Geographic Diffusion of Information and Stock Returns

Transcription

1 Geographic Diffusion of Information and Stock Returns Jawad M. Addoum * University of Miami Alok Kumar University of Miami Kelvin Law Tilburg University February 12, 2014 ABSTRACT This study shows that value-relevant information about firms is geographically distributed across U.S. states and the market is slow in aggregating this information. The earnings and cash flow of firms can be predicted using the past performance of firms in economically relevant geographical regions, but sell-side equity analysts and institutional investors do not fully incorporate this information in their earnings forecasts and trades, respectively. Consequently, firms exhibit stronger post-earnings-announcement drift and stronger momentum in returns when geographic information is more dispersed and difficult to aggregate. A Long Short trading strategy that exploits the slow diffusion of geographic information earns an annual, abnormal risk-adjusted return of about 9%. Keywords: Geographical connections, information diffusion, earnings and return predictability, price momentum, post-earnings announcement drift (PEAD), sell-side analysts, institutional investors. * Please address all correspondence to Alok Kumar, Department of Finance, School of Business Administration, 514G Jenkins Building, University of Miami, Coral Gables, FL 33124; Phone: ; akumar@miami.edu. Jawad Addoum can be reached at or jawad.addoum@miami.edu. Kelvin Law can be reached at (+31) or k.f.law@tilburguniversity.edu. We thank Major Coleman, Xavier Giroud, George Korniotis, Chris Malloy, Chris Parsons, Tao Shu and seminar participants at University of Miami, Tilburg University, Syracuse University, and University of Georgia for helpful comments and valuable suggestions. We are responsible for all remaining errors and omissions.

2 Geographic Diffusion of Information and Stock Returns ABSTRACT This study shows that value-relevant information about firms is geographically distributed across U.S. states and the market is slow in aggregating this information. The earnings and cash flow of firms can be predicted using the past performance of firms in economically relevant geographical regions, but sell-side equity analysts and institutional investors do not fully incorporate this information in their earnings forecasts and trades, respectively. Consequently, firms exhibit stronger post-earnings-announcement drift and stronger momentum in returns when geographic information is more dispersed and difficult to aggregate. A Long Short trading strategy that exploits the slow diffusion of geographic information earns an annual, abnormal risk-adjusted return of about 9%. Keywords: Geographical connections, information diffusion, earnings and return predictability, price momentum, post-earnings announcement drift (PEAD), sell-side analysts, institutional investors.

3 1. Introduction A growing literature in finance demonstrates that value-relevant firm-specific information is distributed geographically across the U.S. states (García and Norli (2012), Giroud (2013), Bernile, Kumar, and Sulaeman (2013)). In particular, this literature finds that a typical U.S. firm has geographical presence in six U.S. states and its business activities often occur at locations away from its corporate headquarters. Consequently, the earnings and cash flow of the firm are likely to originate from activities away from its corporate headquarters. Each of the economically-relevant non-headquarters locations may contain bits and pieces of information about the firm that may not be easily accessible and may be difficult to aggregate. Even firm insiders may have some difficulty accessing and aggregating this geographically dispersed information. In a recent study, Giroud (2013) shows that firm managers are able to improve their monitoring and information gathering activities when plant locations become more accessible to them following the introduction of airline routes. If firm-specific value-relevant information is dispersed geographically, it is natural to ask how quickly this information gets aggregated and incorporated into stock prices. Given the prior evidence in the literature, it is unlikely that the traditional firm-level measures are able to fully capture a firm s geographic exposure across the U.S. states. Consequently, market participants may not be able to aggregate the geographically dispersed information immediately. The frictions generated by physical distance between corporate headquarters and centers of economic activities relevant to the firm are likely to slow down the information aggregation process. This delay in the information aggregation process could subsequently generate predictable patterns in stock returns. To better understand our core idea, consider a hypothetical firm AUTO with corporate headquarters located in the state of Texas. While headquartered in Texas, this automobile parts manufacturer may have manufacturing facilities, research and development centers, customers, 1

4 suppliers, distribution channels, and other relevant economic units spread all across the U.S. Thus, the day-to-day economic activities and operations of the firm are likely to be located in states other than its corporate headquarters. Since firms located in the same geographical regions are likely to experience common local shocks, the future performance of AUTO would be correlated with the performance of other firms located near its corporate headquarters (HQ) and economically connected (EC) states. If value-relevant information from these economically relevant geographical locations is aggregated and incorporated into prices with some delay, the earnings, cash flow, and returns of AUTO may be predictable based on the information contained in these geographic links. In this paper, we generalize this idea and investigate whether the geographical connections of firms contain information that can be used to predict their future fundamentals. In addition, we examine the potential asset-pricing implications of slow diffusion and delayed aggregation of geographically distributed firm-specific information. We also attempt to identify the channels that can potentially impede the diffusion of geographic information. Specifically, we examine how behavior of two key groups of market participants (sell-side equity analysts and institutional investors) impacts the information diffusion process. To measure the economic interest of a firm in a given U.S. state, we follow García and Norli (2012) and conduct textual analysis of firms annual 10-K filings. Specifically, we capture the relative importance of a state to a firm by measuring the annual citation share of each U.S. state, which is defined as the number of times a U.S. state is mentioned in an annual 10-K filing divided by the total number of times all U.S. states are mentioned in the filing. A state with firm-level citation share of one (zero) indicates that the firm has all (none) of its economic activities located in that state. Using these firm-state-year citation share estimates of economically connected states, we measure the geographical dispersion in a firm s economic interests and estimate a series of Fama and MacBeth (1973) type predictive regressions. Specifically, we construct EC Earnings (EC Cash Flow) 2

