The Federal Reserve Board eagle logo links to home page

Skip to: [Printable Version (PDF)] [Bibliography] [Footnotes]
Finance and Economics Discussion Series: 2009-46 Screen Reader version
Firm Volatility and Banks: Evidence from U.S. Banking Deregulation
Ricardo Correa
Gustavo A. Suarez*
Corresponding author: [email protected].
Federal Reserve Board September 23, 2009

Keywords: Bank deregulation, firm volatility, external finance, idiosyncratic volatility

Abstract:

This paper exploits the staggered timing of state-level banking deregulation in the United States during the 1980s to study the causal effect of banking integration on the volatility of non-financial corporations. We find that firm-level employment, production, sales, and cash flows are less volatile after interstate banking deregulation, particularly for firms that have limited access to external finance. This finding suggests that bank-dependent firms exploit wider access to finance after deregulation to smooth out idiosyncratic shocks. In fact, short-term credit becomes less pro-cyclical after out-of-state bank entry is permitted. Finally, lower volatility in real-side variables after deregulation translates into lower idiosyncratic risk in stock returns.

JEL Classification Codes: G21, G32


A growing body of theoretical and empirical research indicates that the ability of firms to access banking finance fosters average growth.1 For example, Rajan and Zingales (1998) find that industries that depend on external financing for investment grow faster in countries with larger banking systems. In addition to its positive effect on average growth, access to banking finance may have an important effect on growth volatility. The effect of financial markets--and banking finance in particular--on volatility is less well understood. This paper studies the relationship between banking integration and the volatility of the corporate sector using data on publicly-traded U.S. firms.

From a macroeconomic point of view, volatility is important because of the growth benefits conferred by stability (Ramey and Ramey (1995), Aghion et al. (2005)). From the point of individual corporations, volatility is important because stable firms face lower expected costs from financial distress (Smith and Stulz (1985)), financial policies are more effective for solving agency problems in stable firms (Stulz (1990)), and investors value firms with smooth cash flows at a premium (Rountree, Weston, and Allayannis (2008)).2

In theory, the direction of the effect of banking finance on firm volatility is ambiguous. On the one hand, wider access to external finance may increase the ability of firms to offset idiosyncratic shocks, thus reducing firm volatility. On the other hand, wider access to external finance may promote specialization and allow firms to pursue riskier and more profitable projects (Thesmar and Thoenig (2009)), thus increasing firm volatility. As a consequence of mutually offsetting forces, the net effect of access to banking finance on firm volatility is an empirical question.3

Recent papers in this area suggest that wider access to banking finance may lower volatility in the corporate sector. Extending the methodology of Rajan and Zingales (1998), Larrain (2006) finds that growth in industries that depend on external finance is less volatile in countries with more bank credit relative to GDP. Using a similar cross-country, cross-industry approach, Raddatz (2006) argues that the volatility-reducing effect of more bank intermediation results partly from the role of the financial system in providing liquidity.4 Cross-country studies results may also be consistent with a reverse causality interpretation, where a more stable corporate sector facilitates the expansion of bank credit.

This paper complements and refines the findings in Larrain (2006) and Raddatz (2006) by using a different identification strategy for isolating the effect of banking finance on corporate volatility. In particular, we use a differences-in-differences approach that exploits the staggered timing of interstate banking deregulation across U.S. states during the 1980s and early 1990s to estimate the effect of banking integration on firm-level volatility. After a state passed an interstate banking deregulation law, out-of-state banks were allowed to acquire banks incorporated in that state, effectively promoting the integration of banking markets. We interpret deregulation that facilitates banking integration, starting from a situation with highly segmented markets, as promoting wider access to banking finance and improving efficiency in intermediation, possibly reflecting that financial institutions become more geographically diversified (Jayaratne and Strahan (1996), Morgan, Rime, and Strahan (2004)). With this interpretation, the results of this paper may inform the debate about the impact of access to banking finance on firm-level volatility.

Our identification strategy is similar to Morgan, Rime, and Strahan's (2004), but we focus on firm volatility--as opposed to state volatility. Firm-level data help us in identifying banking deregulation as the driving force behind our results, because we are able to compare, within each state, the effect of deregulation on bank-dependent firms with the effect on non-bank-dependent firms. Previous research suggests that small firms and firms with limited access to corporate debt face significant asymmetries when accessing credit markets (Almeida, Campello, and Weisbach (2004), Gilchrist and Himmelberg (1995)). These types of firms tend to rely more heavily on banks for their external financing needs and are defined as bank dependent in our regression analysis. We expect bank-dependent firms to be more sensitive to changes in banking conditions within a state.

We find that firms located in states that deregulate interstate banking exhibit a reduction in the volatility of the growth rates of sales, production, cash flow, and employment after deregulation. In our main result, the reduction in volatility is likely to be associated with the changes in banking finance derived from deregulation, because it is concentrated on firms that are more likely to rely on banks for external financing. For example, firms without bond or commercial paper credit ratings or issues reduce their volatility after deregulation by more than firms with credit ratings or issues. Further tests suggest that firms exploit credit markets to smooth temporary cash-flow shocks, as short-term credit becomes more countercyclical after deregulation (as in Larrain (2006)).

In addition, by using firm-level data we are able to examine the implications of wider access to banking finance for the volatility of individual stock returns. We document that the idiosyncratic volatility of stock returns falls after deregulation, particularly for firms that are a priori more likely to rely on banking finance. The residual variance of a market model of excess returns falls after deregulation for firms located in states that open to interstate banking. This finding is robust to adding the size and value factors of Fama and French (1993) and the momentum effect of Jegadeesh and Titman (1993).

At first glance, our results might seem at odds both with the increase in volatility of firm sales growth documented by Comin and Philippon (2005) and Comin and Mulani (2006) and with the increase in the idiosyncratic volatility of individual stock returns documented by Campbell et al. (2001). While these papers describe aggregate trends over the second half of the twentieth century, we restrict our attention to the window of years around deregulation of interstate banking. Importantly, the increasing trend in volatility described by Comin and Philippon (2005) temporarily reverses during the 1980s--the years with the heaviest deregulation activity at the state level and, therefore, the main part of our sample. Similarly, the increasing trend in idiosyncratic volatility described by Campbell et al. (2001) does not apply to the 1980s.5 In this paper, we abstract from aggregate trends in idiosyncratic volatility by controlling for year-fixed effects. We argue that part of the remaining variation in firm-level volatility is explained by interstate banking deregulation.

The rest of the paper is organized as follows. Section I briefly reviews the interstate banking deregulation in the 1980s in the United States and summarizes some related studies. Section II presents our empirical methodology and the data. Section III shows the effect of banking deregulation on firm-level volatility and its differential impact on those firms that are more likely users of banking finance. This section also links volatility in operations and stock return volatility using multifactor models of stock returns at the firm level to isolate the idiosyncratic component of returns. Section IV concludes.

I. The Deregulation of Interstate Bank Entry

The Douglas Amendment to the Bank Holding Company Act of 1956 prohibited Bank Holding Companies (BHCs) from establishing or purchasing bank subsidiaries across state lines unless the state of the target bank authorized the transaction. These restrictions remained in place until Maine passed a law allowing out-of-state BHCs to purchase local banks if the "home" state of the BHC reciprocated. This did not happen until 1982 when Alaska and New York passed similar laws. The same year, as part of the Garn-St Germain Act, federal legislators amended the Bank Holding Company Act to allow failed banks to be acquired by any BHC, regardless of origin and state laws. This regulatory change, coupled with a series of bank and thrift failures during the eighties, triggered a wave of interstate agreements that effectively permitted banking at the national level. By 1994, 49 states and the District of Columbia had deregulated their banking markets allowing out-of-state entry.6

The episode of banking deregulation in the 1980s and early 1990s changed the terms in which nonfinancial firms were able to access banking finance. In particular, Jayaratne and Strahan (1998), Dick (2006), and Rice and Strahan (2009) find that loan prices and spreads decrease after banking deregulation.

A series of studies have analyzed the effect of this episode of banking deregulation on real economic activity. Strahan (2003) finds that interstate deregulation is associated with an increase in incorporations by state and a reduction in the link between state growth and local bank performance.7 In a study on income insurance, Demyanyk, Ostergaard, and Sorensen (2007) find that deregulation, measured as the combination of intrastate and interstate deregulation, decreases the correlation between personal income and state-specific shocks to output. Their result is stronger for proprietor income than for wage income. The authors explain this effect by the closer relationship between banks and small businesses. This result is connected to Morgan, Rime, and Strahan's (2004) finding that geographical bank integration reduces employment volatility within states. This change is due to a decline in the impact of bank capital shocks on state activity.

A limited number of studies have examined the connection between deregulation and firm-level dynamics. Correa (2007) finds that banking deregulation is associated with a decrease in financing constraints for small publicly-traded firms, explained by lower costs of financing after deregulation. From a theoretical perspective, Stebunovs (2006) uses a stochastic general equilibrium model to assess the consequences of banking deregulation on volatility. The author finds that deregulation increases firm entry by reducing local bank monopoly power, thus dampening firm- and aggregate-level fluctuations.

In this paper, we expand on these findings and analyze the empirical effect of interstate bank-entry deregulation on firm volatility. Moreover, we test if changes in the volatility of firm fundamentals are accompanied by changes in the volatility of individual stock returns.

II. Data and Methodology

A. Data

In the empirical analysis of this paper, we use balance sheet and income statement data extracted from the Compustat North-America database. The sample includes observations for firms classified in the manufacturing (SICs 2000 to 3999), wholesale (SICs 5000 to 5199), and retail trade (SIC 5200 to 5999) sectors. To focus on the episode of interstate banking deregulation of the 1980s and early 1990s, our sample includes data between 1976 and 1998.

Compustat's geographic information reports a firm's headquarters location only for the latest year available in the database. To determine whether a firm was affected by bank entry deregulation, we need to find the actual historical location of its headquarters. For this purpose, we collect data from Compact Disclosure between 1988 and 1998. This source contains extracts from SEC filings updated every month, including the firm's address. Using this information, we determine the state where the firm was headquartered during the deregulation period.8 A firm is excluded from the sample if it changed its location between 1988 and 1998. In addition, we consider only firms with non-missing information two years before and two years after interstate bank entry deregulation in the state where headquarters are located.9

After applying these restrictions, the data consists of 39,624 observations for 2,272 firms in 45 states and the District of Columbia.10 Table I shows the distribution of firms by state. Firms in New York and California account for one quarter of the sample; other important states in the Northeast (Connecticut, Maryland, Massachusetts, New Jersey, and Pennsylvania) account for about 20 percent of the sample; and large industrial states in the Midwest (Illinois, Michigan, and Ohio) represent 13 percent of the firms covered. Due to the sample selection criteria, an average firm stays in the sample for 17 years.11

Table I

Stock returns are collected from the Center for Research in Security Prices (CRSP) dataset. Additionally, the one-month Treasury Bill rate and the Fama-French and Jegadeesh-Titman factors are taken from Kenneth French's Data Library at Dartmouth.

