This article highlights some of the common cognitive biases that I have come across in my...
Can FP&A Ever Be Predictive Without Integrating Behavior?
The title might seem a little radical. After all, FP&A makes no provision for integrating behavior as a factor in planning and analysis. So why even raise the issue?
I have been CEO of several companies including one public company in the US. Of course, FP&A was a normal and routine activity in those companies. The most common activity was to plan for future revenues and expenses. Naturally we would go to the head of sales for his thoughts on the matter. When my financial people would present his estimates to me, I would always discount them by maybe 30% because I knew from experience that he was always too optimistic.
And I’d do the same sort of thing with expenses. Here I would ask my CFO for his estimates. In his case I would always increase them around 20-30% because I knew that from experience, he was always too pessimistic about costs and expense would always turn out to be higher than he projected. Usually my revised estimates were pretty good, since I had made these adjustments.
So, you can see that the adjustments I made were adjustments made to account for personal behavior. If I hadn’t made them the estimates would have been way wrong. I think what I did is very common in the business world. CEOs and top people presented with estimates of financial futures will take estimates from their people and will adjust them based on their knowledge of their particular behaviors.
In a nutshell, that’s why behavioral economics and behavioral finance were invented. It’s gradually dawned on everyone connected with economics and finance that our assumptions were way too simplistic. For one, we have assumed that people always make rational decisions. Of course, we don’t. But economics and finance have been founded on that bedrock principle.
And we always believed that when people made economic and financial decisions, we didn’t have any biases one way or the other when we made them. We didn’t let our emotions or thinking patterns affect whatever decision we made. Of course, that was totally wrong too.
Several Nobel prizes have now been awarded in these new disciplines of behavioral economics and finance. The first was to Daniel Kahneman in 2002 and the latest was awarded to Richard Thaler in 2017. These disciplines are based on the truth that humans don’t always make rational decisions. Additionally, we all have unconscious cognitive biases that systematically sway our decisions in ways that are not always rational and lead to suboptimal financial decisions. Al of this occurs without us generally being aware of it.
The idea that your analyses are not impacted by your behavior is belied by a new field of research called behavioral accounting. This focuses on the impacts of behavior on accounting frameworks and their data products.
In other words, we can’t assume that even strict accounting rules are going to shield us from our cognitive biases. This is because we also re-interpret accounting rules according to our cognitive biases. But as most people don’t know what their biases are, it’s difficult for them to realize what is happening to the data they produce in terms of its reliability and level of predictability.
If you want an example go no further than that famed icon of American industry, General Electric. GE was the poster child for financial planning; it wrote the book so to speak. For many years it provided earnings guidance for analysts. We now know that it utilized “creative” accounting techniques to build cushions so that it could “smooth” its earning history.
The use of these cushions was discretionary and therefore dependent on the behaviors of the leaders who dictated the use of these approaches. So, what looked to be predictable earnings was anything but. In fact, without this approach, the earnings would have been unpredictable.
GE’s approach used to be common, but we know that many companies still use questionable techniques to smooth their earnings, albeit in not so obvious a manner. They are doing it because they understand that they must use some type of approach to reduce earnings volatility, which investors don’t like.
But here’s the crux of the matter. Earnings volatility is often the result of behavioral change and volatility, on the part of the leaders of the company, and some of its top players. Sometimes companies use these techniques precisely because they understand that only “unnatural” adjustments, such as illicit or unconventional approaches can seemingly get rid of that volatility.
In other words, the continued existence of these techniques is due to a tacit recognition by many people that you can only make predictions if you somehow adjust for the volatility in earnings that comes from the volatility of human behavior.
This factor has spawned a formal metric, namely “beta”. Beta measures stock volatility. It was developed because people understand that volatility is a natural phenomenon.
GE and many companies use the “smoothing” techniques to hide the impact of natural volatility on their earnings. They understand that if they accounted with legitimate financial accounting and planning techniques, they would not be able to provide the earnings guidance they would like to portray, namely stable, predictable earnings. They use these techniques to make their planning appear more rational than it should be, to be faithful to what is really happening in their operations.
So, there is impeccable academic and practical stock experience evidence for the belief that behaviors, not just products and assets, are a key part of profitability and financial outcomes. That’s why it’s relevant to FP&A, in fact, not just relevant, but vital. The issue is, how do you use this knowledge?
The first thing you must realize is that behavior includes analysts too, that means you yourself, not just the people who actually make the decisions. What are your cognitive biases and how will you find about them, especially how they impact your estimates and analyses?
How do you figure out the cognitive biases of others? That includes your co-workers, data colleagues, and the main actors in the organization for whom you are producing data and analyses. There are ways of doing this using behavioral and psychometric assessments. But the first step is to figure out your own biases and how they affect your own work.
What are some of the main ways you can leverage behavioral analysis to make FP&A much better? Here’s some places you can start:
- Behavioral forensics to improve financial outcomes: there are ways of using cognitive biases to predict the gross margin of both individual executives, the management team and the company. It’s entirely separate from product analysis. There is a way to use another cognitive bias to predict indirect expenses, again entirely separate from conventional cost analysis.
- Use behavioral analysis to predict M&A outcomes. By using these cognitive biases, you can construct a behavioral proforma to compare with the traditional M&A proforma. As we know traditional M&A analysis is very unreliable, so we badly need another approach to predict the outcome of M&A transactions. The behavioral proforma is used as a complement, not as a substitute.
- We can use behavioral analysis as a new approach to performance and operations analysis, both to predict profitability, and to predicting the likely contribution to profitability of individuals and teams.
- It’s clear with all this you can also employ behavioral forensics and analysis to predict investment outcomes; areas include portfolio analysis in private equity portfolios, hedge fund returns, and the returns from other types of equity investments.
- We have already referred to behavioral accounting. Behavioral forensics can also be used to predict the likelihood and existence of fraud and incorrect payments.
- Behavioral analysis is a much-needed upgrade to risk analysis and compliance. Traditional risk and compliance approaches don’t take behavior into account. This constitutes a huge conceptual gap in global risk management approaches. By using behavioral factors and analysis of cognitive biases we can also reduce risk levels everywhere.
Like all disciplines everywhere, FP&A must move with the times. It’s not enough to count on traditional FP&A approaches to preserve the credibility of FP&A analysis. Just like every other traditional approach, these must be upgraded and, in some cases, revolutionized. Utilizing new approaches based on behavioral economics and behavioral finance to is an excellent place to start.