5 measures, defined as the citation-share weighted average earnings (cash flow) of firms located in economically relevant states, excluding the headquarters state. Separately, we also measure the earnings and cash flow of other firms located around firm headquarters. 1 Using the aggregated HQ and EC earnings and cash flow measures, we find that the performance of other firms located in both HQ and EC states contains value-relevant information for the future earnings and cash flow of a given firm, even after we account for the lagged performance and fundamentals of the firm. Specifically, the one-quarter-ahead earnings and cash flow of a given firm are predictable based on the performance of other firms located around corporate headquarters and in regions that are economically relevant for the firm. Further, we demonstrate that the evidence of predictability using information in EC states is twice as strong as the predictive power of HQ-based measures. The information in EC states is valuable even when we exclude firms that are headquartered in EC states and only aggregate information from firms that have economic activities in those EC states. Put differently, the earnings of a given firm i can be predicted using the lagged earnings of other firms that are economically active in states where the economic activities of firm i are concentrated. For example, Dell is a firm headquartered in Texas but has economic activities in California and Florida. Other firms such as Coke and Staples also have economic activities in those two states. Our EC-based earnings predictability results show that the current earnings of Coke and Staples contain valuable information about the future earnings of Dell. 2 Our evidence of earnings and cash flow predictability is incremental over the predictive ability of firm-specific lagged earnings measures (e.g., Fama and French (2000), Vuolteenaho (2002)) 1 Consider our hypothetical firm AUTO that is headquartered in Texas but has economic activities in Florida and California. Suppose, in a given year, it cites Texas, Florida, and California 12, 3, and 7 times, respectively. We assign a citation share weight of 3/10 to Florida and 7/10 to California. The citations to the headquarters state of Texas are excluded from this calculation. See Section 3.2 for additional details. 2 See footnote 7 for additional details about the construction of our EC-based earnings measure when we exclude firms that are headquartered in economically connected states. 3

6 and known market-wide return predictors (e.g., Fama and Schwert (1977), Campbell and Shiller (1988), Fama and French (1988)). We also demonstrate that our results do not merely reflect the direct impact of state-level macroeconomic variables or firm characteristics, as all regressions include controls for state-level macroeconomic variables and firm fundamentals. In the next set of tests, we examine how quickly geographically-dispersed information is aggregated by market participants and subsequently reflected in stock prices. We first examine whether sell-side analysts help incorporate firms geographically dispersed information through their earnings forecasts. Our evidence indicates that equity analysts do not fully incorporate firms geographically dispersed earnings-relevant information into their forecasts. Instead of using dispersed geographical information, they form their forecasts primarily based on lagged firm earnings. As a result, their earnings forecasts do not fully account for the earnings information of other firms headquartered in economically connected states. Specifically, the lagged EC-based earnings measure significantly predicts analysts future forecast errors. A one standard deviation increase in lagged EC Earnings corresponds to a 5.251% increase in analysts consensus forecast error in the next quarter. Further, we find that the explanatory power of EC-based earnings is higher than the lagged earnings or HQ-based earnings measures. Since equity analysts are an important part of the price formation process and geographically dispersed information is not fully incorporated into their earnings forecasts, it is likely that geographic information diffuses slowly and generates predictable patterns in stock returns. We examine the effects of slower diffusion of information in two economic settings: post-earnings announcement drift (PEAD) and momentum in stock returns. Our choice is motivated by the evidence in previous studies, which suggest that slow diffusion of information may be a key mechanism that generates predictable patterns in stock returns in these two economic settings. 4