B. Firm Volatility

The effect of bank entry deregulation on firm volatility is ambiguous from a theoretical standpoint (Larrain (2006), Morgan, Rime, and Strahan (2004)). On the one hand, wider availability of bank credit may dampen the effect of idiosyncratic shocks on productive activities, making firms more stable. On the other hand, improved access to bank financing may increase volatility by allowing firms to undertake riskier and more profitable projects (e.g., adopting new technologies). Hence, we adopt an empirical approach to study the relationship between firm volatility and access to bank credit.

To estimate the effect of interstate bank entry deregulation on volatility we follow Morgan, Rime, and Strahan (2004) and use a two-stage procedure. In the first stage we calculate a time-varying measure of firm volatility. In the second stage we use this measure to determine whether interstate bank entry deregulation had an effect on the volatility of non-financial corporations.

Let i, j, k, and t index firm, state, industry, and year, respectively, and Y_{ijt} be a generic variable. We define the volatility of Y_{ijt} as the deviations in absolute value of Y_{ijt} from its predicted conditional mean. Formally, we estimate the following equation:

\begin{displaymath} Y_{ijt} =\alpha _i +\mu _t +\varphi Integration_{jt} +\eta X_{ijt-1} +\rho Z_{jt} +\kappa W_{kt} +\upsilon _{ijt} \end{displaymath} (1)

where \alpha _i are firm-fixed effects; \mu _t are time effects; Integration_{jt} is an indicator variable that equals one if state j permits out-of-state bank entry through acquisitions in year t; X_{ijt}are firm-specific controls; Z_{jt} represents the growth rate of state per capita income; and W_{kt} is the growth rate in real sales at the industry level (3-digit SIC). We consider three firm-level controls: the log of real sales (log(Sales)), tangible assets divided by total assets (Tangible/Assets), and earnings before taxes, depreciation, and amortization divided by total assets (EBITDA/Assets).12 These measures are associated with firm-leverage (Rajan and Zingales (1995)), and reflect, respectively, the size of the firm, its collateral, and profitability.

After estimating equation (1), the volatility for variable Y_{ijt}, Vol(Y_{ijt}), is defined as \left\vert {\hat {\upsilon }_{ijt} } \right\vert. Notice that this measure of volatility is the absolute deviation of the firm's indicators from the overall trends in the economy, from the state where its headquarters are located, and from the industrial sector that represents its main activity. In addition, we allow for changes in the firms' growth rate after interstate bank entry is permitted. The measure estimated from this empirical equation captures the idiosyncratic component of firms' volatility. Measuring volatility using absolute deviation (as opposed to squared deviations) implies that volatility and growth are conveniently expressed in the same units.

As part of their intermediation function, banks have a comparative advantage in providing liquidity to firms, particularly if the demand for funds is related to individual projects (Larrain (2006)). Therefore, to analyze the effect of banking deregulation on firm volatility, we will focus on its idiosyncratic component. In the second stage of our estimating procedure, we use the volatility measures defined in equation (1) and estimate the following equation:

\begin{displaymath} Vol\left( {Y_{ijt} } \right)=\alpha _i +\mu _t +\delta Integration_{jt} +\beta X_{ijt-1} +\gamma Vol\left( {Z_{jt} } \right)+\phi Vol\left( {W_{kt} } \right)+\varepsilon _{ijt} \end{displaymath} (2)

We add two regressors, in addition to the set included in equation (1), to control for fluctuation at the state and industry level. Vol(Z_{jt}) and Vol(W_{kt}) are volatility proxies for the states' per capita income growth and real sales growth at the industry level, respectively.

In our baseline regressions, Y_{ijt} is the growth rate of production, sales, cash flow, and employment, indicators of real corporate activity. These measures allow us to determine if output fluctuations changed after banking deregulation.13 As in Larrain (2006), firm production is measured by inventory investments plus sales. In addition to this set of variables, we analyze the evolution of the firms' internal cash flow. Panel A in Table II shows summary statistics for the measures of volatility associated with these variables.14

Table II

To test whether firm volatility decreases after interstate bank entry deregulation, we examine the sign of \delta in equation (2). As documented by Comin and Mulani (2005), the volatility of publicly-traded firms increased over the last five decades. The inclusion of time-fixed effects captures this secular trend. Therefore, a negative sign on the Integration coefficient is interpreted as a decrease in the upward trend in volatility.

We control for variation at the industry (W_{kt}) and state level (Z_{jt}) to isolate the effect of deregulation on the idiosyncratic component of firm-specific volatility. The former is measured by the absolute value of deviations of the log change in sales at the 3-digit SIC level, Vol(Industry Sales).15 The state-level component is proxied by the volatility of log changes in per capita income, Vol(Per Capita Income).16 Finally, selecting firms with observations before and after deregulation and controlling for firm-fixed effects alleviates sample biases in which firms that enter after deregulation exhibit a systematically different volatility than firms that exit before deregulation.


C. Firm Volatility and Bank Dependence

The baseline empirical estimation described in the previous section takes advantage of the staggered deregulation dates across states to identify whether there was a change in the firms' volatility measures explained by lower restrictions on out-of-state bank entry. However, this strategy does not fully control for other potential shocks that might have had an effect on firms headquartered within the state at the time of banking deregulation. We take this problem into account by using another layer of differentiation. In particular, we test whether the effects of bank entry deregulation is stronger for bank-dependent firms, as opposed to firms with access to other sources of external financing, within each state.

We define two proxies for firms' Bank Dependence: one based on size and the other based on the use of public corporate debt. These variables have been widely used in the financing constraints literature (e.g., Almeida, Campello, and Weisbach (2004)). Empirical studies typically find that small firms and firms with limited access to corporate debt face significant information asymmetries when accessing credit markets,17 and thus rely more on the use of internal funds or on credit from banks to finance their operations. Dependence on internal funds and bank credit makes these firms more likely to experience a change in financing conditions due to interstate bank entry deregulation. Formally, the following equation is estimated using volatility in production, sales, cash flow, and employment as dependent variables:

\begin{displaymath} \begin{array}{c} Vol\left( {Y_{ijt} } \right)=\alpha _i +\mu _t +\delta Integration_{jt} +\lambda Bank\mbox{ }Dependence_{ijt} \ +\varphi Integration_{jt} \times Bank\mbox{ }Dependence_{ijt} +\beta X_{ijt-1} +\gamma Vol(Z_{jt} )+\phi Vol(W_{kt} )+\varepsilon _{ijt} \ \end{array}\end{displaymath} (3)

where the right-hand side variables are the same as in equation (2), with the exception of Bank Dependence and its interaction with the dummy for bank integration. It is worth stressing that our two measures of Bank Dependence--the one based on firm size and the one based on access to public debt markets--are constructed so that larger values of the variable reflect a heavier dependence on banking finance. If bank-dependent firms benefit more from interstate deregulation, we would expect \phi to have a negative sign. This implies that volatility decreases after deregulation for those firms that rely more heavily on banks for their external financing needs.

In the empirical estimations, the size-based measure of bank dependence is an indicator variable that equals one if the firm's assets are below median assets in a given year and state. Panel B in Table II shows that, according to this criterion, 42.1 percent of firm-year observations in our sample are classified as bank-dependent because of their small size. The second measure, based on the use of corporate debt, is defined through the firm's history of credit ratings and issues between 1970 and 1994. A firm is classified as being bank-dependent if it did not issue debt nor had any credit ratings in this period. Bond and commercial paper credit rating information is retrieved from Compustat. Bond and commercial paper issues are obtained from the Mergent Fixed Income Security Database (FISD) and Moody's Default Risk Service (DRS) Database. Out of the 2,272 firms included in the sample, 1,516 are classified as bank-dependent because they did not have any issues or credit ratings for this period. As shown in Panel B of Table II, this group of firms accounts for 62.5 percent of firm-year observations.


D. Deregulation and the Cyclicality of Short-Term Borrowing

If bank lending availability is enhanced by deregulation, bank-dependent firms faced with idiosyncratic shocks would be able to borrow during periods of output contraction. Hence, bank credit becomes less pro-cyclical and bank entry deregulation would lead to smaller firm volatility. The next step in our analysis is to test whether bank credit became less pro-cyclical for bank-dependent firms after bank entry deregulation. For this purpose, we estimate the following equation:

\begin{displaymath} \begin{array}{c} \Delta \log \left( {Notes\mbox{ }Payable_{ijt} } \right)=\alpha _i +\mu _t +\delta Integration_{jt} +\theta Cyclicality_t + \ \lambda Bank\mbox{ }Dependence_{ijt} +\varphi Integration_{jt} \times Bank\mbox{ }Dependence_{ijt} \ +\nu Integration_{jt} \times Bank\mbox{ }Dependence_{ijt} \times Cyclicality_t \ +\beta X_{ijt-1} +\varepsilon _{ijt} \ \end{array}\end{displaymath} (4)

where \Delta \log \left( {Notes\mbox{ }Payable_{ijt} } \right)is the log change in notes payable for firm i in state j and year t. Although loans from banks are not reported in Compustat, Notes Payable approximate access to credit from financial institutions.18 We consider two alternative measures of the cycle: the first is firm sales scaled by lagged assets, Sales_{ijt}/Assets_{ijt-1}, and the second is the growth rate in real per capita GDP in the United States.

The coefficient on the cyclicality variables measures the co-movement of short-term borrowing with business conditions. In these estimations, we test if \nu is negative and significant. When this coefficient is negative, cyclicality decreases after banking deregulation for the set of firms that most likely use bank credit. This would be evidence that bank entry deregulation reduced the pro-cyclicality of lending to bank-dependent firms.


E. Equity Return Volatility

In addition to real variables, we study the effect of deregulation on equity return volatility. A change in firms' volatility, both of production and profitability, would likely have an effect on stock returns. As shown by Pastor and Veronesi (2003), idiosyncratic return volatility increases with the volatility of profitability. This is particularly relevant for financially constrained firms. A decrease in the volatility of financially constrained firms after bank entry deregulation should also reduce stock return volatility. To test this hypothesis, we measure stock return volatility in our baseline specifications as the standard deviation of the residuals from the following market model estimated with monthly observations:

\begin{displaymath} \left( {R_{ijt} -R_t^f } \right)=\alpha _i +\beta _i \left( {R_t^{vw} -R_t^f } \right)+\varepsilon _{ijt} \end{displaymath} (5)

where R_{ijt} is the return to shareholders of firm i in period t, which in our stock return regressions represents months instead of years. R_t^f is the risk-free return rate proxied by Ibbotson's one-month Treasury Bill rate; R_t^{VW} is the value-weighted return on all NYSE, NASDAQ, and AMEX stocks; and \varepsilon _{ijt} is the idiosyncratic component of excess returns. The market model in equation (5) is estimated for each firm before and after interstate entry deregulation, which allows for different coefficients in the two periods of interest. We also estimate (5) adding the Fama-French size and value factors, and the Jegadeesh-Titman momentum factor. Lastly, we analyze raw excess returns computed as the residuals in (5), without controlling for the excess returns on the market portfolio.

