Who reacts to what information in securities analyst reports? Direct evidence from the investor trade imbalance

Highlights

• Foreign institutions and domestic funds actively respond to analysts' information.

• Individual investors provide liquidity during the release of analyst reports.

• Informed agents trade based on recommendations and target prices but not earnings forecasts.

• Investors react more intensely to a recommendation if it is the only strong signal in the report.

• Stronger target signals do not elicit disproportionally larger investor reactions.

• The relative trading intensity between foreign institutions and mutual funds predict next-period stock returns.

Abstract

Using a comprehensive dataset that distinguishes the buy/sell volume of four investor types, we find that foreign institutions and domestic mutual funds are the primary users of analyst reports. Their buy–sell imbalances move in tandem with analysts' signals and significantly explain the size of cumulative abnormal returns across incidents of analyst report releases. Proprietary traders' buy/sell positions are less related to analyst opinions, and individual investors emerge as de facto liquidity providers to aggressive institutions. Institutional investors respond first to stock recommendations and then use target prices as supplementary information. Earnings forecasts, which contain nuanced information, do not elicit abnormal trade imbalances for any institutional investors. We further discover that trade reactions to target prices and earnings forecasts can be viewed more as a constant multiplier of the signal rather than as an increasing or decreasing function of the signal strength. The trade imbalance caused by analyst information is found to predict next-period stock returns, although the predictive ability is for the short-term period only.

Introduction

Securities analysts play a critical role in price discovery in the capital market. They provide three main quantitative outputs in their reports: stock recommendations, target forecasts, and earnings forecasts. It has been well documented that analyst reports with revised forecasts elicit substantial price reactions.² However, analyst information affects stock prices only through investors' reactions, which is evident in their trading. Investors who receive analyst messages interpret the information in their own unique manner and then make trading decisions. Knowing who trades on what information in analyst reports is crucial to understanding the process of price discovery provided by securities analysts.

In this paper, we examine whose trading responds to sell-side analyst reports and what information in the reports is influential to investors' trading decisions. This paper extends the recent literature concerning trader activity based on analyst-provided information. In previous studies, investors are typically separated into large and small investors according to their trade size, and whether large and small investors react to information differently is examined.³ By leveraging a comprehensive dataset related to the Taiwan stock market, we are able to calculate the buy–sell imbalance of four separate classes of investors: foreign institutions, domestic mutual funds, proprietary traders, and individual investors. Analyzing investor trades occurring around the release of analyst reports can help to identify the role(s)—early-informed agent, late follower, or liquidity provider—of each type of investor.

Identifying who trades during the information-release event is an initial step to understanding the process of price discovery. The next step is to determine to which signals in analyst reports does each type of investor react. Relevant studies tend to focus specifically on recommendation revisions, attributing the trade impact of analyst reports solely to the changes in recommendations and without controlling for other signals in the reports. Evidence strongly suggests that recommendations, when revised, often elicit substantial price changes; nevertheless, the additional information conveyed by target prices and earnings forecasts is important, especially when analysts reiterate their previous recommendations.⁴ Even in a report with an updated recommendation, the presence of a target price and earnings forecast may enhance the resolution of analyst information because of the detailed forecasting of the stock's future performance. To completely understand price discovery provided by analyst reports, it is necessary to investigate investor reactions to all three signals in the reports.

This paper extends the relevant literature by examining the trade responses of four classes of investors to the full matrix of analysts' outputs, namely recommendations, target prices, and earnings forecasts. Employing a finer investor classification and the complete set of analysts' signals enables investigations into how various types of investors react to the three key analyst messages and whether these reactions differ. Specifically, we pursue some research questions previously unanswered: Does each type of investor have a unique preference in responding to specific elements in analyst reports? Does an investor's choice of signals depend on their ability to interpret the information or the trading restrictions to which they are subject to? Which elements in analyst reports serve the price discovery function when analyst stock recommendation is not revised? Do investors increasingly respond to a strong signal when other signals in the same analyst report convey weak information? and How do investors react if inconsistent messages are given by these signals? The answers to these questions can help enrich the understanding of the relationship between price discovery and analyst information, which in turn can enhance the effectiveness of the information provided.

In this study, we use the buy–sell trade imbalance of separate investor groups to determine which investors actively trade based on analyst information. Evidence suggests that foreign institutions and domestic mutual funds are the primary investors who aggressively react to analyst reports in Taiwan. They both exhibit strongly abnormal buy–sell imbalances in the direction indicated by the analyst forecasts around the days on which analyst reports are released, with trade imbalances closely reflecting the patterns of stock returns over the course of several event windows. Moreover, the magnitude of return impacts across the sample events can be significantly explained by the strength of the trade responses of foreign institutions and mutual funds. Proprietary traders, however, exhibit no abnormal buying (selling) behavior upon the release of optimistic (pessimistic) analyst opinions. This finding is consistent with our prior knowledge that proprietary traders in Taiwan tend to employ various in-house strategies not directly driven by analyst information. Individual investors trade in the opposite direction to analysts' suggestions, acting as de facto liquidity providers to fulfill the transaction demand of the better-informed institutional investors when analyst reports are released. Our results reveal that adverse selection is inevitably associated with individuals' information disadvantage, echoing the finding of Barber et al. (2008) that individuals tend to lose by trading against informed traders.