7 We find that both post-earnings-announcement drift (PEAD) and momentum returns are more pronounced among firms that have more geographically dispersed information that is difficult to aggregate. In both instances, we find that the slow diffusion of geographically dispersed information generates predictable patterns in stock returns. In economic terms, our estimates indicate that firms with above-median geographic dispersion have monthly momentum returns that are about 0.35% higher (t-statistic = 2.78) than firms with below-median dispersion. Similarly, we find that a one standard deviation increase in geographic dispersion is associated with 0.20% to 0.35% (t-statistics are between 2.37 and 2.72) higher post-earnings-announcement returns. This evidence is consistent with our broad conjecture and shows that market participants incorporate geographically dispersed information in their decisions with some delay. To quantify the economic magnitudes of these predictable return patterns, we construct outof-sample geography-based trading strategies. We use look-ahead bias-free forecasts of firm-level earnings-per-share (EPS) and sort stocks based on their Expected Earnings Surprise, which is defined as the forecasted EPS minus the analyst consensus. A Long Short trading strategy that takes a Long (Short) position in firms with high (low) expected earnings surprise generates a monthly alpha of 75 basis points or an annual premium of about 9%. The performance of our trading strategy is robust and cannot be explained by standard asset pricing factors such as the market, size, value, momentum, short- and long-term reversal, and liquidity. We also analyze the sensitivity of the trading performance estimates when there is a delay between portfolio formation date and the start of the portfolio performance measurement period. We find that the performance of our trading strategy deteriorates as the gap between the portfolio formation date and the start of the portfolio performance measurement period widens. Consistent with the gradual information diffusion model of Hong and Stein (1999), we find that it takes 5

8 approximately three months for geographically dispersed information to be eventually incorporated into stock prices. To provide additional evidence of our slow information diffusion hypothesis, we directly analyze the impact of financial intermediaries on the speed of information diffusion. Consistent with the findings in Bernile, Kumar, and Sulaeman (2013), we find that institutional investors in economically connected states are able to extract local information more effectively and speed up the information diffusion process. Consequently, the performance of our trading strategy is economically small for the subsample of firms with high concentration of institutions in states where those firms are economically active. In contrast, we find economically significant trading strategy performance for subsamples of firms with high as well as low coverage by analysts located in economically connected states. This evidence suggests that sell-side analysts are relatively less skillful in extracting value-relevant information from various geographical locations. In the last part of the paper, we investigate whether there are frictions in the market that could slow down the information aggregation process even after the earnings information dispersed in various economically connected states is reflected in the prices and returns of firms located in those states. We find that even when geographically dispersed information about firm fundamentals gets aggregated and incorporated into stock prices, it does not get incorporated into the prices of other connected firms immediately. This delay in the information diffusion process generates predictable patterns in stock returns, which can potentially be exploited. Trading strategies that exploit this predictability pattern earn annualized risk-adjusted returns of 5-7%. Collectively, our results provide strong support for our broad conjecture, which posits that geographically dispersed information diffuses slowly and generates predictable patterns in stock fundamentals and returns. These findings contribute to several strands of the finance literature. First, we contribute to the growing literature in finance that exploits the geographical links among firms 6

9 (García and Norli (2012), Giroud (2013), Bernile, Kumar, and Sulaeman (2013)). Specifically, we show that earnings of geographically related firms within the U.S. have the ability to predict the future performance of a given firm. To our knowledge, the evidence of earnings and cash flow predictability using geographically dispersed information is new to this literature. Second, we present novel evidence of predictable patterns in stock returns and contribute to the broad literature on return predictability and market efficiency. Specifically, we demonstrate that market participants incorporate geographically dispersed information about firms with a delay and this slow information diffusion generates predictable patterns in returns. Even relatively more sophisticated market participants such as equity analysts do not successfully incorporate firms geographical information in their forecasts. As a result, earnings surprises of firms are predictable. In addition, firms exhibit stronger post-earnings-announcement drift and stronger momentum in returns when geographic information is more dispersed and more difficult to aggregate. Third, our results contribute to the literature that examines the impact of slow diffusion of information on stock returns. For example, Hong, Lim, and Stein (2000) show that slow diffusion of firm-specific bad news is an important determinant of momentum in stock returns. In another context, Hou and Moskowitz (2005) use a price delay proxy to demonstrate that various types of market frictions slow down the information diffusion process and generate cross-sectional variation in average stock returns. Similarly, Hou (2007) shows that slow diffusion of industry information generates a lead-lag relation between the returns of small and large firms. More recently, Cohen, Diether, and Malloy (2013) show that stock market participants are unable to value the benefits of innovations immediately and, consequently, the returns of innovative firms are predictable. We add a geographic dimension to the information diffusion process and demonstrate that geographical distribution of economic activities of firms within the U.S. slows down the information aggregation process. 7