We define idiosyncratic stock return volatility as the standard deviation of residuals in (5) for each firm and year. Figures 1 and 2 show the evolution of these residual returns before and after deregulation. Following banking deregulation, the median idiosyncratic component of stock return volatility declines for three years. Campbell et al. (2001) and Pastor and Veronesi (2003) document a long-term increase in firm volatility starting in the 1960s. This trend temporarily reverses in the years after deregulation in the 1980s, consistent with the drop in idiosyncratic stock return volatility in Figures 1 and 2.

Figure 1

Figure 2

Figure 3 shows the evolution of median stock return volatility before and after deregulation splitting the sample by bank dependence. Compared to the group of firms with access to public debt markets, excess return volatility is higher for firms that are bank-dependent and experiences a steeper decline after banking deregulation. This evidence suggests that the stabilizing effects from banking deregulation may have benefited more those firms that depend to a greater extent on banks for their financing needs. The next section presents the statistical analysis behind this graphical evidence.

Figure 3

III. Results

A. Firm Volatility

In this section we analyze the effect of interstate bank entry deregulation on firm volatility. The focus is on variables that account for firm output and performance. Table III presents the results of estimating equation (2) with the volatility of production, sales, cash flow, and employment as dependent variables. The coefficient on Integration is negative in all columns and statistically significant in the estimations for the volatility of cash flow and employment. To illustrate the economic magnitudes involved: after deregulation, the volatility of cash flow decreases by roughly 14 percent of the median volatility of cash flow in the sample. These results signal a sizeable decrease in volatility after out-of-state banks were permitted to enter local markets.

Table III

The findings in Table III also suggest that smaller and less profitable firms tend to be more volatile. Firm-specific volatility is not significantly correlated with state-level fluctuations. By contrast, industry-wide fluctuations appear to be an important component of firm volatility, as noted by the positive and significant coefficient on Vol(Industry Sales).

The results in this section are consistent with those in Morgan, Rime, and Strahan (2004), who find a decrease in the volatility of the growth of state-level employment due to banking deregulation. This decrease in volatility is associated with bank geographical diversification, and, therefore, less vulnerability to state-specific shocks. In the following sections, we study the mechanism that triggers the reduction in volatility at the firm level. But first we establish if bank-dependent firms are the ones that benefited the most from banking deregulation.

B. Firm Volatility and Bank Dependence

Small firms and firms without access to public debt markets are more likely to use bank credit to finance their operations. In the next set of tests, we compare the volatility of bank-dependent firms with the volatility of firms with access to other sources of finance, before and after the state deregulation of bank entry. These estimations allow us to control for changes in firm volatility within a state that are unrelated to decreases in out-of-state bank entry restrictions. Formally, we test whether volatility for bank-dependent firms decreased by more after interstate banking deregulation.

Table IV reports the results of estimating equation (3). In the regressions in Panel A, bank dependence is measured as lack of access to public debt markets, while, in Panel B, bank dependence is proxied by small firm size.19 The coefficient on the interaction between bank dependence and the deregulation dummy is negative and significant in all specifications. This finding reflects the importance of deregulation on the observed decrease in volatility for the sample of bank-dependent firms. Bank deregulation makes bank-dependent firms more stable, but does not significantly alter the volatility of those firms that have ample access to non-bank sources of funding. The results in Table IV suggest that the effect of bank deregulation on firm volatility operates through direct relationship between firms and banks.

Table IV

The differential effect of deregulation on bank-dependent firms is substantial in most specifications, but it is particularly sizeable for cash-flow volatility. After deregulation, the reduction in the volatility of cash flow for small firms or for firms without access to public debt markets is about 27 percent relative to the median volatility of firm cash flow in the sample. The magnitude of the decrease in volatility of bank-dependent firms is somewhat smaller but still substantial when considering production, sales, and employment. As discussed by Booth and Cleary (2008), firms with more volatile cash flows hold larger amounts of financial slack to finance their investment. Lower volatility of output and cash flow and better access to external finance should decrease the need for cash holdings and increase their profitability.

The results outlined in this section confirm the significant effect of bank entry deregulation on the volatility of bank-dependent firms. After checking the robustness of our main results to different specifications, sample restrictions, and controls in the next section, we will assess whether the reduction in firm volatility is explained by a decrease in the pro-cyclicality of bank credit.

C. Robustness Checks

This section tests the sensitivity of our basic results in equations (2) and (3) to different specifications. In particular, Table V considers alternative volatility and deregulation measures, different sample selection criteria, and additional control variables. For the sake of compactness, Table V only reports the coefficients on the interstate deregulation dummy and its interaction with the bank dependence proxy (as measured by the firm's access to commercial paper and bond markets).20

Table V

Panels A and B of Table V consider volatility measures alternative to the one defined in equation (1). Panel A reports estimates of regressions (2) and (3) computing the volatility of a generic variable Y_{ijt} as squared (instead of absolute) deviations of the variable from its conditional mean. Using the notation of equation (1), we use \hat {\upsilon }_{ijt}^2 instead of \left\vert {\hat {\upsilon }_{ijt} } \right\vert. Panel A suggests that giving a heavier weight to larger deviations in computing volatility does not affect qualitatively our basic results.

Panel B compares, for each firm, the five years before deregulation to the five years after deregulation, collapsing all observations into two periods. The dependent variables are the standard deviations of production, sales, cash flow, and employment in each period. Similarly, control variables are included in the regression as period averages. The results in Panel B show that the standard deviation of production, sales, and employment is lower in the five-year period after deregulation with respect to the five-year period after deregulation. The measure of volatility in Panel B is less noisy than the one we use in our baseline regressions, but it also makes hard to distinguish a causal effect of banking deregulation on firm volatility from an aggregate trend towards stability over the sample period.

Panel C replaces the banking deregulation dummy with a continuous measure of interstate banking integration: the share of deposits held by multi-state banks in each state and year. We instrument this continuous variable using the banking deregulation dummy. The results in Panel C indicate that firms become more stable when the fraction of deposits held by multi-state banks increases for the state where firms are located. In other words, actual integration of banks across states--the goal of deregulation--reduces volatility in non-financial firms, particularly for bank-dependent firms. The results in Panel C suggest that as the lending capacity of a state's "financial system" becomes more diversified through the participation of out-of-state banks, firms located in the state become more stable.

Panels D and E explore alternative sample selection criteria for regressions (2) and (3). Some states deregulated intrastate banking around the same time they deregulated interstate banking. In order to isolate the effect interstate deregulation, Panel D excludes observations in states where intrastate banking was deregulated within a year of interstate bank-entry deregulation. Excluding these observations reduces the sample by about a sixth. Panel E exclude observations from the two states with the highest number of firms, California and New York, which account for about a quarter of the sample. The baseline results are insensitive to applying the more restrictive selection criteria of Panels D and E.

Panel F includes all available firm-year observations, unlike our baseline regressions, which include only firms with non-missing observations two years before and two years after deregulation. The results in Panel F are very similar to our baseline results. Hence, our finding that bank-dependent firms become more stable after deregulation is dominated by the decrease in volatility for surviving firms, leaving only a relatively small role for differences in across firms that entered after deregulation and those that exited before deregulation.

Panels G and H expand the set of control variables of the baseline regressions. In Panel G, we add a proxy for banking concentration to the right-hand side of equations (2) and (3). We measure banking concentration using the Herfindahl-Hirshman index of demand deposits by state. Panel H considers a larger set of firm-control variables. First, it includes leverage as measured by total debt scaled by assets. Second, it substitutes firm age, measured as time from IPO, for log sales. (Age and log-sales are not included simultaneously, as they are highly correlated.) The results from panel H suggest that the reduction in volatility after deregulation in previous sections is not explained by firms becoming more stable as they mature. Furthermore, in results not reported here, we estimate equations (2) and (3) replacing the actual banking deregulation dummy with a "placebo" deregulation dummy, constructed as an indicator that equals one starting two and four years after the actual deregulation took place in each state. We find that the effect of this "placebo" banking deregulation on the volatility of bank-dependent firms is not statistically significant.

The results summarized in Table V suggest that our main findings are robust to different volatility and deregulation measures, to alternative sample selection criteria, and to the inclusion of additional controls. In practically all regressions reported in Table V, the interaction between the interstate deregulation dummy and the bank dependence proxy is negative and significant. This evidence suggests that firms that are more likely to rely on bank credit tend to be more stable, on average, after interstate banking deregulation.21

D. Cyclicality of Short-Term Borrowing

In this section we explore the channel that leads to the decrease in real volatility for our sample of publicly-traded firms. As Larrain (2006) points out, lower volatility has to be accompanied by increased countercyclical borrowing. As banks become more geographically diversified through deregulation, the correlation between bank capital and economic growth at the state level decreases (Morgan, Rime, and Strahan (2004)). Therefore, firms exploit bank credit to buffer from negative shocks.

We test the link between deregulation and firm volatility by checking the cyclicality of short-term borrowing. Table VI summarizes the results of estimating equation (4) using notes payable as the measure for short-term borrowing. Column (1) shows that notes payable is pro-cyclical on average. The coefficient on the interaction between the cyclicality variables and the deregulation dummy is negative but statistically insignificant. As expected, more profitable firms and with larger shares of tangible assets have higher borrowing growth rates.

Table VI

Columns (2) and (3) add proxies for bank dependence: public debt access and size. In both columns, the coefficient on the triple interaction between cyclicality, bank-dependence, and deregulation is negative and significant. Hence, short-term borrowing for small firms becomes less pro-cyclical after deregulation. The same pattern is observed for firms that do not use bonds or commercial paper to finance their operation. This result is consistent with the decrease in the volatility of firm output and short-term credit described above. Geographically diversified banks are likely able to buffer firms from state-specific shocks by smoothing the credit they provide over the cycle.

E. Stock Return Volatility

In previous sections, we determined that output and cash flow volatility decreased after banking deregulation, especially for bank-dependent firms. This effect was accompanied by a reduction in the pro-cyclicality of credit. Based on our previous findings, we expect that lower volatility in real and financial variables after deregulation translates into less volatile stock returns. To test this hypothesis, we estimate the following equation:

\begin{displaymath} \begin{array}{c} Vol\left( {R_{ijt} } \right)=\alpha _i +\mu _t +\delta Integration_{jt} +\lambda Bank\mbox{ }Dependence_{ijt} \ +\varphi Integration_{jt} \times Bank\mbox{ }Dependence_{ijt} +\beta X_{ijt-1} +\varepsilon _{ijt} \ \end{array}\end{displaymath} (6)

where Vol(R_{ijt}) is the standard deviation of idiosyncratic returns as defined in Section II.B. Our focus is on \phi , the coefficient on the interaction between access to external finance and banking deregulation.