We further analyze the types of messages that are considered to be valuable by institutional investors and that are the information basis for their trading. For each investor type, we empirically assess the trade imbalance caused by analysts' recommendations, target forecasts, and earnings forecasts. We find that among the three signals provided in an analyst report, recommendation revisions prompt the largest trade imbalance, which is entirely attributable to trading by foreign institutions and domestic mutual funds. Target prices are also influential; however, they serve a secondary information role, stimulating smaller trade reactions. Earnings forecasts, which require the application of some valuation models to be transformed into an investment decision, have little impact on the trades of any type of institutional investor.

Investors react more intensely to a recommendation if it is the only strong signal in the report. For analyst reports that reiterate a recommendation, investors react strongly to target prices but not earnings forecasts. It appears that the market relies solely on target prices for price discovery when analysts reiterate their previous recommendations. Stronger target signals, however, do not elicit disproportionally larger investor reactions. Investors respond to target signals in a consistent manner, such that their trade imbalance can be viewed more as a constant multiplier of analyst signals rather than as an increasing or decreasing function of the strength of analyst signals. This is true regardless of the strength of the other two signals in the same analyst report. The results indicate that each type of signal has a unique information role and may not be a substitute for the others.

If analyst reports contain contradictory information concerning the recommendation and target price, investors react primarily to the recommendations. Nevertheless, target prices continue to be influential, such that a recommendation revision induces a statistically larger trade imbalance if supportive information concerning the target forecast is provided rather than if contradictory information concerning the target price is provided.

Our results suggest that both foreign institutions and mutual funds in Taiwan prefer eye-catching and easy-to-interpret information in analyst reports (i.e., recommendations and target prices); thus, they are unlike the large traders in the United States, who react to all types of analyst messages including earnings forecasts (Malmendier and Shanthikumar, 2014). The difference in information usage can be attributed to the degree of information competition in the emerging market. Because the major volume contributors in Taiwan, namely individual investors, are not privy to analyst information, informed institutional investors can therefore enjoy the privilege of reacting to the most significant signals while ignoring the messages containing nuanced information. If individual investors receive the same analyst information and switch from being liquidity providers to information competitors, institutional investors would likely begin to “read between the lines” and actively respond to earnings forecasts.

We further show that the trade imbalance of informed traders during the event window can provide important information on returns in the subsequent week. Specifically, the difference in the trade imbalance between mutual funds and foreign institutions during the 5-day event windows (−6, −2) and (−1, 3) can be used to effectively predict the direction and magnitude of stock returns in the following week. This finding indicates the formation of a possible trading strategy by late-informed investors who observe the transactions of early-informed traders.

Our paper contributes to the literature in two ways. First, this study extends the recent work of Kadan et al. (2018), which is the first to reveal the trading strategies employed by early-informed investors during the period in which analysts' signals are released. Our analysis differs from theirs in that their sample contains only the occasions when analysts revise recommendations, whereas we include additional occurrences when analysts reiterate recommendations while continuing to provide revised target prices and/or earnings forecasts. Because only a small proportion of analyst reports provide revised recommendations, the additional information concerning target prices and earnings forecasts is the key to understanding the price discovery function of analyst reports most of the time. The inclusion of a wider set of analyst signals enables us to study the relative importance of these key analyst signals and the reasons why some information is used but some is neglected.

Second, we provide rare evidence of information usage and competition in an emerging market with a highly different investor composition from that of a developed market. Unlike the US stock market, where institutional investors own/trade the majority of shares, the Taiwanese stock market is characterized by a large number of individual investors, and their trading accounts for more than 70% of the trading volume; their dominant participation in the market results in an annual turnover ratio of over 300% (Barber et al., 2008). This characteristic is shared by many emerging markets such as those in China (Tan et al., 2008), South Korea (Kim et al., 2013), and India (Lao and Singh, 2011). To date, no study has detailed the trade responses to the information released by analysts in any emerging market. We present evidence supporting that investor composition may dramatically affect the manner in which analyst information is reflected in market prices. The analysis provides the first evidence of the price discovery function of analyst reports in a typical emerging market.

The remainder of the paper is organized as follows. In Section 2, we present a literature review and the institutional background of the Taiwanese stock market. Section 3 provides a description of our methodologies and sample. In Section 4, we then present our main results and specifically detail how each type of investor responds to analyst reports and each quantitative element in them. Section 4 also presents an investigation into how investors' trades are affected by the relative strength of analyst signals and by the consistency between signals. Subsequently, an explanation is provided for the ability to use the investors' event-window trade imbalance to predict abnormal stock returns in subsequent periods. Section 5 provides the conclusion of the paper.