10 Using Compustat segment data, recent studies show that value-relevant information from international locations are incorporated into stock prices with a delay (Nguyen (2012), Huang (2013)). Our results complement these earlier findings and indicate that pieces of information generated even within the U.S. are not immediately incorporated into stock prices, which generates predictable patterns in stock returns. This evidence of within-country geography-based informational friction is perhaps more surprising since information barriers within the U.S. are likely to be small. Our study also relates to prior studies, which demonstrate that local investors and equity analysts exhibit local bias and may possess an informational advantage. 3 Our findings suggest that a significant part of local information still remains unexploited. Further, our paper is related to recent studies that examine economic links among firms that emerge through customer-supplier networks (e.g., Cohen and Frazzini (2008), Menzly and Ozbas (2010)). Our evidence suggests that common exposures to geographical regions even within the U.S. generate economic links among firms. More broadly, our findings complement the evidence from recent studies that examine the importance of locational factors on corporate behavior. In particular, Dougal, Parsons, and Titman (2013) investigate the impact of time-varying locational factors (i.e., urban vibrancy ) around corporate headquarters on corporate investments. While we also emphasize the importance of local factors on firm performance and stock returns, unlike their study, we focus on local factors away from corporate headquarters. We show that those local factors useful contain information about the future earnings and cash flow of firms. The rest of the paper is organized as follows. Section 2 provides a summary of the various data sources used in the empirical analysis. Section 3 presents evidence of earnings and cash flow predictability using geographic connections. Sections 4 and 5 examine whether market participants 3 For example, see Coval and Moskowitz (1999, 2001), Huberman (2001), Hau (2001), Grinblatt and Keloharju (2001), Malloy (2005), Ivković, Sialm, and Weisbenner (2008), Teo (2009), van Nieuwerburgh and Veldkamp (2005, 2009). 8

11 are able to quickly aggregate this geographically dispersed information into stock prices. We conclude in Section 6 with a brief discussion. 2. Data and Methods In this section, we provide details about our data sources and present summary statistics for the main variables used in the empirical analysis. 2.1 Measuring Economic Connections We identify the locations of economic activities and corporate headquarters of all U.S. publicly listed firms by retrieving their annual Form 10-K filings from the Electronic Data Gathering, Analysis, and Retrieval (EDGAR) system. Our sample period is from 1994 to To comply with listing disclosure requirements, all U.S.-based firms are required to file Form 10-K with the U.S. Securities and Exchange Commission (SEC) annually. Each 10-K has a prescribed reporting structure and disclosure items containing a comprehensive overview of a firm s business operations, summary of financial conditions, and audited financial statements. Apart from containing the operational details relating to a firm s business activities (e.g., organizational structure, executive compensation, industry competition, auditor s reports, and regulatory issues etc.), it also lists detailed information regarding physical locations of a firm s assets (e.g., factories, warehouses, distribution centers, and sales offices, etc.), segment reporting, and management discussions and analysis of financial condition or results of operations. 4 We measure the geographic exposure of a firm to each U.S. state by conducting a textual analysis of its annual 10-K filings (García and Norli (2012), Bernile, Kumar, and Sulaeman (2013)). For each firm-year, we parse the 10-K filings and count the number of times references are made to 4 See for additional details about the information reported in Form 10-K. Also, see Appendices C and D for examples and direct excerpts from 10-K filings, respectively. 9

12 each of the 50 U.S. states and Washington D.C. Specifically, we focus on the occurrence of geographic references in the following four specific sections: (a) Item 1: Business, (b) Item 2: Properties, (c) Item 6: Consolidated Financial Data, and (d) Item 7: Management s Discussion and Analysis. As these four sections summarize the locality of a firm s main business operations including a firm s plant and equipment, major physical assets, store locations, office locations, and acquisition activities the citation count measure captures the economic ties between a firm and its geographically distributed economic interests. While our textual analysis based measure of economic connections may contain measurement noise, it has several advantages. First, instead of relying on disclosure in firm s filings (e.g., geographical segment reporting), our method greatly expands to a broader sample. Moreover, as some of the voluntary disclosures may be subject to management s discretion or interpretation (e.g., use of geographical regions in segment reporting, defining materiality in Exhibit 21), content analysis is less prone to this potential measurement bias. Specifically, firm managers have a great discretion in deciding a reportable geographical segment (under accounting standards SFAS 131 or IFRS 8), and have an incentive to pool geographical segments to avoid disclosing commercially valuable information to competitors that is unavailable elsewhere (Harris (1998)). After obtaining the total count of each state s mentions in the 10-K filing, we compute a citation share (CS) for each state in the filing. It is defined as the number of times a U.S. state is cited divided by the total number of citations of all U.S. states in the firm s 10-K filing in year t. The maximum (minimum) of citation share is one (zero), where a higher (lower) citation share implies that a given firm s economic activities are more (less) concentrated in a given state. 2.2 State-Level Measures of Economic Activity We use the Korniotis and Kumar (2013) method to compute the economic activity index for each U.S. state. Specifically, we define the Economic Activities Index as a summary index of state-level 10