Table VII presents the result of estimating equation (6). The dependent variable in columns (1) through (3) is a measure of idiosyncratic stock return volatility computed as the standard deviations of excess returns on the firm stock (over the market return). The dependent variable in columns (4) through (6) measures idiosyncratic stock return volatility as the standard deviations of the residuals after estimating the market model for each firm. Idiosyncratic volatility in columns (7) through (8) is derived by adding the size and value factors from Fama and French, and the momentum factor from Jegadeesh and Titman (1993).

Table VII

In columns (1), (4), and (7) the coefficient on the deregulation indicator is negative but not statistically significant. On average, the decrease in idiosyncratic return volatility is minimal. On the contrary, if we focus on bank-dependent firms, we find a considerable decrease in idiosyncratic return volatility after deregulation. These findings confirm the results in Pastor and Veronesi (2003) in a dynamic setting. As uncertainty about bank-dependent firms' future profitability decreases due to better access to external financing, stock return volatility for these firms also declines. In sum, the financing environment has a significant effect on a firm's real and financial volatility.


IV. Conclusion

The empirical methodology of this paper exploits the staggered timing of state deregulation of interstate banking in the United States in the 1980s and early 1990s. In previous decades, bank acquisition activity across state lines was generally very restricted. Deregulation of interstate banking promoted integration of state banking markets and enhanced banking competition. Part of the initial benefits of deregulation accrued to firms that relied on banks for funding, by improving the terms in which they were able to access banking finance. For example, Jayaratne and Strahan (1998), Dick (2006), and Rice and Strahan (2009) find that loan prices and spreads decrease after banking deregulation.

In our first set of results, we find that firms located in states that experienced interstate banking deregulation become more stable after deregulation. In particular, the growth rates of sales, production, cash flow, and employment become more stable for firms located in states that opened to interstate banking. Since the effect is stronger for those firms that depend on banks for external finance (because they are either small or have no credit ratings or issues), the stabilization is likely to be explained by the changes in the banking system induced by deregulation. When we exclude the proxies for bank dependence, the control group is the set of firms located in states that are yet to pass deregulation laws. When we include the proxies for bank dependence, we refine our control group to those firms that depend less on banks, because they can access external finance through public markets.

Our second set of results suggests that the stabilizing effects of banking deregulation are connected to the ability of firms to exploit credit to smooth out idiosyncratic shocks. In particular, short-term credit becomes more countercyclical after deregulation. Firms may benefit from stability by smoothing investment when external financing is costly (Froot, Scharfstein, and Stein (1993)), reducing expected costs of financial distress, and lowering expected tax liabilities (Smith and Stulz (1985)). The concern for stability may be more pronounced for publicly traded firms, as Rountree, Weston, and Allayanis (2008) find that cash-flow volatility is negatively valued by investors.

In our third set of results, using a multifactor model of stock returns, we find that the idiosyncratic component of stock return volatility falls after deregulation. This finding suggests that the increased stability in operations (employment and production) and financial statements (cash flows) after deregulation translates into greater stock market stability. A reduction in idiosyncratic volatility may have an important impact on returns, because higher idiosyncratic volatility is associated with lower average returns (Ang et al. (2006a) and (2006b)).

Interpreting bank integration as a proxy for wider access to bank finance, the findings of this paper bridge two sets of results arguing that banking finance may have a stabilizing effect on the corporate sector. On the one hand, Larrain (2006) and Raddatz (2006) argue that industries that depend on banks are more stable in countries with more developed banking systems. On the other hand, Morgan, Rime, and Strahan (2004) show that banking deregulation had a stabilizing effect on state-level business cycles in the United States. Similar to Morgan, Rime, and Strahan (2004), we exploit interstate banking deregulation in the United States during the 1980s and early 1990s. This quasi-experiment addresses some potential concerns of using cross-country data. Similar to Larrain (2006) and Raddatz (2006), we exploit cross-sectional differences in the intensity of banking dependence. Unlike Larrain (2006), Raddatz (2006), and Morgan, Rime, and Strahan (2004), we are able to study the effect of banking development on individual stock returns, because we focus on firm-level data.

Recent research suggests that the reduction in macroeconomic volatility (the "Great Moderation" documented, for example, by McConnell and Perez-Quiroz (2000) and Stock and Watson (2002)) occurred despite an increase in volatility for publicly-traded firms over the second half of the 20th century.22 Our results do not run against the long-term increase in firm-level volatility documented by Campbell et al. (2001), Comin and Philippon (2005), and Comin and Mulani (2006). In fact, the upward trend in volatility of those papers is temporarily halted or even reversed during the deregulation years we study in this paper. Our results do suggest, however, that the increase in firm volatility may have been steeper without interstate banking deregulation. In other words, firm-level volatility has increased despite of--not because of--banking deregulation. Naturally, the link between firm-level volatility and aggregate volatility deserves further attention.23

In our paper, the stabilizing effects of interstate banking deregulation were obtained starting from a situation where banking markets were highly fragmented across states. For those initial conditions, interstate banking deregulation likely generated efficiency gains that improved access to banking finance for the corporate sector. A deregulatory experiment starting from different conditions may not produce the same effects we find in this paper. Similarly, other forms of deregulation may trigger a different response on banks and, thus, on nonfinancial firms.

While the development of some financial institutions, like banks, may promote stability in the corporate sector, the development of other institutions, like equity markets, may have different effects. For example, Thesmar and Thoenig (2009) use a theoretical model to study the effect of international capital integration on the volatility of publicly-traded and privately-held firms. Similarly, using a panel of countries, Bekaert, Harvey, and Lundblad (2006) show that macroeconomic volatility falls after equity market liberalizations. Understanding the role of different institutions and their interaction is an interesting area for further research.24


References

Acemoglu, Daron, Simon Johnson, James A. Robinson, and Yunyong Thaicharoen, 2003, Institutional causes, macroeconomic symptoms: Volatility, crises, and growth, Journal of Monetary Economics 50, 49-123.

Aghion, Philippe, George-Marios Angeletos, Abhijit Banerjee, and Kalina Manova, 2005, Volatility and growth: Financial development and the cyclical composition of investment, NBER Working paper No. 11349.

Almeida, Heitor, and Murillo Campello, 2007, Financial constraints, asset tangibility, and corporate investment, Review of Financial Studies 20, 1429-1429.

Almeida, Heitor, Murillo Campello, and Michael S. Weisbach, 2004, The cash flow sensitivity of cash, Journal of Finance 59, 1777-1804.

Ang, Andrew, Robert J. Hodrick, Yuhang Xing, and Xiaoyan Zhang, 2006a, The cross-section of volatility and expected returns, Journal of Finance 61, 259-299.

Ang, Andrew, Robert J. Hodrick, Yuhang Xing, and Xiaoyan Zhang, 2006b, High idiosyncratic volatility and low returns: International and further U.S. Evidence, Mimeo, Columbia University.

Bekaert, Geert, Campbell R. Harvey, and Christian Lundblad, 2006, Growth volatility and financial liberalization, Journal of International Money and Finance 25, 370-403.

Booth, Laurence, and Sean Cleary, 2008, Cash flow volatility, financial slack, and investment decisions, China Finance Review 2, 63-86.

Calomiris, Charles W., Charles P. Himmelberg, and Paul Wachtel, 1995, Commercial paper, corporate finance, and business cycle: A microeconomic perspective, Carnegie-Rochester Series on Public Policy 42, 203-250.

Campbell, John Y., Martin Lettau, Burton G. Malkiel, and Yexiao Xu, 2001, Have individual stock returns become more volatile? An empirical exploration of idiosyncratic risk, Journal of Finance 56, 1-43.

Comin, Diego, and Sunil Mulani, 2006, Diverging trends in aggregate and firm volatility, Review of Economics and Statistics 88, 374-383.

Comin, Diego, and Thomas Philippon, 2005, The rise in firm-level volatility: Causes and consequences, in Mark Gertler and Kenneth Rogoff, eds.: NBER Macroeconomics Annual (MIT Press, Cambridge, MA).

Correa, Ricardo, 2007, Bank integration and financial constraints: Evidence from U.S. firms, Working paper, Board of Governors of the Federal Reserve System.

Davies, Steven J., John Haltiwanger, Ron Jarmin, and Javier Miranda, 2006, Volatility and dispersion in business growth rates: Publicly traded versus privately held firms, in Daron Acemoglu, Kenneth Rogoff, and Michael Woodford, eds.: NBER Macroeconomics Annual (MIT Press, Cambridge, MA).

Demyanyk, Yuliya, Charlotte Ostergaard, and Bent E. Sorensen, 2007, U.S. banking deregulation, small business, and interstate insurance of personal income, Journal of Finance 62, 2763-2801.

Dick, Astrid A., 2006, Nationwide branching and its impact on market structure, quality, and bank performance, Journal of Business 79, 567-592.

Dynan, Karen E., Douglas W. Elmendorf, and Daniel E. Sichel, 2006, Can financial innovation help to explain the reduced volatility of economic activity? Journal of Monetary Economics 53, 123-150.

Fama, Eugene F., and Kenneth R. French, 1993, Common risk factors in the returns on stocks and bonds, Journal of Financial Economics 33, 3-56.

Froot, Kenneth A., David S. Scharfstein, and Jeremy C. Stein, 1993, Risk Management: Coordinating corporate investment and financing policies, Journal of Finance 48, 1629-1658.

Gilchrist, Simon, and Charles Himmelberg, 1995, Evidence on the role of cash flow for investment, Journal of Monetary Economics 36, 541-572.

Huang, Rocco, 2008, Evaluating the real effect of bank branching deregulation: Comparing contiguous counties across US state borders, Journal of Financial Economics 87, 678-705.

Jayaratne, Jith, and Philip E. Strahan, 1996, The finance-growth nexus: Evidence from bank branch deregulation, Quarterly Journal of Economics 111, 639-670.

Jayaratne, Jith, and Philip E. Strahan, 1998, Entry restrictions, industry evolution, and dynamic efficiency: Evidence from commercial banking, Journal of Law and Economics 41, 239-274.

Jegadeesh, Narasimhan, and Shreidan Titman, 1993, Returns to buying winners and selling losers: Implications for stock market efficiency, Journal of Finance 48, 65-92.

Kashyap, Anil, Owen Lamont, and Jeremy C. Stein, 1994, Credit conditions and the cyclical behavior of inventories, Quarterly Journal of Economics 109, 565-592.

Krozner, Randall S., Luc Laeven, and Daniela Klingebiel, 2007, Banking crises, financial dependence, and growth, Journal of Financial Economics 84, 187-228.

Larrain, Borja, 2006, Do banks affect the level and composition of industrial volatility? Journal of Finance 61, 1897-1925.

Levine, Ross, 2005, Finance and growth: Theory and evidence, in Philippe Aghion and Steven N. Durlauf, eds.: Handbook of Economic Growth, Vol. 1A (Elsevier, Amsterdam).

