with this stock if an event similar to the 2008 crisis would cause reductions in revenue from 10% to 75%. He found this particularly useful for assessing companies in the natural resource sector, including oil and rare earth.
- Credit/debt: In this section, he wanted to know what would happen if credit markets collapse. First, he explored the implications should its weighted average cost of capital go up by 100% or 500%. He also asked what would happen if credit was no longer available at all (which happened to many companies after the events of 2008).
- Management: Can management allocate capital effectively? Far too often they cannot and, to make things worse, often get involved in bad merger and acquisition deals, 80% of which end up destroying capital. From there, he considers two factors: dilution and addition of debt.
- Competitive moat: This refers to the ability of a company to protect its pricing power; it includes propriety technology and regulatory fences. For Macpherson, this involves estimating what "commodification" could to do to a company's revenues and earnings.
- Regulatory circumstances: He notes that both state and federal governments keep increasing their reach, so risk assessment should or must include regulatory pressures. One area in which this “must” must be considered is the health industry.
Macpherson summed up by recalling how tech investors were hammered in 2000 and financial stocks were pushed to the brink, and over, of bankruptcy in 2008:
“Nothing can devastate long-term returns than an investment that permanently impairs your capital. Just like a technology investor in 2000 or financials services investor in 2007 - 2009, you will find how difficult it is to recover from catastrophic losses. I think the best way to manage risk against such events is to make them happen before you invest. By 'getting to zero' before they invest a dime, investors can improve their investment returns over the long term.”
Summed up, use a case-breaking strategy to protect yourself against confirmatory bias.
Conclusion
In chapter eight of "The Little Book of Behavioral Investing: How Not to Be Your Own Worst Enemy," Montier introduced us to a silent bias, confirmatory bias, that leads us to put too much emphasis on evidence that confirms what we already believe. Flipping the concept the other way, we might say it keeps us from properly considering other perspectives.
Investors should fear this bias because it can lead them to ignore warnings or problems with stocks. As Montier said, "It appears the more sure people were that they had the correct view, the more they distorted new evidence to suit their existing preference".
To overcome this, he offered the examples of Robertson and Berkowitz. The latter has a process by which he tries to minimize the confirmatory bias.
Similarly, GuruFocus contributor Macpherson has a system in which he tries to get an investing case to go to zero. He manipulates five key criteria to see what might break a company's future success.
About Montier
The author is a member of the asset allocation team at GMO, the firm founded by Jeremy Grantham (Trades, Portfolio) in 1977. According to his Amazon profile, he was previously co-head of global strategy at Société Générale (XPAR:GLE, Financial). The author of three books, he is also a visiting fellow at the University of Durham and a fellow of the Royal Society of Arts. The book we are discussing was published in 2010.
(This article is one in a series of chapter-by-chapter reviews. To read more, and reviews of other important investing books, go to this page.)
