In the first piece of exclusive content written for Investor’s Guild members, Patrick Cairns chatted to Gryphon’s Chief Investment Officer, Reuben Beelders about the emphasis that the Gryphon places on constantly re-evaluating its decision-making process.
Gryphon’s chief investment officer, Reuben Beelders, believes that there is a big mistake that is easy to make when evaluating the quality of an investment firm’s decision making.
“We have a human tendency called ‘resulting’,” Beelders says. “When we ask ourselves whether or not we made a good decision, we don’t think about the process – we think about the result.
“We’ll say we’ve made a good decision not because the process was good, but because the result was good. But that also means that when the decision goes against us, we assume it was a bad decision, even when maybe it wasn’t.”
In an industry where even the best decision makers are getting at least 40% of their calls wrong, Beelders believes that only looking at the outcome is never sufficient. It’s important to be able to evaluate a decision at the time it was made.
In order to facilitate this, Gryphon uses three approaches for testing and improving their decision making.
The first is keeping decision diaries that record everything about a decision at the time it is taken.
“Annie Duke talks about hindsight bias,” Beelders says. “When we look into the past and see only the thing that happened, it seems to have been inevitable.
“But a decision diary is an attempt to force you to write down the information you have at the time you made the decision. In this industry, it’s easy to look back on something and say it was a bad decision because it didn’t work out. But, based on the information available at that point in time, maybe it was the correct assessment. Keeping a decision diary means you have an accurate representation of what could have happened and not a version edited by hindsight.”
The second approach is to use a pre-mortem. This assesses a decision before it is made by imagining a future in which the result of that decision has been poor, and asking what has happened for that to be the case.
“The big thing with pre-mortem is to ask yourself if this decision goes wrong, what are the things going to be that could cause it go wrong?” Beelders says. “It forces you to ask: How well do I know this subject? Have I made this kind of decision before? Have other people made this decision before?”
Identifying a ‘reference class’
The last of those three questions is perhaps the most useful, since it allows you to learn from others’ mistakes. Michael Mauboussin talks about this as the ‘reference class’.
For example, Beelders points to the example of a locally listed retailer that made a decision to acquire a company in Australia in 2014. A number of its peers had already tried to enter Australia without much success. The reference class was clear and obvious in this case, but ignored.
The third approach is to follow Charlie Munger’s dictum of always understanding the alternative hypothesis.
“When our indicator says we should be out of equities, for example, we don’t just leave it there,” Beelders says. “We need to understand the argument for why we should be long equities.”
This particular question is one that Gryphon has been engaged with for some time to ensure that they don’t succumb to confirmation bias in maintaining their current position.
“After Covid, our buy indicators didn’t kick in,” Beelders says. “We thought that maybe our indicators had missed something, so we tested that. We said that if there was a pickup in the movement of goods between countries, our indicators may be wrong.
“One of the ways we investigated that was to see whether volumes in shipping lines were increasing or not. We looked at the underlying volumes, and came to the conclusion that they were not growing and so our market-based indicators were in fact correct.”
Checking the signals
As a rules-based manager, Gryphon is essentially continually evaluating if there could be a reason why the rules it follows will not work.
For example, Gryphon’s primary buy signal is the 10-year bond yield minus the three-year bond yield. However, Beelders and his team have had to question whether market dynamics might be affecting this reading.
“The reason is that the R186 is significant in South Africa in terms of market value and volume of trade, but it was a 10-year bond 10 years ago, and it’s now a one-year bond,” Beelders explains. “We wondered whether the fact that the R186 had increasingly become the three-year bond had impacted the pricing of the generic three-year.
“We considered the possibility that the bond was priced where it was because of market liquidity, when it should have been higher. But we investigated that, and it proved not to be the case.”
As rules-based investors, it is critical that Gryphon’s team doesn’t override the signals. That would be a fundamental break from their philosophy. But it is equally important that they are constantly evaluating whether the world has changed and if the signals are still relevant.
For Beelders, this isn’t about whether rules-based decisions always lead to good outcomes. But whether the correct decisions are being made based on what is known at any given time.
“It’s a question of saying: this is what I know, this is what I expect, and these are the metrics I am basing the decision on. It’s not a question of whether the result was good, but whether I went through the right process in making the decision,” Beelders explains.
“Because you can govern the decision process. That is something that is within your control. The outcome is not in your control. The better you become at the process, the more likely you are to deliver the outcomes you want.”