McConnell, Margaret M., and Gabriel Perez-Quiros, 2000, Output fluctuations in the United States: What has changed since the early 1980s? American Economic Review 90, 1464-1476.

Morgan, Donald P., Bertrand Rime, Philip E. Strahan, 2004, Bank integration and state business cycles, Quarterly Journal of Economics 119, 1555-1584.

Pastor, Lubos, and Pietro Veronesi, 2003, Stock valuation and learning about profitability, Journal of Finance 58, 1749-1789.

Raddatz, Claudio, 2006, Liquidity needs and vulnerability to financial underdevelopment, Journal of Financial Economics 80, 677-722.

Rajan, Raghuram G., and Luigi Zingales, 1995, What do we know about capital structure? Some evidence from international data, Journal of Finance 50, 1421-1460.

Rajan, Raghuram G., and Luigi Zingales, 1998, Finance dependence and growth, American Economic Review 88, 559-586.

Ramey, Garey, and Valerie A. Ramey, 1995, Cross-country evidence on the link between volatility and growth, American Economic Review 85, 1138-1151.

Rice, Tara, and Philip E. Strahan, 2009, Does credit competition affect small-firm finance? Journal of Finance, forthcoming.

Rountree, Brian, James P. Weston, and George Allayannis, 2008, Do investors value smooth performance? Journal of Financial Economics 90, 237-251.

Smith, Clifford W., and René Stulz, 1985, The determinants of firms' hedging policies, Journal of Financial and Quantitative Analysis 20, 391-405.

Stebunovs, Viktors, 2006, Finance as a barrier to entry: U.S. bank deregulation and volatility, Working paper, Boston College.

Stock, James H., and Mark W. Watson, 2002, Has the business cycle changed and why? in Mark Gertler and Kenneth Rogoff, eds.: NBER Macroeconomics Annual 2002 (MIT Press, Cambridge, MA).

Stulz, René, 1990, Managerial discretion and optimal financing policies, Journal of Financial Economics 32, 263-292.

Strahan, Philip E., 2003, The real effects of U.S. banking deregulation, Federal Reserve Bank of St. Louis Review 85, 111-128.

Thesmar, David, and Mathias Thoenig, 2009, Contrasting trends in firm volatility: Theory and evidence, Working paper, CEPR.


Appendix 1: Dates of Deregulation of Interstate Bank Entry by State


State Year
ALASKA 1982
ALABAMA 1987
ARKANSAS 1989
ARIZONA 1986
CALIFORNIA 1987
COLORADO 1988
CONNECTICUT 1983
DISTRICT OF COLUMBIA 1985
DELAWARE 1988
FLORIDA 1985
GEORGIA 1985
HAWAII 1995
IOWA 1991
IDAHO 1985
ILLINOIS 1986
INDIANA 1986
KANSAS 1992
KENTUCKY 1984
LOUISIANA 1987
MASSACHUSETTS 1983
MARYLAND 1985
MAINE 1978
MICHIGAN 1986
MINNESOTA 1986
MISSOURI 1986
MISSISSIPPI 1988
MONTANA 1993
NORTH CAROLINA 1985
NORTH DAKOTA 1991
NEBRASKA 1990
NEW HAMPSHIRE 1987
NEW JERSEY 1986
NEW MEXICO 1989
NEVADA 1985
NEW YORK 1982
OHIO 1985
OKLAHOMA 1987
OREGON 1986
PENNSYLVANIA 1986
RHODE ISLAND 1984
SOUTH CAROLINA 1986
SOUTH DAKOTA 1988
TENNESSEE 1985
TEXAS 1987
UTAH 1984
VIRGINIA 1985
VERMONT 1988
WASHINGTON 1987
WISCONSIN 1987
WEST VIRGINIA 1988
WYOMING 1987

Table I
Firm Distribution across States

The sample includes annual observations between 1976 and 1998 of Compustat firms classified in the manufacturing (SIC codes 2000 to 3999), wholesale (SIC codes 5000 to 5199), and retail trade (SIC codes 5200 to 5999) sectors. We consider only firms with non-missing information 2 years prior and 2 years after deregulation of interstate bank entry in the state where firms are headquartered.

State Number of firms Pct. of total firms Number of observations Pct. of total observations Average years per firm
ALABAMA 13 0.6% 239 0.6% 18.4
ARKANSAS 8 0.4% 155 0.4% 19.4
ARIZONA 18 0.8% 284 0.7% 15.8
CALIFORNIA 312 13.7% 5,011 12.6% 16.1
COLORADO 41 1.8% 664 1.7% 16.2
CONNECTICUT 96 4.2% 1,585 4.0% 16.5
DISTRICT OF COLUMBIA 2 0.1% 46 0.1% 23.0
FLORIDA 84 3.7% 1,467 3.7% 17.5
GEORGIA 35 1.5% 661 1.7% 18.9
IOWA 14 0.6% 288 0.7% 20.6
IDAHO 3 0.1% 69 0.2% 23.0
ILLINOIS 137 6.0% 2,575 6.5% 18.8
INDIANA 42 1.8% 785 2.0% 18.7
KANSAS 13 0.6% 210 0.5% 16.2
KENTUCKY 12 0.5% 180 0.5% 15.0
LOUISIANA 8 0.4% 144 0.4% 18.0
MASSACHUSETTS 106 4.7% 1,809 4.6% 17.1
MARYLAND 34 1.5% 547 1.4% 16.1
MAINE 4 0.2% 59 0.1% 14.8
MICHIGAN 76 3.3% 1,449 3.7% 19.1
MINNESOTA 87 3.8% 1,574 4.0% 18.1
MISSOURI 46 2.0% 874 2.2% 19.0
MISSISSIPPI 2 0.1% 46 0.1% 23.0
MONTANA 1 0.0% 17 0.0% 17.0
NORTH CAROLINA 46 2.0% 845 2.1% 18.4
NEBRASKA 5 0.2% 98 0.2% 19.6
NEW HAMPSHIRE 9 0.4% 144 0.4% 16.0
NEW JERSEY 139 6.1% 2,379 6.0% 17.1
NEW MEXICO 2 0.1% 28 0.1% 14.0
NEVADA 5 0.2% 35 0.1% 7.0
NEW YORK 315 13.9% 5,104 12.9% 16.2
OHIO 122 5.4% 2,294 5.8% 18.8
OKLAHOMA 13 0.6% 244 0.6% 18.8
OREGON 15 0.7% 279 0.7% 18.6
PENNSYLVANIA 108 4.8% 2,065 5.2% 19.1
RHODE ISLAND 11 0.5% 183 0.5% 16.6
SOUTH CAROLINA 10 0.4% 179 0.5% 17.9
TENNESSEE 20 0.9% 380 1.0% 19.0
TEXAS 125 5.5% 2,179 5.5% 17.4
UTAH 15 0.7% 232 0.6% 15.5
VIRGINIA 43 1.9% 819 2.1% 19.0
VERMONT 3 0.1% 47 0.1% 15.7
WASHINGTON 31 1.4% 538 1.4% 17.4
WISCONSIN 39 1.7% 785 2.0% 20.1
WEST VIRGINIA 1 0.0% 15 0.0% 15.0
WYOMING 1 0.0% 14 0.0% 14.0
Total 2,272   39,624   17.4

Table II
Summary Statistics: Volatility and Access to External Finance

Panel A summarizes the properties of the firm-level volatility measures over the sample period, 1976-1998. Vol( Y ) represents the volatility of Y as defined in Section II.B. Production is measured as Sales plus the change in inventories, and Employment is measured as total number of employees. These variables are taken from Compustat, and nominal variables are deflated using the GDP deflator. R_{0} are firm-level excess returns over the Treasury Bill. R_{1} is the residual of running a market-model regression of firm-level monthly excess returns on the excess return on the market portfolio and an intercept. R_{2} is the residual obtained by expanding the market-model regression to include the size and value factors of Fama and French (1993) and the momentum factor of Jegadeesh and Titman (1993). Data on returns and factors are taken from CRSP and Kenneth French's Data Library at Dartmouth. Panel B summarizes firm-level proxies for Bank Dependence. The size-based measure of bank dependence equals 1 for firm-year observations when the firm assets are below the state-year median and 0 otherwise. The rating-issues-based measure equals 1 for firms that did not issue bonds or commercial paper and did not have credit ratings from 1970 to 1994. Data on bond and commercial paper issues are from Mergent Fixed Income Security Database and Moody's DRS. Ratings are taken from S&P as reported in Compustat.

Panel A
  Observations Mean Median Standard Deviation
Vol(Production) 36,883 0.152 0.092 0.198
Vol(Sales) 36,905 0.132 0.081 0.159
Vol(Cash Flow) 29,679 0.353 0.196 0.489
Vol(Employment) 35,623 0.134 0.077 0.196
Vol(R0) 28,270 0.119 0.104 0.070
Vol(R1) 28,270 0.107 0.091 0.069
Vol(R2) 28,270 0.103 0.087 0.066


Panel B
  Bank Dependence Proxied by Small Firm Size, Observations Bank Dependence Proxied by Small Firm Size, Percentage Bank Dependence Proxied by No Debt Issues and Ratings, Observations Bank Dependence Proxied by No Debt Issues and Ratings, Percentage
Not Dependent on Bank Finance 22,960 57.9% 14,849 37.5%
Dependent on Bank Finance 16,664 42.1% 24,775 62.5%

Table III
The Effect of Banking Deregulation on Firm-Level Volatility

This table reports the results from the regression:

\begin{displaymath} Vol\left( {Y_{ijt} } \right)=\alpha _i +\mu _t +\delta Integration_{jt} +\beta X_{ijt-1} +\gamma Vol(Z_{jt} )+\varphi Vol(W_{kt} )+\varepsilon _{ijt} \end{displaymath}

