Behavioral Finance and Analyst Forecasts

Financial statement models are not immune to behavioral biases. Analysts must be aware of the impact of behavioral biases and solutions to improve investment decisions and forecasts. The five key behavioral biases are overconfidence, conservatism, confirmation bias, the illusion of control, and representativeness.

Overconfidence

Overconfidence is a behavioral bias where analysts may overestimate their ability to forecast accurately. An analyst might place excessive trust in their financial analysis models, neglecting to consider potential external factors that can drastically affect a company’s performance. This misplaced confidence can lead to inaccurate and unreliable predictions, potentially causing substantial financial misjudgments.

To address overconfidence bias, analysts should record and openly discuss their forecasts, consistently reviewing them to identify both accurate and inaccurate predictions. For instance, given the wide variability in most financial variables, analysts will often find that their predictions are incorrect as frequently, or more so, than they are correct. The aim is to recognize the high rate of forecast errors and take measures to broaden the confidence intervals of their forecasts. One effective method is scenario analysis, where analysts ask themselves, “Where might I be wrong and by how much?” to generate various forecast scenarios.

Illusion of Control

The illusion of control, a bias linked to overconfidence, involves overestimating one’s ability to control uncontrollable factors, leading to ultimately futile actions in pursuit of control. This bias often leads analysts to believe that forecasts can be made more accurate by gathering more information and opinions from experts and by creating more detailed and complex models.
While additional information and complexity can enhance forecast accuracy, the benefits diminish over time. The finite amount of relevant information for an investment means that adding irrelevant data can be misleading. For example, overly complex models may be overfitted to historical data and fail to perform well in diverse environments with unprecedented outliers. Moreover, excessive data and model complexity can obscure assumptions and complicate the process of updating forecasts with new information.

To mitigate the illusion of control, it is essential to recognize that uncertainty is inherent in investments. This can be managed by limiting modeling variables to those regularly disclosed by the company, focusing on the most impactful variables, and consulting only those with unique or significant insights.

Conservatism

Conservatism bias refers to the tendency to stick to prior views or forecasts by inadequately incorporating new information. This often occurs in forecasting when an analyst fails to update their predictions after receiving contradictory information, such as disappointing earnings results or actions by competitors. While the most common form of conservatism is the reluctance to include new negative information into a forecast, analysts can also fail to adequately incorporate positive information, leading to overly low estimates.

Another term for conservatism bias in this context is anchoring and adjustment, where an analyst uses their previous estimates as an “anchor” and makes subsequent adjustments. While updating a previous forecast is not inherently problematic, the prior forecast often has a significant influence, causing adjustments to be insufficient and resulting in an updated forecast that remains too close to the original one.

Mitigating conservatism bias can be achieved by having an investment team review forecasts and models regularly, such as quarterly, and by creating flexible models with fewer variables to facilitate easier assumption changes.

Moreover, since conservatism bias is related to overconfidence and the illusion of control, addressing these biases can also help reduce it.

Confirmation Bias

Confirmation bias is the tendency to favor information that supports existing beliefs while disregarding or undervaluing information that contradicts them.

In the context of investment analysis, this bias often manifests when analysts focus only on positive news or specific criteria or conduct research with a limited scope. This bias is closely linked to overconfidence and representativeness biases.

To mitigate confirmation bias in forecasting, analysts should consult research from those with negative opinions on the security in question and seek input from colleagues who are not economically or emotionally invested in the security.

Representativeness Bias

Representativeness bias involves the inclination to classify new information based on past experiences and familiar categories, even when the new information might be different and require a new perspective. This bias can lead to an incorrect understanding that influences future thinking about the information.

A common form of representativeness bias in forecasting is base-rate neglect. This occurs when a larger population’s overall incidence rate or characteristics (the base rate) are ignored in favor of specific information about the situation or individual. The base rate perspective is sometimes referred to as the “outside view,” while the situation-specific perspective is the “inside view.”

For instance, when an analyst models operating costs and margins for a biopharmaceutical company, the “inside view” would focus on company-specific factors, such as the types of drugs sold and the number of salespeople required in different regions. The “outside view” would consider the company as part of the larger biopharmaceutical industry and use industry averages for metrics like gross margin and R&D expenses as a percentage of sales.

A superior forecast typically involves considering both perspectives. One effective approach is to start with the base rate and then adjust for factors that distinguish the target company from the industry average.

Question

Sophia, a financial analyst, has been closely monitoring the growth of a start-up tech company, TechGrowth Inc. The firm has experienced consistent revenue growth over the past five years, leading many analysts to predict a continuation of this trend. Despite a recent report highlighting potential legal issues that could affect TechGrowth Inc.’s operations, Sophia is convinced the company’s revenue will continue to grow unimpeded. She bases her projection on the company’s past performance, assuming it will consistently replicate its success in the future.

Which of the following biases is Sophia most likely exhibiting in her analysis?

1. Representativeness Bias
2. Conservatism Bias
3. Illusion of Control Bias

Solution

The correct answer is A.

Sophia exhibits representativeness bias. This bias occurs when individuals unjustly categorize new information based on past experiences or classifications, often leading to base rate neglect. Sophia’s assumption that TechGrowth Inc. will persistently replicate its past success in the future, despite new information about potential legal issues, is a manifestation of this bias. She erroneously believes that the firm’s future growth will mirror its past growth based on the pattern observed in the previous years.

B is incorrect. Conservatism bias involves the reluctance to revise one’s belief upon receiving new information. Although it might appear that Sophia is ignoring the report about potential legal issues, her decision is based on the perceived relevance of past performance to future growth rather than an unwillingness to adjust her predictions in light of new information, making representativeness bias a more accurate characterization of her behavior.

C is incorrect. The illusion of control bias would entail Sophia believing that she can influence or control outcomes that are actually beyond her control. In this scenario, Sophia is not attempting to control the outcomes; rather, she is making a predictive error based on past performance, aligning more with representativeness bias.