Capital Market Conditions and the Financial and Real Implications of Cash Holdings* Aziz Alimov University of Arizona Wayne Mikkelson University of Oregon This draft: October 18, 2009 Abstract We investigate the financial and real implications of corporate cash holdings over different capital market conditions in the period 1972-2008. A recent theoretical work implies that the value of corporate liquidity varies with the supply of external capital and should be higher during market-wide liquidity shocks. We find that for an average firm a marginal dollar in cash holdings has the same value in high and low liquidity periods. However, the additional dollar of cash is more valuable for research-intensive firms, which are more likely to be shut out of capital markets during financial crises. We also study the real implications of cash holdings across the market-wide liquidity cycle. Our goal is to study whether and how cash reserves affect a firm‟s performance and investment relative to its industry rivals in regular and adverse market conditions. Our results indicate that having more cash relative to industry rivals allows a firm to achieve a higher rate of sales growth and spend more on capital and inventory expenditures. However, we do not find that larger cash reserves relative to industry rivals lead to more investment and better performance when capital market liquidity is low. Overall, we conclude that in a market downturn large holdings of cash do not give a firm a strategic advantage over its industry rivals. Keywords: Cash holdings; Firm Value; Performance; Capital Market Conditions. Preliminary and Incomplete Comments Welcome * Contact information: Alimov ( aalimov@email.arizona.edu) and Mikkelson (wmikkels@lcbmail.uoregon.edu) 1. Introduction The ongoing financial crisis and the resultant freezing of credit markets suddenly emphasized the advantage of having a stockpile of cash and liquid investments for firm operations and its performance. The credit crisis has also had a significant effect on how investors view and value corporate holdings of cash and liquid investments. For example, a recent article in The Economist (Nov 20, 2008) states “How times change. Not long ago companies with cash piles were assailed by corporate activists to return money to shareholders. Nowadays it is only a slight exaggeration to say that the more cash that investors see in a firm‟s coffers, the happier they are. “ The statement above suggests that capital market conditions are important determinants of the value that investors assign to corporate liquid assets. Economic theory also suggest that capital market conditions, by affecting the supply of external financing, can have significant effects on firms‟ investment and financial decisions, and might even determine which firm survives tough times. The cash reserves can protect a firm and even give it a competitive advantage over its industry rivals during liquidity shocks. In fact, the typical rationale given for the recent tendency of U.S. firms to accumulate high levels of liquid assets is that a stockpile of cash gives a firm the flexibility to internally finance value-increasing investments when external capital is unavailable or costly (Passov, 2003). Yet the financial and real implications of firm internal liquidity during periods of market-wide liquidity shocks have been little explored empirically. The question is also clearly relevant for monetary policy. The goal of this paper is therefore twofold. First, we examine the impact of variations in capital market conditions on the value of corporate liquidity. Second, we investigate whether high levels of cash lead firms to gain a competitive advantage over their rivals with less cash during limited supply of capital. 1 A period of market-wide shock to the supply of capital is a natural event to examine the value that investors assign to corporate liquid assets and whether and how cash holdings affect competitive policies and outcomes. During normal times, cash holdings can have both positive and negative valuation effects due to benefits and costs associated with cash holdings. On the one hand, a benefit of holding cash is the ability to finance valuable projects. Hence, cash reserves should be particularly valuable for firms that face problems in obtaining external financing, or financially constrained firms. However, as Almeida, Campello and Weisbach (2004) point out, cash holdings are costly for constrained firms because higher cash require a reduction in current period investment. In addition, Jensen (1986) argues that cash holdings could be costly for shareholders if excess cash induces managers to waste it on value-destroying projects. Therefore, in good times cash holdings can either enhance or reduce firm value, making it difficult to ascertain the value of cash. In contrast, studying periods of market-wide shortage of capital allow us to shed light on which of these channels drive the value of corporate cash. We hypothesize that, in rational markets, the value of corporate cash should increase during market-wide liquidity squeeze because internal reserves of liquidity enable a firm to avoid financial distress and to readily fund investment. Further, if fluctuations in capital market liquidity indeed play an important role in the way market views and values cash holdings, one can expect the value of cash to be greater for firms that should be hit the hardest by the negative shock to market liquidity. As a result, we expect cash reserves to be particularly valuable during financial crises for smaller and more research-intensive firms. Small and research-intensive firms face greater financing frictions and could even be rationed out of the capital markets due to severe informational problems and low collateral. 