13 macroeconomic conditions in a state. We consider three state-level economic indicators that are likely to capture business cycle variation at the local state-level: 1) income growth, 2) unemployment rate, and 3) housing collateral ratio. All these state-level macroeconomic data are available at the quarterly frequency. The Economic Activities Index is defined as the sum of standardized values of state-level income growth and state-level housing collateral ratio, minus the standardized value of the relative state-level unemployment ratio, divided by three. In this definition, the state-level income growth is the log difference between state income in a given quarter and state income in the same quarter in the previous year. This measure is motivated by prior studies that interpret this measure as a proxy for the return to human capital (e.g., Jagannathan and Wang (1996), Campbell (1996)). A high-level of income growth reflects positive macroeconomic conditions in a state. State-level unemployment rate is the ratio of the current unemployment rate to the moving average of the unemployment rates from the previous 16 quarters. It is a recession indicator for the state economy. The moving average serves as a proxy for the expected level of unemployment and a positive (negative) deviation from this projected or expected level signals a good (bad) signal for the local economy. The unemployment rates are obtained from the Bureau of Labor Statistics (BLS). State-level housing collateral ratio is the log ratio of state-level housing equity to state labor income following Lustig and van Nieuwerburgh (2005, 2009). It is computed using the Lustig and van Nieuwerburgh (2005) method, where the data are obtained from Stijn van Nieuwerburgh s website. This measure captures the tightness of borrowing constraints and the degree of risk sharing at the local state level. A high housing collateral ratio predicts high consumption growth at the statelevel, as the consumption of individuals is better insulated against adverse labor income shocks. 11

14 2.3 Other Datasets We use the Central Index Key (CIK) to merge firms 10-K filings data with firm fundamentals from Compustat. CIK is assigned by the SEC to uniquely identify registered firms for meeting disclosure requirements. After merging, we retrieve firms headquarter location from the CRSP-Compustat Merged (CCM) file. The fundamentals of firms with at least $10 million of average total assets, two years of data, and $1 closing price are obtained from Compustat. All firm-year observations with SIC codes in the ranges of (utility firms) and (financial institutions) are excluded. Corresponding price and return data are then obtained from the Center for Research on Security Prices (CRSP). Monthly data on the risk-free rate, market excess return (MKT), size factor (SMB), value factor (HML), and momentum factor (UMD), short-term reversal (STR), and long-term reversal (LTR) factors are obtained from Ken French s website. To define analyst forecasts, we obtain split-adjusted earnings per share (EPS) forecasts from Thomson-Reuters IBES unadjusted detail file. To identity analysts geographical location, we use the names and brokerage house information in the Broker Translation File to hand-match records in Nelsons Directory of Investment Research (Malloy (2005)). 3. Earnings and Cash Flow Predictability We begin our empirical analysis by estimating a series of predictive regressions. Specifically, we investigate whether the earnings of a firm can be predicted using the earnings information of other firms headquartered in the same state where the firm is headquartered or has an economic presence. Our conjecture is that the earnings information of other firms located in the same geographical area would contain information about a firm s future earnings. Thus, if a firm s economic activities are geographically dispersed, then the lagged performance of firms with 12

15 economic presence in those economically connected states may contain valuable information about the future performance of the firm. We test this conjecture using data on both earnings and cash flow. Specifically, we investigate whether the earnings and cash flow of other firms headquartered in EC states have any incremental power to predict a given firm s future earnings and cash flow. We also compare the predictive ability of EC based measures and HQ based earnings measures obtained using firms around corporate headquarters. 3.1 Predictability Using Information Around Corporate Headquarters Our first test examines whether a firm s earnings and cash flow can be predicted using the financial performance of other firms headquartered in the same state. Specifically, we estimate the following predictive Fama and MacBeth (1973) style regression:,,,,, (1) At the end of each quarter, we regress firm j s Earnings in quarter t+1 on an intercept, lagged Earnings in quarter t, lagged HQ Earnings in quarter t, and a vector of control variables (X), which includes firm-specific characteristics and measures of the HQ state s economic environment in quarter t. Earnings is defined as operating income after depreciation divided by average total assets (Richardson, Sloan, Soliman, and Tuna (2005)). The lagged Earnings is included in the specification as prior studies show that firm s earnings is persistent (Fama and French (2000), Dichev and Tang (2009), Frankel and Litov (2009)). HQ Earnings is the value-weighted Earnings of firms headquartered in the same state as firm j. β 2 and β 3 jointly measure the one-quarter-ahead predictability of firm earnings, where the main coefficient of interest is β 2. 13

16 Our main conjecture is that HQ Earnings in quarter t would contain incremental explanatory power beyond the predictive ability of firm-specific lagged Earnings in quarter t. A positive coefficient estimate would indicate that the earnings information of other firms headquartered in the same state can be used to predict a given firm s earnings in quarter t+1. As the error terms are likely to be auto-correlated in the Fama and MacBeth (1973) regressions, all standard errors are adjusted using the Newey and West (1987) method with 4-year lag (i.e., 16 quarters). To minimize the noise in our measures, we require at least two other firms when we compute the HQ-based earnings measures. X is a vector of control variables, which includes the following firm attributes: 1) Firm Size, 2) Leverage, 3) Loss, 4) Market-to-Book, 5) Dividend Yield, 6) No-Dividend Indicator, 7) Dividend-Price, and 8) Economic Activities Index. Firm Size is the natural logarithm of total assets. Leverage is the sum of shortterm and long-term debts, divided by total assets. Loss (No-Dividend Indicator) is a dummy that takes a value of one when operating income (dividend) is negative (zero), and zero otherwise. Market-to-Book is the sum of market equity, short-term debt, and long-term debt, divided by total assets. Dividend Yield is the dividends divided by shareholders equity. Together, these control variables should account for differences in firm size, growth opportunities, operations, and profitability, which could affect firm earnings. Additional details about all variables are available in Appendix A. Beyond these firm attributes, we include two variables that capture the differences in statelevel business cycles across headquarter states. The first control variable is Dividend-Price, defined as the value-weighted average of the log of one plus the dividend-price (D/P) ratio of firms headquartered in the same state. Dividend D is the sum of the past four quarterly dividends, whereas P is the end-of-month stock price. The monthly stock prices are obtained from CRSP, and the quarterly stock-level dividends are obtained from Compustat. We also include the Economic Activities 14

