Cao, 1997; Hau, 2001). However, some papers suggest that foreign investors who participate in a market can be better informed than local investors (Seasholes, 2000;
Froot et al., 2001). These studies provide mixed conclusions regarding whether local or foreign investors have information advantage.
In contrast to most studies focus on investors choice and performance, Bae et al.
(2008) directly examined whether analysts resident in a country make more precise earnings forecasts for firms in that country than analysts who are not resident in that country. They found that local analysts have a significant information advantage over foreign analysts for a large sample of countries. The result is the same as Malloy (2005) who witnessed that in the U.S. analysts who are closer to the headquarters of firms have information advantages.
However, Bacmann and Bolliger (2001) investigated the relative performance of local and foreign financial analysts on Latin American emerging markets. They found
that foreign financial analysts outperform local analysts on these markets. Foreign analysts produce more timely and more accurate forecasts. A significant price reaction is observed following their downward forecast revisions. Also, the evidences on the performance of local and foreign analysts are mixed.
Bae et al. (2008) defined an analyst as a “local analyst” if the country location of an analyst is the same as that of the firm he covers, regardless of whether an analyst is working for a local research firm or a research firm from a foreign country. Unlike the analysts definition of Bae et al.(2008)10, this study defines an analyst as a “local (foreign) analyst” if the analyst is working for a local (foreign) firm, regardless of whether the country location of an analyst is the same as that of the firm he reports. It means that local analysts may not in the same country location where they cover, but foreign analysts may in the same country location where they cover. Then, this study discusses forecast accuracy of analysts and optimism in the next part.
2.2 Forecast accuracy of analysts and optimism
A large number of papers employ analysts’ forecasts as a measure of expected earnings and refer to negative mean forecast errors as evidence of the phenomenon of optimism in analysts forecast. Ke and Yu (2006) found that analysts’ forecasts tend to be optimistic at the beginning of the period while pessimistic at the end of the period to produce more accurate earnings forecasts and they are thus less likely to be fired.
Clement (1999) found that forecast accuracy is positively associated with analysts' experience and employer size, and negatively associated with the number of firms and industries followed by the analysts. Duru and Reeb (2002) found greater corporate international diversification associated with less accurate and more optimistic analysts' forecasts. Hope (2003) witnessed strong enforcement of
10 We do not have the data of the country locations of analysts, hence we use the properties of broker- age firms that analysts work for dividing them into local or foreign analysts.
accounting standards and firm-level disclosures are associated with higher forecast accuracy. Loh and Mian (2006) applied absolute forecast errors examining the relation between the accuracy of analysts’ earnings forecasts and the profitability of their stock recommendations. They documented the recommendations of superior earnings forecasters significantly outperform those of inferior ones.
Combining papers above, we find, in order to maintain good relationships with management and avoid being fired, analysts are usually more optimistic at the beginning of the period. However, corporations with less international diversification, larger size, strong enforcement of accounting standards and firm-level disclosures will induce analysts’ earnings forecasts more accurate. More accurate earnings forecasts often accompany superior stock recommendations. Are local analysts or foreign analysts more accurate in earnings forecasts? We will discuss in this chapter.
3. Data and methodology
3.1 Sample description
We use Nikkei 225 Index and MSCI Taiwan Index inclusions to study the earnings forecast changes of analysts over the period through September 1991 to March 2008 and May 1999 to May 2007, respectively. We examine whether their forecasts are different in newly added firms and the matching firms. Our matching firms are listed firms that are not included in the Nikkei 225 Index and MSCI Taiwan Index, which are two benchmark portfolios. Also, whether local analysts and foreign analysts have the same earnings forecasts or not is analyzed. Nikkei 225 Index and MSCI Taiwan Index have undergone composite stocks adjusted for many times in these periods. There are 112 firms and 105 firms included in the Nikkei 225 Index and MSCI Taiwan Index during the sample periods, respectively. The sample excludes newly added firms that are a spin-off or if they engaged in a merger or takeover
around the inclusion event.
In order to compare the differences of earnings forecast changes of analysts for newly added firms and the matched firms, we adopt two benchmarks for the matched firms. First, we include all companies in the I/B/E/S database that we can compute a current/ one-year-ahead median EPS forecast for the same pre-announcement and post- announcement time periods as for the addition firms11.
For the second benchmark, we separate firms into different industry sectors. The log market capitalization is proxied for size. The liquidity is defined as the trading volume dividend by the number of shares outstanding for MSCI Taiwan Index. For Nikkei 225 Index, market liquidity is measured by its trading volume. Each newly added stock is matched with its appropriate industry, size and liquidity12. We regard the first criterion set as “all other firms” while the second criterion as the “industry, size and liquidity (ISL) matched firms”. In Appendix A and B, we exhibit the descriptive statistics of price, market value, turnover rate and volume for Nikkei 225 Index and MSCI Taiwan Index addition firms and the matched firms.
Earnings per share (EPS) forecasts and actual EPS for the current year and one-year-head are obtained from the I/B/E/S database. The price, market value and trading volume of Japanese stocks are obtained from the Datastream. Information on announcement dates for Nikkei 225 Index adjustment is obtained from the Nikkei Interactive website. Price, market value, trading volume and the number of shares
11 The first criterion for our research is consistent with Denis et al. (2003).
12The second criterion for Denis et al. (2003) is that each company in the I/B/E/S database first sorted into 1 of the 12 Fama French industry portfolios. Each industry portfolio is divided into 3 portfolios on the basis of market capitalization, with one-third of the firms in each market-value portfolio. Finally, within each industry and market-value portfolio, firms are sorted into 3 liquidity portfolios, where liquidity is defined as the five-year average of annual trading volume divided by the number of shares outstanding. This sorting procedure results in 108 portfolios. Each newly added stock is matched with its appropriate industry, size, and liquidity portfolio. We have not adopted this classification in this paper due to data limitation.
outstanding data for Taiwanese stocks are obtained from the TEJ (Taiwan Economic Journal). Furthermore, announcement date information for MSCI Taiwan Index adjustment is obtained from the UDN (United Daily News) data.