2 Studying whether investors assign a different value to corporate liquidity across different capital market conditions is only one goal of our study. Our second main objective is to investigate whether firms with higher cash levels are better buffered against the market liquidity shocks thus gain a competitive advantage over their rivals holding less cash. To this end, we investigate whether cash-rich firms invest more and perform better than their cash-poor peers during periods of low market liquidity. We test our conjectures using a panel data set containing information for U.S. public firms covering the period 1976-2008. We split our sample period into times of high and low market liquidity, and compare the financial and real implications of cash holdings under those two market circumstances. The definition of what constitutes a period of high or low capital market liquidity is critical for our study. Our main classification method is based on the spread between rates on commercial paper and Treasury bills (CP-Bill yield spread). The CP-Bill yield spread is commonly used in empirical studies and Krishnamurthy (2002) argues the spread should reflect aggregate demand for liquidity because commercial paper is relative illiquid compared to Treasury bills. To ensure that the CP-Bill yield spread does not contain any credit risk, we orthogonalize the CP-Bill spread with respect to the spread between rates in Baa and Aaa rates corporate bonds. We find that, after controlling for other determinants of financing decisions, the average firms raises less capital during periods of low market liquidity. Our main analysis proceeds in two steps. First, we follow the methodology developed by Faulkender and Wang (2006) and examine the effect of an additional dollar of cash on market value of equity during different capital market conditions. We find that, for an average firm, the value of a marginal dollar in cash holdings does not depend on capital market conditions. However, we find different results when we partition our sample according to ex-ante needs for 3 external financing. Specifically, we find that a dollar of cash is worth more for researchintensive firms, which are more likely to be denied external financing when they need it most . Next, we probe the economic sources of the time-varying value of cash. Because the obvious link between liquidity constraints and firm value is through the impact of liquidity on investment and subsequent performance, we analyze the impact of cash on sales growth, operating performance and investment spending during the recession. Our goal is to determine whether cash reserves allow firms to gain market share at the expense of cash poor industry rivals during market-wide shortage of capital. We examine whether higher cash reserves relative the industry average during the high low capital market liquidity periods lead to higher levels of investment and subsequent shortterm higher sales growth and better performance. We find evidence that having more cash reserves, allows a firm to spend more on capital and inventory expenditures. As a result, cash rich firms are able to increase their sales growth rates relative to industry rivals. However, we do not find that large cash reserves relative to industry rivals lead to more investment and better performance during tight capital market conditions. Overall, we conclude that in a capital market downturn having large cash reserves does not give a firm a strategic advantage and thereby enhance subsequent performance. Our paper contributes to a recent literature that shows that the relation between stock prices and cash holdings. Two studies demonstrate that investors seem to recognize the potential benefits and inefficiencies associated with high cash balances and value corporate liquidity. Faulkender and Wang (2006) find that that the marginal value of cash is higher for financially constrained firms than for unconstrained firms and for firms smaller cash balances. Dittmar and Mahrt-Smith (2007) find that the marginal value of cash is lower for poorly governed firms. 4 However, each of these studies assumes that investors' valuation of corporate liquidity is the same over different stages of the business cycle. Our paper instead focuses on the role played by the business cycle in the valuation of a dollar of cash. We also contribute to research on the effect of a firm‟s cash holdings on its investment and performance. Harford (1999) firms with excess cash holdings are more likely to undertake acquisitions and these acquisitions are more likely to be value-decreasing. Mikkelson and Partch (2003) examine firms that consistently hold a high cash balance, and find that high cash reserves has no adverse effect on the operating performance of their sample of firms. Other related evidence is provided by Opler and Titman (1994) and Campello (2003), who focus on the effects of leverage, rather than cash holdings, and find evidence of benefits of financial conservatism in an industry or economy wide downturn. The remainder of the study is organized as follows. The next section discusses the potential relation between the business cycle and the financial and real implications of cash holdings. Section 3 describes our sample and descriptive statistics. The empirical results are reported in Section 4. We present our concluding remarks in Section 5. 2. Background and theoretical predictions We start by reviewing the theoretical connection between the state of the capital markets and the value and use of corporate cash holdings. In a Modigliani and Miller (1958) world of frictionless capital markets this question is uninteresting because firms can always raise external financing at no cost. The connection between cash holdings and capital market conditions, which determine the supply of external financing, starts with the fact that capital market imperfections such as asymmetric information, moral hazard, and agency problems lead to a wedge between 5 the costs of internal and external funds. Seminal articles on this subject include Myers (1977), Myers and Majluf (1984), and Jensen (1986). Bernanke and Gertler (1989) further suggest that the market imperfections are likely to exacerbate during market downturns, therefore increasing the cost of external financing. In addition, Stiglitz and Weiss (1981) further argue lower credit quality firms are also more likely to be rationed out of the credit markets in times of tight credit. In sum, theory suggests during poor capital market conditions, firms are likely to receive external financing only at substantially higher cost and the lowest credit quality firms cannot raise financing at all. The existence of market frictions, which exacerbate