Intangible Assets and Financial Analysts Herding Behaviour

Abstract

The question of whether financial analysts provide unbiased forecasts based on the information available to them has been widely discussed in the finance and accounting literature. Financial analysts play a vital role in disseminating information, and their forecasts are important inputs to financial markets. Hence, there is a pressing need to comprehensively understand their forecasting behaviour. Prior studies suggest that financial analysts tend to “herd”, i.e., produce forecasts that imitate each other’s, to avoid the risks to their future careers and reputations inherent in producing bad forecasts. However, the increasing presence and significance of firms’ intangible assets, combined with the difficulty of their valuation, has made analysts’ forecasting tasks increasingly challenging. This study investigates analyst herding behaviour, particularly that which is associated with the intensity of firms’ intangible assets. I used three accounting-based proxies to identify a positive association between firm-specific intangible assets and analyst herding behaviour. More specifically, I first investigated analysts’ herding tendencies at the individual level by examining revised earnings forecasts. As expected, all three proxies were positively and highly significantly, demonstrating that the probability of issuing a “herd forecast” increases with the amount of intangibles of the firm being analysed. The empirical results also demonstrate that analysts’ general experience does not significantly decrease their tendency to herd, while firm-specific experience does. This indicates that firm-specific knowledge or networking may help analysts to obtain high quality private information which helps them to avoid herding. Secondly, at the aggregate level (firm level), I examined the combination of cross-sectional forecast errors and the standard deviations of earnings forecasts. These were used to measure analyst herding behaviour. I found a positive association between firm-specific intangibles and analyst herding behaviour after controlling for various firm characteristics. In addition, compared with analysts who only provided earnings forecasts, analysts who issued both earnings and cash flow forecasts were less likely to herd when covering firms with intensive intangible assets. Finally, consistent with prior research, I found a negative association between analyst herding behaviour and firm market value. An additional determinant of analyst herding behaviour was found, which implies that financial analysts are less confident in their private information when the firms they are analysing have a high intensity of intangible assets. In addition, the results of this thesis confirm the notion that herding behaviour is correlated with task difficulty. This thesis has valuable implications for information dissemination mechanisms in financial markets. The private information of analysts tends to be underweighted and not fully reflected in their earnings forecasts.