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 in Forecasting

Overconfidence bias is a situation in which an analyst has unwarranted faith in their abilities. Studies have shown that people are more overconfident when forecasting what others do not see or expect. One way an analyst can mitigate overconfidence bias is by recording and sharing their forecasts and reviewing them on a regular basis to identify accurate and inaccurate forecasts. Scenario analysis can be used to widen the forecast confidence interval.

Illusion of Control

The illusion of control is a habit of overestimating the ability to control what is beyond one’s control and taking futile actions to gain control. They are often manifested in an analyst’s belief that forecasts can be made more accurate by creating more complex models and obtaining more information. While more information and complex models increase accuracy, marginal returns diminish. Adding immaterial information can lead to misleading forecasts, and complex models tend to overfit historical data sets. More complex data can also conceal assumptions, making updating new information for forecasting difficult.

The illusion of control can be mitigated by only using modeling variables a company regularly discloses, emphasizing the most impactful or important.

Conservatism Bias

This is a bias where an analyst maintains previous forecasts or views because new information conflicts with earlier forecasts. The most common conservatism is the reluctance to incorporate new negative information into a forecast. Nevertheless, analysts may also fail to incorporate positive information, leading to lower estimates. Conservatism bias can also be referred to as anchoring and adjustment.

An investment team can mitigate conservatism bias by regularly reviewing models and forecasts. Alternatively, they can mitigate it by using simpler models with fewer variables, making changing assumptions easier. Since overconfidence and illusion of control are related to conservatism bias, mitigating those biases will mitigate conservatism bias.

Representativeness Bias

Representative bias is the tendency to categorize information based on known classifications and past experiences. New information may seem representative of familiar elements already categorized, yet it is very different and can be viewed from a different perspective. A common form of representativeness bias is base-rate neglect. In this case, a phenomenon’s incidence rate in a larger population or a larger characteristic class to which a specific member belongs (its base rate) is neglected in favor of member or situation-specific information. The situation-specific information is also known as the “inside view,” while the base rate is known as the “outside view.”

Considering both the outside and inside view makes a superior forecast. To do this, an analyst will first consider the base rate so as to determine the factors that make a company different from the base rate and the implications of those differences.

Confirmation Bias

Confirmation bias is the habit of ignoring whatever contradicts prior beliefs and paying attention to those that confirm prior beliefs. Confirmation bias commonly involves tailoring the research process to yield positive news. Overconfidence and representative bias are closely related to confirmation bias.

It is good practice for analysts to speak with management in their research process. Besides, they must be aware of the inherent bias, especially if a company has been involved in past controversies. An analyst should always seek differing perspectives. Two ways of mitigating confirmation bias are to read or speak to an analyst with different perspectives about the security being researched and to seek perspectives from colleagues who have neither psychological nor economic investments in the security.

Question

James, an analyst at AGI Capital, received news that a stock he had been researching on had declined revenue in the previous year due to increased competition. He, however, does not include this information in his analysis. The most likely bias that James demonstrates is:

1. Overconfidence in forecasting.
2. Confirmation bias.
3. Conservatism bias,

Solution

The correct answer is C. Conservatism bias is where an analyst maintains previous forecasts or views because new information conflict with prior forecasts. The most common conservatism is the reluctance to incorporate new negative information into a forecast.

A is incorrect. Overconfidence bias is when an analyst has unwarranted faith in their abilities.

B is incorrect. Confirmation bias is the habit of ignoring whatever contradicts prior beliefs and paying attention to those that confirm prior beliefs. Confirmation bias commonly involves tailoring the research process to yield positive news.

Reading 17: Financial Statement Modeling

LOS 17 (h) Explain how behavioral factors affect analyst forecasts and recommend remedial actions for analyst biases.