Where i indexes firms, j denotes states, k indexes industries, and t represents years, and the sample period is 1976-1998. Vol(Y ) is the volatility, as defined in Section II.B, of the growth rate of variable Y. We report results for 4 dependent variables: (1) Vol(Production), where Production is measured as Sales plus the change in inventories; (2) Vol(Sales); (3) Vol(Cash Flow); and (4) Vol(Employment), where Employment is measured as total number of employees. Integration is a dummy variable that equals 1 for state-year observations after laws that deregulated interstate banking were passed. X is a vector of 3 lagged firm controls: log(Sales); EBITDA/Assets, where EBITDA is earnings before income, tax, depreciation, and amortization; and Tangible Assets/Assets, where Tangible Assets are defined as in Almeida and Campello (2007). Vol(Z) is the volatility of state per capita income.Vol(W) is the volatility of industry sales, aggregated at the 3-digit SIC code level. Firm- and time-fixed effects are not reported. Firm-level variables are taken from Compustat, and nominal variables are deflated using the GDP deflator. State per capita income is from the Bureau of Economic Analysis. Robust standard errors clustered by state-year are reported in brackets below the coefficients. Significance: * significant at 10%; ** significant at 5%; *** significant at 1%.
Dependent variable: Vol(Production) (1) Vol(Sales) (2) Vol(Cash Flow) (3) Vol(Employment) (4)
Integration -0.004 -0.006 -0.027** -0.007*
Integration standard errors [0.004] [0.003] [0.012] [0.004]
log(Salest-1) -0.021*** -0.021*** -0.004 -0.022***
log(Salest-1) standard errors [0.003] [0.003] [0.007] [0.003]
EBITDAt-1 / Assetst-1 -0.130*** -0.088*** -1.236*** -0.091***
EBITDAt-1 / Assetst-1 standard errors [0.025] [0.014] [0.064] [0.018]
Tangiblet-1 / Assetst-1 -0.011 -0.001 -0.194*** -0.006
Tangiblet-1 / Assetst-1 standard errors [0.021] [0.016] [0.048] [0.027]
Vol(State p.c. Income) -0.004 -0.030 -0.089 0.053
Vol(State p.c. Income) standard errors [0.111] [0.102] [0.381] [0.127]
Vol(Industry Sales) 0.109*** 0.103*** 0.087*** 0.044***
Vol(Industry Sales) standard errors [0.012] [0.009] [0.029] [0.010]
Observations 36,883 36,905 29,679 35,623
Firms 2,270 2,270 2,209 2,262
R-squared 0.02 0.03 0.04 0.01

Table IV
The Effect of Deregulation on Firm Volatility for Bank-Dependent Firms

This table reports the results from the regression:

\begin{displaymath} Vol\left( {Y_{ijt} } \right)=\alpha _i +\mu _t +\delta Integration_{jt} +\lambda Bank\;Dependence_{ijt} +\varphi Integration_{jt} \times Bank\mbox{ }Dependence_{ijt} +\beta X_{ijt-1} +\gamma Vol(Z_{jt} )+\phi Vol(W_{kt} )+\varepsilon _{ijt} \end{displaymath}

Where i indexes firms, j denotes states, k indexes industries, and t denotes years, and the sample period is 1976-1998. Vol(Y ) is the volatility, as defined in Section II.B, of the growth rate of a generic variable Y. We report results for 4 dependent variables: (1) Vol(Production), where Production is measured as Sales plus the change in inventories; (2) Vol(Sales); (3) Vol(Cash Flow); and (4) Vol(Employment), where Employment is measured as total number of employees. Integration is a dummy variable that equals 1 for state-year observations after laws that deregulated interstate banking were passed. In Panel A, Bank Dependence equals 1 for firm-year observations when the firm assets are below the state-year median and 0 otherwise. In Panel B, Bank Dependence equals 1 for firms that did not issue bonds or commercial paper and did not have credit ratings from 1970 to 1994. X represents lagged firm controls: log(Sales); EBITDA/Assets and Tangible Assets/Assets. Vol(Z) is the volatility of state per capita income.Vol(W) is the volatility of industry sales, aggregated at the 3-digit SIC code level. Firm- and time-fixed effects are not reported. Firm-level variables are taken from Compustat, except for bond and commercial paper issues, which are from FISD and Moody's DRS. State per capita income is from the Bureau of Economic Analysis. Nominal variables are deflated using the GDP deflator. Robust standard errors clustered by state-year are reported in brackets below the coefficients. Significance: * significant at 10%; ** significant at 5%; *** significant at 1%.


Panel A - Bank Dependence Proxied by No Debt Issues and Ratings
Dependent variable: Vol(Production) (1) Vol(Sales) (2) Vol(Cash Flow) (3) Vol(Employment) (4)
Integration 0.004 0.003 0.016 0.002
Integration standard errors [0.005] [0.004] [0.013] [0.005]
Bank Dependence - - - -
Integration \times Bank Dependence -0.013*** -0.013*** -0.069*** -0.013***
Integration \times Bank Dependence standard errors [0.004] [0.003] [0.011] [0.004]
log(Salest-1) -0.021*** -0.021*** -0.006 -0.023***
log(Salest-1) standard errors [0.003] [0.003] [0.007] [0.003]
EBITDAt-1 / Assetst-1 -0.130*** -0.088*** -1.245*** -0.091***
EBITDAt-1 / Assetst-1 standard errors [0.025] [0.014] [0.064] [0.018]
Tangiblet-1 / Assetst-1 -0.007 0.004 -0.157*** -0.001
Tangiblet-1 / Assetst-1 standard errors [0.021] [0.016] [0.050] [0.027]
Vol(State p.c. Income) 0.000 -0.026 -0.080 0.057
Vol(State p.c. Income) standard errors [0.112] [0.102] [0.377] [0.128]
Vol(Industry Sales) 0.109*** 0.103*** 0.088*** 0.044***
Vol(Industry Sales) standard errors [0.012] [0.009] [0.029] [0.010]
Observations 36,883 36,905 29,679 35,623
Firms 2,270 2,270 2,209 2,262
R-squared 0.02 0.03 0.04 0.01


Panel B - Bank Dependence Proxied by Small Firm Size
Dependent variable: Vol(Production) (1) Vol(Sales) (2) Vol(Cash Flow) (3) Vol(Employment) (4)
Integration 0.003 0.002 -0.003 -0.002
Integration [0.004] [0.003] [0.013] [0.004]
Bank Dependence -0.004 -0.001 0.043*** 0.004
Bank Dependence [0.005] [0.004] [0.016] [0.005]
Integration \times Bank Dependence -0.016*** -0.017*** -0.065*** -0.011**
Integration \times Bank Dependence [0.004] [0.003] [0.013] [0.004]
log(Salest-1) -0.023*** -0.023*** -0.003 -0.023***
log(Salest-1) [0.004] [0.003] [0.008] [0.003]
EBITDAt-1 / Assetst-1 -0.129*** -0.087*** -1.249*** -0.091***
EBITDAt-1 / Assetst-1 standard errors [0.025] [0.014] [0.064] [0.018]
Tangiblet-1 / Assetst-1 -0.006 0.005 -0.163*** -0.002
Tangiblet-1 / Assetst-1 standard errors [0.021] [0.016] [0.049] [0.027]
Vol(State p.c. Income) -0.006 -0.032 -0.099 0.052
Vol(State p.c. Income) standard errors [0.111] [0.101] [0.379] [0.127]
Vol(Industry Sales) 0.109*** 0.103*** 0.086*** 0.044***
Vol(Industry Sales) standard errors [0.012] [0.009] [0.029] [0.010]
Observations 36,883 36,905 29,679 35,623
Firms 2,270 2,270 2,209 2,262
R-squared 0.03 0.03 0.04 0.01

Table V
Robustness Checks

This table reports the results from the following two regressions:

\begin{displaymath} Vol\left( {Y_{ijt} } \right)=\alpha _i +\mu _t +\delta Integration_{jt} +\beta X_{ijt-1} +\gamma Vol(Z_{jt} )+\varphi Vol(W_{kt} )+\varepsilon _{ijt} \end{displaymath}


\begin{displaymath} Vol\left( {Y_{ijt} } \right)=\alpha _i +\mu _t +\delta Integration_{jt} +\lambda Bank\;Dependence_{ijt} +\varphi Integration_{jt} \times Bank\mbox{ }Dependence_{ijt} +\beta X_{ijt-1} +\gamma Vol(Z_{jt} )+\phi Vol(W_{kt} )+\varepsilon _{ijt} \end{displaymath}

Where i indexes firms, j denotes states, k indexes industries, and t denotes years, and the sample period is 1976-1998. Vol(Y ) is the volatility, as defined in Section II.B, of the growth rate of a generic variable Y. We report results for 4 dependent variables: (1) Vol(Production), where Production is measured as Sales plus the change in inventories; (2) Vol(Sales); (3) Vol(Cash Flow); and (4) Vol(Employment), where Employment is measured as total number of employees. \alpha _{i} denotes firm-fixed effects and \mu _{t} denotes year-fixed effects. All regressions include year-fixed effects, except for Panel B. Integration is a dummy variable that equals 1 for state-year observations after laws that deregulated interstate banking were passed. Bank Dependence equals 1 for firms that did not issue bonds or commercial paper and did not have credit ratings from 1970 to 1994. Data on debt issues are from FISD and Moody's DRS. X represents lagged firm controls: log(Sales); EBITDA/Assets and Tangible Assets/Assets. Vol(Z) is the volatility of state per capita income.Vol(W) is the volatility of industry sales, aggregated at the 3-digit SIC code level. Firm-level variables are taken from Compustat, except for bond and commercial paper issues, which are from FISD and Moody's DRS. State per capita income is from the Bureau of Economic Analysis. Nominal variables are deflated using the GDP deflator. Robust standard errors clustered by state-year are reported in brackets below the coefficients. Significance: * significant at 10%; ** significant at 5%; *** significant at 1%.


Panel A - Dependent variable is volatility, measured by the squared value of deviations from the firm-level conditional mean
Dependent var.: Vol(Production) (1) Vol(Production) (2) Vol(Sales) (3) Vol(Sales) (4) Vol(Cash Flow) (5) Vol(Cash Flow) (6) Vol(Employment) (7) Vol(Employment) (8)
Integration 0.003 0.010 -0.005 0.003 -0.062* 0.002 -0.006 0.005
Integration standard errors [0.008] [0.008] [0.003] [0.003] [0.033] [0.037] [0.007] [0.010]
Integration \times Bank Dependence   -0.011*   -0.011***   -0.104***   -0.016*
Integration \times Bank Dependence standard errors   [0.006]   [0.003]   [0.033]   [0.010]
Observations 36,883 36,883 36,905 36,905 29,679 29,679 35,623 35,623
Firms 2,270 2,270 2,270 2,270 2,209 2,209 2,262 2,262
R-squared 0.01 0.01 0.02 0.02 0.02 0.02 0 0
                 


Panel B - Dependent variable is volatility, measured as the standard deviation five years before and five years after banking deregulation
Dependent var.: Vol(Production) (1) Vol(Production) (2) Vol(Sales) (3) Vol(Sales) (4) Vol(Cash Flow) (5) Vol(Cash Flow) (6) Vol(Employment) (7) Vol(Employment) (8)
Integration -0.007 0.005 -0.005 0.010* 0.051*** 0.048*** 0.004 0.015
Integration standard errors [0.007] [0.007] [0.006] [0.005] [0.013] [0.015] [0.007] [0.009]
Integration \times Bank Dependence   -0.017***   -0.024***   0.005   -0.017**
Integration \times Bank Dependence standard errors   [0.006]   [0.005]   [0.023]   [0.008]
Observations 2,929 2,929 2,929 2,929 2,649 2,649 2,868 2,868
Firms 1,469 1,469 1,469 1,469 1,420 1,420 1,454 1,454
R-squared 0.03 0.03 0.04 0.05 0.08 0.08 0.04 0.04