17 Index in the specification, which is a summary index of state-level macroeconomic conditions in a firm s headquarters state. For our second set of baseline tests, we use a similar specification as equation (1) to examine the predictability of one-quarter-ahead cash flow. Specifically, we replace the dependent variable Earnings with Cash Flow, and estimate the following specification:,,,,, (2) In this equation, Cash Flow is the cash flow from operating activities divided by average total assets. Similar to the Earnings specification, the main coefficient of interest is β 2. A positive coefficient would indicate that the cash flow of other firms situated in the same state predicts a given firm s future cash flow. Our choice of firm fundamentals (Earnings and Cash Flow) is motivated by prior studies that widely study these metrics as proxies of financial performance of firms (e.g., Sloan (1996), Fama and French (2000)). The primary difference between these two measures is that the former captures the accrual component of a firm s earnings, which typically includes future or deferred cash flow, depreciation, and allowances. In contrast, the latter measures the actual cash flow component of earnings, and depends on the actual timing of the earnings realization. We present the summary statistics of all variables in Panel A of Table I. The mean Earnings (Cash Flow) is (0.017), which indicates that the average firm earnings (cash flow) is 0.9% (1.7%) of total assets each quarter. These estimates are consistent with prior studies finding that cash flow from operations is on average higher than earnings (e.g., Dichev and Tang (2009)). We also find that firm-specific earnings (Earnings) are more volatile than aggregate state-level earnings (HQ Earnings), as the standard deviation of Earnings is about 3.5 times the standard deviation of HQ Earnings. The 15

18 same pattern is observed with the Cash Flow measure, as the aggregate state-level cash flow (HQ Cash Flow) is less volatile than firm-specific cash flow (Cash Flow). Panel B of Table I reports the correlations among these measures. We find that the aggregate state-level earnings/cash flow (i.e., HQ Earnings/HQ Cash Flow) is strongly and positively correlated with firms earnings/cash flow (i.e., Earnings/Cash Flow) at the 1% level. Table II reports the estimates from the earnings and cash flow regression specifications. Specifically, columns 1 and 4 report the univariate regression results, columns 2 and 5 report the results from regressing firm s Earnings/Cash Flow on lagged Earnings/Cash Flow and lagged HQ Earnings/Cash Flow, while columns 3 and 6 report the estimates from the full specification that includes all control variables. First, consistent with prior evidence, we find that firms earnings and cash flow are persistent, as the estimated coefficients on lagged firm-specific Earnings and Cash Flow are positive and statistically significant. The average R 2 is 0.680, confirming the findings in prior literature that the firm-specific earnings process is fairly persistent. The economic magnitudes of lagged Earnings and Cash Flow coefficient estimates are also similar to the findings in prior studies. 5 Second, we find that controlling for firms lagged performance, the earnings and cash information of other firms headquartered in the same state predicts the firm s future financial performance. Specifically, the estimated coefficients on HQ Earnings and HQ Cash Flow are positive and statistically significant, ranging from (t-statistic = 2.87) to (t-statistic = 3.58). Examining the economic magnitudes of these estimates, we find that a one standard deviation change in HQ Earnings (HQ Cash Flow) leads to a 0.075% (0.197%) change in Earnings (Cash Flow) in quarter t+1. Relative to average quarterly earnings (cash flow) of 0.9% (1.7%), these magnitudes are economically meaningful. 5 For instance, see Richard, Sloan, Soliman, and Tuna (2005), Dichev and Tang (2009), and Frankel and Litov (2009). 16

