Do international acquisitions by emerging-economy firms create shareholder value? The case of Indian firms

Abstract

While overseas acquisitions by emerging-economy firms are gaining increased attention from the business press, our understanding of whether and why this inorganic mode of international expansion creates value to acquirer firms is limited. We argue that international acquisitions facilitate internalization of tangible and intangible resources that are both difficult to trade through market transactions and take time to develop internally, thus constituting an important strategic lever of value creation for emerging-economy firms. Furthermore, the magnitude of value created will be higher when the target firms are located in advanced economic and institutional environments: country markets that carry the promise of higher quality of resources, and therefore, stronger complementarity to the existing capabilities of emerging-economy firms. An event study of 425 cross-border acquisitions by Indian firms during 2000–2007 supports our predictions.

APPENDIX

Event Study Methodology

The impact of an economic event such as an acquisition can be assessed by observing the price change in the acquirer's security over a relatively short period through a technique known as event study methodology (Haleblian et al., 2006; MacKinlay, 1997). The underlying assumption is that the market is efficient, and therefore, all information related to the firm and its expected future performance is incorporated in its security or stock price (Binder, 1998). The methodology involves three primary steps (Brown & Warner, 1985; McWilliams and Siegel, 1997):

• 1) Identify the event and define the event window and the estimation period.

• 2) Determine abnormal returns.

• 3) Cumulate the abnormal returns over the event window and test for their significance.

According to McWilliams and Siegel (1997) event study makes three fundamental assumptions:

• 1) The market is efficient.

• 2) The events are unanticipated.

• 3) There are no other, confounding events.

Given that the first assumption may not be totally tenable in emerging economies such as India, the effect of inefficiency can be minimized by selecting a fairly long event window. However, too long an event window can also be problematic in terms of reducing the statistical power of the test and exacerbating the difficulty of controlling for confounding events (McWilliams & Siegel, 1997). Also, long-event windows increase the likelihood of contemporaneous and intertemporal correlations of residuals resulting in significant underestimates of standard errors (Salinger, 1992). In the past, studies have employed various lengths of event window, ranging from as low as 3 days (announcement day ±1 day) to as high as 181 days (announcement day ±90 days) (McWilliams & Siegel, 1997). In this study we employ a moderate event window of 11 days (announcement day ±5 days), and a preceding estimation period of clear 240 days which excludes the event window so as to not contaminate the normal performance model parameter estimates (MacKinlay, 1997).

Appraisal of the event's impact requires a measure of the abnormal return (MacKinlay, 1997), calculated as the difference between the stock market return associated with a given event involving a firm (i.e., actual return) and the firm's historical return (i.e., normal returns) (Merchant & Schendel, 2000). Here, the normal return is defined as that expected if the event did not take place, and is measured by the return obtained with the market model (Capron & Pistre, 2002). Following the procedure outlined by Brown and Warner (1985) and others, we calculate the daily abnormal returns (AR _it ) to the shareholders of acquiring firm i for day t by controlling for both the market return and the firm-specific return as per the model below,

where R _it is the observed firm i's return and R _it is the return on a market index, both being for day t. In the above equation the term within the brackets is the expected share price normal return, and is calculated by regressing daily share price return on a daily market portfolio (index) return over a predetermined estimation period preceding the event (e.g., 250–50 days prior to the event). The corresponding constant and the coefficient obtained from the above regression are α _i and β _i , respectively.

We used SENSEX, the benchmark index of the Bombay Stock Exchange, to assess the fluctuation in daily share price of listed and publicly traded acquirer firms. As a cross-check, we used other indices, such as NIFTY, BSE-500, and S&P CNX-500, as alternate benchmarks to assess abnormal fluctuation in the share prices of acquiring firms. The values of expected normal returns obtained across the indices correlated with those obtained using SENSEX. Once the daily abnormal returns around the event have been determined, it is customary to cumulate the abnormal return obtained over the “event window”. This aggregation is undertaken to account for the capital markets’ reaction to announcements that may have been made after trading hours (Merchant & Schendel, 2000), and to account for any information leakage prior to the official announcement of the event. We cumulate the abnormal return over the 11-day window. Finally, to assess whether the event under observation has a significant impact on values of the firms, a suitable test statistic (in our case the average of CAR over the 11-day window for all events) is assessed for statistical significance (McWilliams & Siegel, 1997).