Panel C - Bank integration is measured as the share of deposits held by multi-state banks in each state and year (This measure is instrumented by the deregulation dummy)
Dependent var.: Vol(Production) (1) Vol(Production) (2) Vol(Sales) (3) Vol(Sales) (4) Vol(Cash Flow) (5) Vol(Cash Flow) (6) Vol(Employment) (7) Vol(Employment) (8)
Integration -0.040 -0.018 -0.051 -0.030 -0.232* -0.134 -0.063 -0.041
Integration standard errors [0.044] [0.043] [0.032] [0.031] [0.120] [0.118] [0.040] [0.039]
Integration \times Bank Dependence   -0.030***   -0.030***   -0.158***   -0.031***
Integration \times Bank Dependence standard errors   [0.010]   [0.008]   [0.031]   [0.010]
Observations 36,881 36,881 36,903 36,903 29,634 29,634 35,614 35,614
Firms 2,268 2,268 2,268 2,268 2,164 2,164 2,253 2,253
R-squared 0.02 0.02 0.02 0.02 0.03 0.04 0.01 0.01


Panel D - Sample excludes states with simultaneous intrastate and interstate banking deregulation
Dependent var.: Vol(Production) (1) Vol(Production) (2) Vol(Sales) (3) Vol(Sales) (4) Vol(Cash Flow) (5) Vol(Cash Flow) (6) Vol(Employment) (7) Vol(Employment) (8)
Integration -0.006 0.004 -0.006 0.004 -0.024** 0.014 -0.007 0.004
Integration standard errors [0.004] [0.005] [0.004] [0.004] [0.012] [0.014] [0.004] [0.006]
Integration \times Bank Dependence   -0.014***   -0.014***   -0.062***   -0.017***
Integration \times Bank Dependence standard errors   [0.004]   [0.004]   [0.012]   [0.004]
Observations 30,652 30,652 30,672 30,672 24,605 24,605 29,636 29,636
Firms 1,896 1,896 1,896 1,896 1,843 1,843 1,890 1,890
R-squared 0.03 0.03 0.03 0.03 0.04 0.04 0.01 0.02


Panel E - Sample excludes firms in California and New York
Dependent var.: Vol(Production) (1) Vol(Production) (2) Vol(Sales) (3) Vol(Sales) (4) Vol(Cash Flow) (5) Vol(Cash Flow) (6) Vol(Employment) (7) Vol(Employment) (8)
Integration 0.002 0.012* -0.001 0.009* -0.02 0.036** -0.008 0.001
Integration standard errors [0.006] [0.006] [0.005] [0.005] [0.015] [0.016] [0.006] [0.006]
Integration \times Bank Dependence   -0.014***   -0.015***   -0.089***   -0.014***
Integration \times Bank Dependence standard errors   [0.004]   [0.003]   [0.012]   [0.005]
Observations 27,540 27,540 27,558 27,558 22,690 22,690 26,565 26,565
Firms 1,645 1,645 1,645 1,645 1,608 1,608 1,639 1,639
R-squared 0.02 0.02 0.02 0.02 0.04 0.04 0.01 0.01


Panel F - Sample includes all firms with non-missing observations between 1976 and 1998
Dependent var.: Vol(Production) (1) Vol(Production) (2) Vol(Sales) (3) Vol(Sales) (4) Vol(Cash Flow) (5) Vol(Cash Flow) (6) Vol(Employment) (7) Vol(Employment) (8)
Integration -0.002 0.006 -0.002 0.006* -0.030*** 0.009 -0.005 0.004
Integration standard errors [0.004] [0.004] [0.004] [0.004] [0.011] [0.013] [0.004] [0.005]
Integration \times Bank Dependence   -0.013***   -0.012***   -0.062***   -0.014***
Integration \times Bank Dependence standard errors   [0.004]   [0.003]   [0.010]   [0.004]
Observations 60,939 60,939 60,987 60,987 45,723 45,723 58,442 58,442
Firms 7,288 7,288 7,290 7,290 6,046 6,046 7,104 7,104
R-squared 0.02 0.02 0.02 0.02 0.03 0.04 0.01 0.01


Panel G - Controls include Banking Concentration as measured by Herfindahl-Hirshman index of deposits by state
Dependent var.: Vol(Production) (1) Vol(Production) (2) Vol(Sales) (3) Vol(Sales) (4) Vol(Cash Flow) (5) Vol(Cash Flow) (6) Vol(Employment) (7) Vol(Employment) (8)
Integration -0.004 0.004 -0.006* 0.003 -0.027** 0.016 -0.007* 0.002
Integration standard errors [0.004] [0.005] [0.003] [0.004] [0.012] [0.013] [0.004] [0.005]
Integration \times Bank Dependence   -0.013***   -0.013***   -0.069***   -0.013***
Integration \times Bank Dependence standard errors   [0.004]   [0.003]   [0.011]   [0.004]
Observations 36,883 36,883 36,905 36,905 29,679 29,679 35,623 35,623
Firms 2,270 2,270 2,270 2,270 2,209 2,209 2,262 2,262
R-squared 0.02 0.02 0.03 0.03 0.04 0.04 0.01 0.01


Panel H - Controls include age (as measured by years from IPO) instead of log(Sales), and firm leverage, measured as total debt divided by total assets
Dependent var.: Vol(Production) (1) Vol(Production) (2) Vol(Sales) (3) Vol(Sales) (4) Vol(Cash Flow) (5) Vol(Cash Flow) (6) Vol(Employment) (7) Vol(Employment) (8)
Integration -0.008* -0.001 -0.008** -0.003 -0.037*** -0.009 -0.006 -0.002
Integration standard errors [0.005] [0.005] [0.003] [0.004] [0.013] [0.015] [0.004] [0.005]
Integration \times Bank Dependence   -0.011**   -0.009***   -0.049***   -0.006
Integration \times Bank Dependence standard errors   [0.004]   [0.003]   [0.012]   [0.004]
Observations 30,592 30,592 30,604 30,604 25,221 25,221 30,025 30,025
Firms 2,015 2,015 2,015 2,015 1,939 1,939 2,007 2,007
R-squared 0.02 0.02 0.02 0.02 0.04 0.04 0.01 0.01

Table VI
Banking Deregulation and Countercyclical Borrowing

This table reports the results from the regression:

\begin{displaymath} \begin{array}{c} \Delta \log \left( {Notes\mbox{ }Payable_{ijt} } \right)=\alpha _i +\mu _t +\delta Integration_{jt} +\theta Cyclicality_t + \ \lambda Bank\mbox{ }Dependence_{ijt} +\varphi Integration_{jt} \times Bank\mbox{ }Dependence_{ijt} \ +\nu Integration_{jt} \times Bank\mbox{ }Dependence_{ijt} \times Cyclicality_t \ +\beta X_{ijt-1} +\varepsilon _{ijt} \ \end{array}\end{displaymath}

Where i indexes firms, j denotes states, k indexes industries, and t denotes years, and the sample period is 1976-1998. The dependent variable is the log change in Notes Payable. Estimations in Columns (4) to (6) do not include year-fixed effects. Integration is a dummy variable that equals 1 for state-year observations after interstate banking deregulation laws were passed. The table reports results with and without two proxies for bank dependence. The size-based measure of Bank Dependence equals 1 for firm-year observations when the firm assets are below the state-year median and 0 otherwise. The rating-issues-based measure of Bank Dependence equals 1 for firms that did not issue bonds or commercial paper and did not have credit ratings from 1970 to 1994. X represents lagged firm controls: log(Sales); EBITDA/Assets and Tangible Assets/Assets. Firm-level variables are taken from Compustat, except for bond and commercial paper issues, which are from FISD and Moody's DRS. Nominal variables are deflated using the GDP deflator. Robust standard errors clustered by state-year are reported in brackets below the coefficients. Significance: * significant at 10%; ** significant at 5%; *** significant at 1%.


Cyclicality measured as: Cyclicality measured as Sales_{t}/ Assets_{t-1 }, No Bank Dependence Proxy (1) Cyclicality measured as Sales_{t}/ Assets_{t-1 }, Bank dependence measured as No Debt Issues and Ratings (2) Cyclicality measured as Sales_{t}/ Assets_{t-1 }, Bank dependence measured as Small Firm Size (3) Cyclicality measured as \Delta log(Real GDP), No Bank Dependence Proxy (4) Cyclicality measured as \Delta log(Real GDP), Bank dependence measured as No Debt Issues and Ratings (5) Cyclicality measured as \Delta log(Real GDP)_{ }, Bank dependence measured as Small Firm Size (6)
Integration -0.007 -0.027 -0.032 0.046 0.046 0.046
Integration standard errors [0.061] [0.059] [0.058] [0.036] [0.036] [0.036]
Cyclicality 0.127*** 0.139*** 0.148*** 3.837*** 3.825*** 3.819***
Cyclicality standard errors [0.029] [0.027] [0.026] [0.533] [0.532] [0.531]
Bank Dependence     0.114**     0.092*
Bank Dependence standard errors     [0.055]     [0.053]
Integration \times Cyclicality -0.019 0.056* 0.047 -1.577 -0.275 -0.257
Integration \times Cyclicality standard errors [0.028] [0.030] [0.030] [0.966] [1.127] [1.051]
Integration \times Bank Dependence \times Cyclicality   -0.100*** -0.125***   -2.395** -3.941***
Integration \times Bank Dependence \times Cyclicality standard errors   [0.024] [0.029]   [0.954] [0.975]
log(Sales_{t-1}) -0.044* -0.053** -0.047* -0.059** -0.064*** -0.059**
log(Sales_{t-1 }) standard errors [0.024] [0.024] [0.025] [0.024] [0.024] [0.025]
EBITDA_{t-1 }/ Assets_{t-1 } 0.678*** 0.672*** 0.670*** 0.768*** 0.770*** 0.770***
EBITDA_{t-1 }/ Assets _{t-1 standard errors} [0.099] [0.098] [0.098] [0.104] [0.104] [0.104]
Tangible_{t-1 }/ Assets_{t-1 } 0.507*** 0.542*** 0.532*** 0.624*** 0.666*** 0.678***
Tangible_{t-1 }/ Assets _{t-1 standard errors} [0.194] [0.194] [0.191] [0.179] [0.181] [0.180]
Observations 16,953 16,953 16,953 16,953 16,953 16,953
Firm 1,875 1,875 1,875 1,875 1,875 1,875
R-squared 0.02 0.02 0.02 0.01 0.01 0.01

Table VII
Banking Deregulation and Idiosyncratic Stock Return Volatility

This table presents the results from the following regression:

\begin{displaymath} Vol\left( {R_{it} } \right)=\alpha _i +\mu _t +\delta Integration_{jt} +\lambda Bank\mbox{ }Dependence_{ijt} +\varphi Integration_{jt} \times Bank\mbox{ }Dependence_{ijt} +\beta X_{ijt-1} +\varepsilon _{ijt} \end{displaymath}

Where i indexes firms, j denotes states, and t denotes year, and the sample period is 1976-1998. R_{0} are firm-level raw excess returns over the Treasury Bill. R_{1} is the residual of running a firm-level market-model regression of monthly excess returns on the excess return on the market portfolio and an intercept. R_{2} is the residual obtained by expanding the market model regression to include the size and value factors of Fama and French (1993) and the momentum factor of Jegadeesh and Titman (1993). Vol( R_{it} ) is the volatility of R_{it}, measured as its standard deviation in year t. Integration is a dummy variable that equals 1 for state-year observations after interstate banking deregulation laws were passed. The size-based measure of Bank Dependence equals 1 for firm-year observations when the firm assets are below the state-year median and 0 otherwise. The rating-issues-based measure of Bank Dependence equals 1 for firms that did not issue bonds and commercial paper and did not have credit ratings from 1970 to 1994. X represents lagged firm controls: log(Sales); EBITDA/Assets and Tangible Assets/Assets. Firm-level variables are taken from Compustat, except for bond and commercial paper issues, which are from FISD and Moody's DRS. Nominal variables are deflated using the GDP deflator. Robust standard errors clustered by state-year are reported in brackets below the coefficients. Significance: * significant at 10%; ** significant at 5%; *** significant at 1%.