in bad economic times, provides the main rationale for firms to hold cash. When financial market conditions are poor, a stockpile of cash and liquid investments gives a firm the flexibility to internally finance future valueincreasing investments. Richard Passov (2003), the treasurer of Pfizer, states that the potential inability to raise external financing during poor capital market conditions is the primary reason why knowledge-based firms often hold large cash positions. In addition, firms with large cash holdings can use cash to finance aggressive actions against their cash-poor rivals, such as cutting . prices, opening new facilities, and even acquiring rivals at fire-sale prices etc . The notion that particular financial policies can result in a competitive advantage to firms within an industry has been documented by Chevalier (1995) and Pulvino (1998). Anecdotal evidence also abounds. For example, an article in the Wall Street Journal (Feb 17, 2009) describes how Oracle is using its large cash reserves to buy distressed smaller rivals. In sum, the arguments above therefore implies that in periods of low market liquidity larger cash reserves should result in more investment and give a firm a competitive advantage over rivals whose cash reserves are not sufficient to fund profitable investment opportunities. 6 We also predict the value of corporate liquidity to increase in bad times because internal reserves of liquidity enable a firm to avoid financial distress and to readily fund investment. Firms holding greater cash reserves in a period of low market liquidity should invest more and eventually experience better performance than firms whose cash reserves are insufficient to support profitable investment. The foregoing discussion assumes that managers‟ and stockholders‟ interests are aligned and that the additional investment facilitated by cash reserves is value-increasing. However, as Jensen (1986) points out, large cash holdings can lead to value-destroying investment decisions because of agency problems between shareholders and managers. In financial crises, therefore, the decline in internally generated cash flow raises the costs of financing precisely at a time when the amount of profitable investment opportunities might be shrinking, providing a constraint on managers‟ spending Furthermore, this constraint is desirable if managers‟ interests are not well-aligned with stockholders and managers have personal incentives to spend available cash, behavior suggested by Jensen‟s (1986) theory of agency costs of free cash flow. Thus, this argument implies that internally generated cash flow varies in the same direction as changes in a firm‟s profitable investment opportunities. As a result, managers are disciplined to align cash flow and investment spending, leading to the creation of value. The reduced spending discipline argument implies that large cash reserves enable managers to avoid reductions in investment at times when profitable investment opportunities contract. Moreover, cash reserves allow managers to avoid the scrutiny of outside providers of capital, scrutiny that can also serve to monitor managers. Similar to the hypothesis of lower financing costs, firms with greater cash reserves will invest more during an industry downturn and the investment of cash rich firms will be less sensitive to the downturn in revenues. Reduced 7 spending discipline implies that managers invest too much and spend in ways that reduce firm value. This implies that, relative to industry rivals, cash rich firms eventually perform worse than their industry rivals and that their poorer performance is related to the additional investment made possible by cash reserves. 3. Sample and Data 3.1 Sample Construction and Variable Definitions Our initial sample includes all U.S. firms covered by the CRSP and quarterly COMPUSTAT databases during the period 1976-2008. We start our sample in 1976 because necessary quarterly data items exist on Compustat from 1976. The advantage of using the quarterly data is that it allows us to closely align the timing of observations with the timevariation in capital market conditions. Because capital market conditions can change dramatically during the fiscal year, the annual data might lead to incorrect classification of low and high market liquidity periods and thus introduce noise into our tests. However, the drawback is that quarterly data are not available before 1976 and quarterly R&D data are not available before 1989. Consistent with the previous literature, we exclude financial institutions (SIC codes 6000 to 6999), utilities (SIC codes 4900 to 4999) and non-for-profit organizations (SIC code 8000 to 9000). We require that firm-quarters to have non-missing data for cash and cash equivalents, non-missing common shareholders‟ equity and operating income, and sufficient data to calculate our variables. We convert all data to real values in 1972 dollars using the consumer price index (CPI). Because we study whether the level of cash reserves affects firm investment behavior and performance, we restrict the sample to firms with active cash and investment policies. 8 Specifically, we require that firms have at least $0.1 million (in 1972 dollars) in cash, and to have positive operating assets and plant, properties and equipment items in their balance sheets. 3.2. Identification of inter-temporal and cross-sectional liquidity constraints In our tests we exploit intertemporal and cross-sectional contrasts in firms‟ ability to access capital market to examine the impact of capital market conditions on the value and use of corporate cash. We expect that during periods of tight external financing, such as financial crises, the value that investors assign to corporate cash reserves increase. We consider several measures of market tightness, all commonly used in empirical studies. Our main measure of market-wide liquidity is computed monthly as the spread between rates on 3 month commercial (nonfinancial) paper and 3 month Treasury bills (CP-Bills yield spread). Krishnamurthy (2002) argues that while the commercial paper is relatively safe instrument, the paper is illiquid relative to Treasury bills. Consequently, the rate on commercial paper includes a liquidity premium and the difference in rates should reflect investors‟ aggregate demand to own liquid securities. Kashyap, Stein and Wilcox (1993) provide evidence supporting the idea that CP-Bills yield spread measures capital supply in the economy. Therefore, based on the results of the Kashyap et al. paper, we use the yield spread as a proxy for availability of external capital or overall market liquidity. We obtain information on monthly commercial paper and T-bills rates from the Federal Reserve‟s website. We compute the spread at the quarterly frequency by averaging monthly paper-bills spreads over the firm‟s fiscal quarter. Figures 1 graph the CP-Bills yield spread over the period 1975-2009. As have been previously noted in the literature (e.g. Kashyap, Stein and Wilcox, 1993), the spread is negative related (with a slight lead) to the overall economic activity. 