19 Overall, the estimates from earnings and cash flow regression specifications indicate that the information contained in the financial performance of other firms headquartered in the same state predicts a given firm s earnings and cash flow. Since we estimate Fama and MacBeth (1973) type regressions, these results do not reflect the effects of broad macroeconomic variables on firm earnings and cash flow. 3.2 Predictability using Information From Economically Relevant Regions Next, we examine whether the fundamental information in a firm s economically connected states has the ability to predict its future performance, incremental over the ability of firm fundamentals of other firms located in the same state. To test this conjecture, we re-estimate our baseline regressions with three additional independent variables using the following specification:,,,,,, (3) Similar to HQ Earnings, EC Earnings is the citation-share weighted Earnings of firms located in economically connected states, excluding the firms in the HQ state. As before, all EC-based measures require a minimum of two firms. The following example illustrates the intuition behind our identification strategy: consider again our hypothetical automobile firm AUTO that is headquartered in Texas but 2/3 of its economic activities is concentrated in California and 1/3 of its activities are in Florida. EC Earnings is the sum of a) 2/3 of the average earnings of firms that are headquartered in California or have economic presence in that state, and b) 1/3 of the average earnings of all firms with corporate headquarters or economic interests in Florida. We do not include the earnings of firms located in Texas where the firm is headquartered. If EC locations contain value-relevant information about AUTO, EC Earnings should predict the firm s future performance, as captured by the coefficient β 3. A positive β 3 estimate would indicate 17

20 that the earnings information of firms located in a given firm s economically connected states can help predict the firm s earnings next quarter. Beyond the standard firm-level control variables, two additional control variables (EC Dividend-Price and EC Economic Activities Index) are included in the specification to capture the macroeconomic conditions in EC states. 6 These control variables are important, as EC Earnings could potentially capture the overall economic environment of firms economically-relevant states. Including these additional variables helps mitigate this potential concern. Specifically, EC Dividend- Price is constructed following HQ Dividend-Price, defined as the citation-share weighted dividend-price index (Dividend-Price) of all firms located in economically connected states, excluding the HQ state. Similarly, EC Economic Activities Index is the citation-share weighted Economic Activities Index of all firms in economically connected states, with HQ state excluded. We estimate the following predictive cash flow regression specification:,,,,,, (4) Here, EC Cash Flow is the citation-share weighted Cash Flow of firms located in economically connected states, excluding the HQ state. A positive coefficient of β 3 would indicate that the cash flow information of firms in a firm s economically connected states can predict its future cash flow. The summary statistics in Panel A of Table I show that EC Earnings/EC Cash Flow exhibit high and significantly positive correlations with firm-specific Earnings/Cash Flow. EC-measures also have the lowest volatilities, as on average they have less than a quarter of the standard deviation of Earnings/Cash Flow. This evidence is not surprising. Since firms typically have economic presence in multiple states, a diversified geographical presence lowers the volatility of these measures. 6 Our Fama and MacBeth (1973) style estimation implicitly captures and controls for known market-wide return predictors (e.g., Fama and Schwert (1977), Campbell and Shiller (1988), Fama and French (1988)). 18

21 The estimation results for EC based earnings and cash flow regressions are reported in Table III. Similar to the previous results, columns 1 and 5 report the results regressing firms Earnings/Cash Flow on 1) lagged Earnings/Cash Flow, 2) lagged HQ Earnings/Cash Flow, lagged EC Earnings/Cash Flow, and control variables, whereas columns 3 and 7 report the results for the full specification that includes all control variables. In columns 4 and 8, we report the full specification results when EC Earnings/Cash Flow are calculated using only those firms that have economic interests in connected states. We exclude other firms that are headquartered in those economically connected states. With the modified earnings and cash flow measures, we want to determine whether indirect geographic connections also contain valuable information about a given firm s earnings and cash flow. 7 We find that EC Earnings contain useful information about the future earnings of a firm. This effect is incremental over the ability of HQ Earnings to predict a firm s future earnings. Specifically, the estimated coefficients on EC Earnings are all positive and statistically significant, ranging from (t-statistic = 12.95) to (t-statistic = 24.13) across columns 1-3. In economic terms, a one standard deviation increase in EC Earnings corresponds to a 0.156% increase in firms Earnings in the next quarter, which is economically meaningful. This economic magnitude is greater than the predictive power of HQ Earnings. In this expanded regression specification, we also find that the estimated coefficients on HQ Earnings remain positive and statistically significant across all specifications. However, when EC Earnings is included in the specification, the magnitudes of HQ Earnings weaken in comparison to the corresponding estimates reported in Table II, and ranges from (t-statistic = 6.42) to (t-statistic = 6.50). 7 To define EC earnings and cash flow measures, we consider firms headquartered in multiple states. For example, when we predict the earnings of a given firm j that is headquartered in state A but has economic interests in states B, C, and D, we use the lagged earnings of firms located in states B, C, and D. But when we identify the firms located in those three states, we only consider other firms that also have economic interests in those states and we exclude the firms that are headquartered in those locations. These firms with economic interests in states B, C, and D would be headquartered in various other states. So, effectively, we are aggregating the earnings information from firms headquartered in multiple states. The common link between firm j and these firms is that they all have economic interests in at least one common geographical location (i.e., state B, C, or D). 19

