When analyst bias becomes fund manager blind spots

The role of air quality in analyst bias

Professor Dasgupta and his collaborators, Shi Yushui and Xia Ying of Monash University, and Wu Weili of Central University of Finance and Economics, conducted their research using Chinese data, examining how fund managers assess information from their familiar analysts and other analysts in their portfolio decision-making process. They also tested their findings using US data and found consistent results.

The researchers examined the Crystal Ball Awards for Sell-Side Analysts in China from 2011 to 2019, which allow fund managers to vote for their preferred analysts. Analysts who received a fund manager’s vote in the prior year were considered connected to that manager. Hence, the researchers call them “connected analysts”.

The researchers analysed how fund managers used information from these connected analysts. Specifically, they looked at companies in a fund’s portfolio that the connected analysts have visited physically and issued earnings forecasts within 15 days of their visits. Fund analysts often make visits to corporations as part of their research and analysis process.

Although it is commonly assumed that fund managers use analyst information in their portfolio decisions, proving this causality is challenging due to other variables that can influence them. Therefore, the researchers analysed how exogenous shocks, such as air quality during corporate visits, affected the analysts’ forecasts and, eventually, fund managers’ decisions.

Air quality changes are highly unpredictable and influenced by unrelated factors, making it a suitable instrument for isolating the impact of biases that arise. Fund managers normally do not take part in the same site visits. “Air quality on a given day at a firm’s location can affect an analyst’s outlook or mood the forecast the analyst issues of the firm’s earnings per share,” Professor Dasgupta says.

Connected analysts vs. unconnected ones

The team used the air quality index on the visiting days to measure how environmental factors influenced the optimism of the analysts’ forecasts. The results revealed that analysts exposed to higher levels of air pollution during site visits tended to issue less optimistic earnings forecasts. Conversely, better air quality led to more optimistic predictions.

The next question was whether these biases influenced fund managers’ decisions. As expected, fund managers tended to increase a company’s stock holdings if their connected analysts issued more optimistic forecasts following visits with good air quality. These managers increased the stock holdings by a sizeable amount, equal to 6.55 per cent of the typical trading changes.

To further explore fund managers’ decision-making, the researchers examined whether unconnected analysts carried similar weight. The researchers compared forecasts from unconnected analysts to the average predictions from other analysts over the previous three months, as well as assessing the importance of a stock to a mutual fund by comparing the stock’s trading value to the total value of all stocks in the fund.

The analyses found no evidence that biases in unconnected analysts’ predictions influenced fund managers. “Found managers are choosers and assign very little weight to unconnected analysts’ forecasts, thereby leaving their actions vulnerable to biases that affect the connected analysts’ forecasts,” Professor Dasgupta explains.

Expertise helps reduce bias

The researchers then examined the performance of stocks that fund managers traded based on connected analysts’ biased forecasts. The findings indicated that air-quality-influenced optimism from connected analysts is negatively associated with the performance of fund managers’ trading decisions. The connected managers tended to increase the holdings of stocks that later showed lower returns.

“Fund managers make poor trading decisions when they are affected by biases that inform their connected analysts’ forecasts,” says Professor Dasgupta.

It is worth noting that not all fund managers rely heavily on their connected analysts. Professor Dasgupta observes when fund managers have superior industry knowledge, or when a given stock is less difficult to value, they will rely less on connected analysts. “Their expertise and knowledge could allow them to identify biases in analysts’ earnings forecasts.”

The team found that fund managers rely more on connected analysts when the specific industry sector is thriving. In such competitive environments, selecting the right stocks becomes more challenging, increasing the reliance on familiars, Professor Dasgupta explains.

The flip side of familiarity

While fund managers gain valuable information from their networks, this advantage can come at a cost. A heavy focus on connected analysts’ information exposes fund managers to biases, a phenomenon Professor Dasgupta refers to as “tunnel vision,” where fund managers only focus on information from their connected analysts while undervaluing information from other analysts outside their networks.

“When institutional investors are marginal investors, individual analysts’ opinions can affect asset prices and individual biases can be reflected in trading volume and affect price efficiency,” he says.

RELATED ARTICLE

Where Guanxi Matters: The Modern Chinese Financial Sector

Even the most experienced financial professionals are not immune to cognitive biases, which have far-reaching implications for the financial industry, particularly in the areas of investment strategy, market efficiency, and regulatory oversight. Fund managers and analysts alike should recognise the risks associated with tunnel vision.

For analysts, Professor Dasgupta advises being mindful of external factors that could unintentionally influence their forecasts. He also recommends providing clear and comprehensive explanations of the reasoning behind their predictions to help fund managers understand the analysis better.

For fund managers, diversifying information sources is crucial to better their decision-making. “Fund managers can avoid over-reliance on analysts within their networks and incorporate diverse viewpoints to mitigate biases associated with connected analysts,” Professor Dasgupta adds.