Dependent variable Bank dependence measured as: Vol(R_{0}) No Bank Depend. Proxy (1) Vol(R_{0}) No Debt Issues and Ratings (2) Vol(R_{0}) Small Firm Size (3) Vol(R_{1}) No Bank Depend. Proxy (4) Vol(R_{1}) No Debt Issues and Ratings (5) Vol(R_{1}) Small Firm Size (6) Vol(R_{2}) No Bank Depend. Proxy (7) Vol(R_{2}) No Debt Issues and Ratings (8) Vol(R_{2}) Small Firm Size (9)
Integration -0.001 0.005*** 0.002 -0.001 0.003* 0.001 -0.001 0.002 0.001
Integration standard errors [0.001] [0.002] [0.001] [0.001] [0.002] [0.001] [0.001] [0.002] [0.001]
Bank Dependence     0.011***     0.011***     0.010***
Bank Dependence standard errors     [0.002]     [0.002]     [0.002]
Integration \times   -0.008*** -0.006***   -0.006*** -0.006***   -0.005*** -0.005**
Integration \times standard errors   [0.002] [0.002]   [0.001] [0.002]   [0.001] [0.002]
log(Salest-1) -0.013*** -0.013*** -0.011*** -0.013*** -0.013*** -0.012*** -0.013*** -0.013*** -0.011***
log(Salest-1) standard errors [0.001] [0.001] [0.001] [0.001] [0.001] [0.001] [0.001] [0.001] [0.001]
EBITDAt-1 / Assetst-1 -0.071*** -0.071*** -0.073*** -0.081*** -0.081*** -0.083*** -0.076*** -0.076*** -0.078***
EBITDAt-1 / Assetst-1 standard errors [0.009] [0.009] [0.009] [0.008] [0.008] [0.008] [0.008] [0.008] [0.008]
Tangiblet-1 / Assetst-1 -0.047*** -0.044*** -0.045*** -0.049*** -0.047*** -0.047*** -0.045*** -0.043*** -0.044***
Tangiblet-1 / Assetst-1 standard errors [0.007] [0.007] [0.007] [0.007] [0.007] [0.007] [0.007] [0.007] [0.007]
Observations 28,167 28,167 28,167 28,167 28,167 28,167 28,167 28,167 28,167
Firms 1,637 1,637 1,637 1,637 1,637 1,637 1,637 1,637 1,637
R-squared 0.09 0.10 0.10 0.07 0.07 0.07 0.06 0.06 0.06


Figure 1


Figures 1 shows the evolution of these residual returns before and after deregulation. Following banking deregulation, the median idiosyncratic component of stock return volatility declines for three years. Campbell et al. (2001) and Pastor and Veronesi (2003) document a long-term increase in firm volatility starting in the 1960s. This trend temporarily reverses in the years after deregulation in the 1980s, consistent with the drop in idiosyncratic stock return volatility in Figures 1.

Figure 1. Volatility of idiosyncratic returns estimated from a market model. The figure shows the median idiosyncratic component of firm volatility for the three years before and after interstate bank entry deregulation. Volatility is measured as the standard deviation of residuals from a market model.



Figure 2

Figures  2 shows the evolution of these residual returns before and after deregulation. Following banking deregulation, the median idiosyncratic component of stock return volatility declines for three years. Campbell et al. (2001) and Pastor and Veronesi (2003) document a long-term increase in firm volatility starting in the 1960s. This trend temporarily reverses in the years after deregulation in the 1980s, consistent with the drop in idiosyncratic stock return volatility in Figures 2.

Figure 2. Volatility of idiosyncratic returns estimated from a four-factor model. The figure shows the median idiosyncratic component of firm volatility for the three years before and after interstate bank entry deregulation. Volatility is measured as the standard deviation of residuals from a four-factor model.



Figure 3

Figure 3 shows the evolution of median stock return volatility before and after deregulation splitting the sample by bank dependence. Compared to the group of firms with access to public debt markets, excess return volatility is higher for firms that are bank-dependent and experiences a steeper decline after banking deregulation. This evidence suggests that the stabilizing effects from banking deregulation may have benefited more those firms that depend to a greater extent on banks for their financing needs.

Figure 3. Volatility of idiosyncratic returns and bank dependence. The figure shows the median idiosyncratic component of firm volatility for the three years before and three years after interstate bank entry deregulation. Volatility is measured as the standard deviation of residuals from a market model (Panel A) and a four-factor model (Panel B). The sample of firms is divided according to bank dependence. A firm is classified as being bank dependent if it did not access public debt markets between 1970 and 1994.



Footnotes

* We thank Andres Almazan, Robert Avery, Allen Berger, Dan Covitz, Eric Engstrom, Arturo Galindo, Rocco Huang, Borja Larrain, Ugur Lel, Greg Nini, Andrew Powell, Rafael Repullo, and seminar participants at the Federal Reserve Board, the Latin American Finance Network, the AFA Annual meetings in San Francisco, and the MFA Annual meetings in San Antonio for valuable comments. The usual disclaimer applies. The authors' views do not necessarily reflect those of the Board of Governors of the Federal Reserve or the Federal Reserve System. Return to Text
1. See Levine (2005) for a survey. Return to Text
2. The existence of corporate hedging activity suggests that firms try to mitigate some sources of risk and volatility. See, for example, Froot, Scharfstein, and Stein (1993). Return to Text
3. In addition, the effect of wider access to banking finance on volatility may depend on the relative strength of credit supply and credit demand shocks (Morgan, Rime, and Strahan (2004)). Return to Text
4. In a related branch of the literature, Krozner, Laeven, and Klingebiel (2007) study the impact of banking crises across sectors depending on their degree of dependence on external finance. Return to Text
5. In fact, excluding the market crash of 1987 in Figure 4 of Campbell et al. (2001), idiosyncratic volatility trended down during the 1980s. Return to Text
6. Interstate branching was permitted nationwide with the Riegle-Neal Interstate Banking and Branching Efficiency Act, which became effective in June 1997. Montana and Texas took advantage of a clause in the Act and opted out at an earlier date. The table in Appendix 1 lists deregulation dates by state. Return to Text
7. Huang (2007) exploits the variation of neighboring counties across state borders to understand the effects of this episode of banking deregulation. Return to Text
8. The earliest year for which we have location information comes after the first bank entry deregulation agreement was implemented (1982). Although, some firms might have changed their location between these dates, the number is likely to be small. Using information from Compact Disclosure, we determine that less than 10 percent of firms covered by this database changed the state where they were headquartered between 1988 and 2005. Return to Text
9. Due to this selection criterion, our results are explained by changes in the volatility of existing firms rather than changes in the volatility of entering and exiting firms. All results are robust to include firms that exit before deregulation and enter after deregulation. Return to Text
10. We exclude firms headquartered in Delaware and South Dakota due to their incorporation and consumer finance regulations, respectively. In addition, as Hawaii did not deregulate before 1994, we also exclude firms located in that state. Return to Text
11. The discontinuity at deregulation of our empirical design suggests that our results are not simply a reflection of firms becoming more stable as they age, and we conduct additional checks of our baseline results. Return to Text
12. Following Almeida and Campello (2007), tangible assets are defined as: Tangible=0.715\times \mbox{Re}ceivables+0.547\times Inventory+0.535\times Capital, where Receivables is Compustat item #2, Inventory is item #3, and Capital is item #8. EBITDA is defined as Compustat item #13. Return to Text
13. Nominal variables are deflated using the GDP deflator. Comin and Mulani (2005) test different price deflators, and do not find any significant differences in their volatility measures. Return to Text
14. Sales is Compustat item #12, employment is item #29, cash flow equals item #18 plus item #14, and inventories is item #3. Return to Text
15. Formally, Vol\left( {Y_{kt} } \right)=\left( {Y_{kt} -\bar {Y}_k } \right)^2 represents the volatility of the log change in Industry Sales, for each 3-digit SIC sector k. \bar {Y}_k is computed separately for the periods before and after bank entry deregulation. Return to Text
16. This measure is computed using the absolute value of the error term, after estimating equation (1) for the log changes in per capita income at the state level. Return to Text
17. Gilchrist and Himmelberg (1995) use size to define financially constrained firms. These authors and Kashyap, Lamont, and Stein (1994) also use access to the bond market to define firms that face greater informational asymmetries in the financial markets. Calomiris, Himmelberg, and Wachtel (1995) describe the characteristics of commercial paper issuers. Return to Text
18. Notes payable is Compustat item #206. Return to Text
19. Since the external finance proxy for debt issues and ratings is time-invariant, it drops from estimating equation (3) using firm-fixed effects. Return to Text
20. Results are very similar when we measure bank-dependence using a proxy related to firm size. Return to Text
21. In other results not reported here, we replace for each state the actual deregulation date with a randomly generated date. The bank integration proxy and its interaction with the bank dependence measure become not significant. Return to Text
22. In contrast to the publicly-traded firms' evidence, Davies et al. (2006), find a 23 percent decrease in employment growth volatility for privately held firms between 1978 and 2001. Return to Text
23. This paper complements existing evidence on the link between financial markets and financial innovation, on the one hand, and volatility, on the other. In the case of the United States, for example, Dynan, Elmendorf, and Sichel (2006) argue that financial innovation may help explain the stabilization of economic activity in the mid 1980s. An interesting avenue for exploration consists of studying the interaction between banking deregulation and financial innovation (in the form of more efficient pricing of risk, for example). Return to Text
24. See, for example, Acemoglu et al. (2003). Return to Text

This version is optimized for use by screen readers. Descriptions for all mathematical expressions are provided in LaTex format. A printable pdf version is available. Return to Text