22 Further, the results in column 4 show that the predictive power of EC Earnings is strong even when we define the measure using the earnings of firms that have economic interests in a given firm s economically connected states but are not headquartered in one of those states. This evidence suggests that indirect economic connections among firms also contain valuable information about earnings of a given firm. A similar pattern is present in the Cash Flow regression results. We find that EC Cash Flow exhibits a strong ability to predict firms cash flow in the next quarter. Specifically, the estimated coefficients on EC Cash Flow are positive and statistically significant across columns 5-7, ranging from (t-statistic = 2.35) to (t-statistic = 4.47). Similar to our earlier evidence, EC Cash Flow has stronger predictive ability than HQ Cash Flow, where a one standard deviation increase in EC Cash Flow predicts a 0.468% increase in firms Cash Flow in the next quarter. As in the case of earnings, the results in column 8 show that this predictive power extends to the case where firms headquartered in a firm s economically connected states are excluded from the analysis. In addition, since all regressions specifications include the EC Economic Activities Index, our results are unlikely to reflect the direct effects of the overall macroeconomic environment in economically-relevant states. Collectively, these predictive regression results show that the fundamental information of firms located in economically connected states contains valuable incremental information about the future performance of firms that are not headquartered in those states. 3.3 Earnings Predictability: Robustness Tests We conduct several additional tests to examine the robustness of our baseline results. Specifically, we examine whether our baseline estimates are robust to an alternative definition of economically connected states. We also conduct a falsification exercise, examine the robustness of our results for various subsamples, and report estimates after accounting for industry effects. The 20

23 results from these tests are summarized in Table IV. For brevity, we report these results using only the earnings measure, but our results are very similar when we use the cash flow based measure. In the first set of tests, instead of using all cited states, we alternatively compute the earnings measure using only the top-three states and top-one state with the highest citation-share, excluding the HQ state. It is possible that the states that are mentioned infrequently are not economically relevant for the firm and do not contain valuable information about the earnings of the firm. We find very similar results when we use these alternative definitions of earnings (see columns 1 and 2). Specifically, the coefficients on EC Earnings remain strongly positive, with a value of (0.107) for the top-three states (top-one state), and statistically significant at the 1% level. Further, we find that the EC earnings measure continues to exhibit stronger predictive power relative to the HQ earnings measure. In particular, given that the mean and standard deviation of EC Earnings (HQ Earnings) are and (0.003 and 0.015), respectively, a one standard deviation change in top-three EC Earnings (HQ Earnings) leads to 0.120% (0.062%) increase in firms Earnings in the next quarter, which is economically significant. 8 In the second set of tests, we examine the robustness of our regression specification. Specifically, instead of using current quarter earnings as a predictor of earnings in the next quarter, we examine the ability of EC Earnings and HQ Earnings to predict the change in earnings over the next quarter. The results in column 3 show that EC Earnings is a strong predictor of the change in earnings over the next quarter, with an estimated coefficient of and t-statistic of 4.11 indicating significance at the 1% level. On the other hand, HQ Earnings has virtually no predictive power when we consider changes in quarterly earnings. This is further evidence that the EC Earnings measure has stronger predictive power than the HQ Earnings measure. 8 Our evidence of predictability is very similar when we use the top five EC states to define the earnings measures. We also find that our evidence is robust to eliminating the most heavily-cited EC states (California, New York, and Texas). 21

24 To further test the robustness of our EC Earnings and HQ Earnings coefficient estimates, we conduct a falsification exercise. We randomly re-assign HQ and EC locations across firms in the sample each quarter, maintaining the cross-sectional distribution of HQ and EC states. We repeat this procedure 1,000 times, re-estimating our baseline specification for each trial and collecting the EC Earnings and HQ Earnings coefficients. We examine the distributions of the coefficients and report the average coefficients in column 4 of Table IV. The average coefficients across the trials are economically small and statistically insignificant (p-values are and 0.290). 9 In the next test, we address the potential concern that our results are influenced by firms incorporated in Delaware. We exclude all sample firms that are incorporated in Delaware and reestimate our baseline regression. The results reported in column 5 of Table IV indicate that our estimates remain similar. This finding suggests that our results are unlikely to be driven by excessive citation of Delaware in the 10-K reports that is unrelated to economic activities of firms in that state. To ensure that our results are not sensitive to our choice of a quarterly measurement period, we re-estimate our key regression using annual data. Although the number of cross-sections drops substantially from 64 to 16, we find in column 6 of Table IV that the strong predictive power of EC Earnings and EC Cash Flow remains similar. This evidence suggests that our results are not very sensitive to the data frequency used in the analysis. In the next set of tests, we examine whether our results capture potential earnings predictability at the industry level. We re-estimate the baseline regression after including the Fama and French 38 (FF38) industry fixed effects as well as quarter fixed effects in the specification. The 9 We also perform an alternative placebo test where we include an additional predictor in the regression specification: the earnings or cash flow of a randomly chosen state, which is neither the HQ state nor one of the EC states of the firm. The results from this test are very similar to the results from the falsification exercise reported in column 4 of Table IV. The coefficient estimate of this random predictor is small and statistically insignificant (mean estimate = , p- value = 0.459). 22