9 Specifically, the figure shows that the spread rises before the NBER-defined recessions and drops before expansions. This pattern in the CP-Bills spread is consistent with the argument that the availability of financing affects overall economic growth. However, one must consider the possibility that the CP-Bills yield spread may contain the default risk, which also varies with the state of the economy. To remove the default risk, we orthogonalize CP-Bills yield spread with respect to the commonly used proxy for the default risk, the spread between rates on BAA and AAA rated corporate bonds. The data on bond rates is also obtained from the Federal Reserve‟s website. The figure shows that orthogonalizing the spread to the default risk does not affect any of the properties of original series. Nevertheless, we use the orthogonalized spread as our measure of market liquidity. We classify market liquidity in a particular quarter as a high or low liquidity market based on the orthogonalized CP-Bills spread. Low liquidity quarters are quarters in which the yield spread is above its sample median and high liquidity quarters are all other quarters. To provide additional evidence on the effect of market liquidity fluctuations, we also study the cross-sectional differences across firms. If firms do not have equal access to external financial markets, the sensitivity of the value of cash to market liquidity will differ across firms. As discussed in our review of the literature, finance theory predicts that the access to capital should become more difficult for firms facing ex-ante greater financing frictions. We use these predictions to examine whether the valuation of internal funds for firms with greater financing frictions exhibit more cyclical behavior. We consider two cross-sectional variables as proxy for ex ante high financing frictions, firm size and the research and development (R&D) intensity. First, following large literature on capital market imperfections (e.g. Almeida et al., 2004), we use firm size as a proxy for financial frictions. Small firms are typically young, less 10 known, and more vulnerable to capital market imperfections and, thus, are more likely to face financing frictions in obtaining external capital. We define a firm as financially unconstrained if the book value of its assets lies above the 67th percentile and constrained if its assets lie below the 33rd percentile. We obtain similar results if we use firm age, payout ratio and credit rating as a measure of firms' financing constraints. Investment in R&D is a major investment for a large number of firms, especially firms in operating in the technology and science-based industries. It is widely accepted that R&D investments are more difficult to finance with external funds due to limited collateral value and potentially greater information problem (Hall, 2002). Therefore, one can expect a shortage of external financing to have particularly strong implications for investment and performance of research-intensive firms. We define research-intensive firms as firms with R&D-to-sales ratio above 5%. 3.3 Sample Summary Statistics Our final sample includes 250,840 firm-quarter observations. Table 1 reports the mean and medians of our main variables. To provide univariate statistics on the impact of capital market conditions on firm characteristics, we divide the sample into high, neutral and lowliquidity subperiods. There are 133,410 firm-quarter observations during high market liquidity and 117,430 firm- quarter observations during low market liquidity quarters. Firms in both sub-periods have similar levels of book assets and sales. There is slight difference in cash holdings across the liquidity cycle. On average, cash and marketable securities represent 41 percent of total assets during high liquidity periods and 33 percent of assets during 11 low liquidity. The median firm holds 12 and 9 percent of its total assets in the form of cash and marketable securities during the high and low liquidity periods, respectively. The average and median change in cash holdings is positive in both sub-samples, suggesting that, on average, firms are increasing their cash holdings over time. We observe that performance of our sample firms is cyclical and worsens substantially during the poor market conditions. The average next-4 quarter rate of sales growth for firms during high liquidity quarters is 43 percent and 39 percent during low liquidity quarters. The stock returns and marketto-book ratio also differ significantly between the two periods. The quarterly raw stock return for an average firm is 6% during high liquidity and only 1% during low liquidity. The average market-to-book assets ratio is 1.97 during high-liquidity periods and 1.85 during low-liquidity periods. Finally, both research and development expenditures and capital expenditures seem to be fairly stable across both subperiods. 3.4 Financing patterns across different market conditions To establish the validity of our proxy for the supply of external capital, we first examine whether the fluctuations in capital market conditions affect firms’ financing activity. To do so, we estimate the relation between firm financing activity and market liquidity conditions using the following multivariate equation investment equation: Financing i,t a + b1Tight Capital +b 3Log of sales i,t -1 Net Assetsi,t -1 +b 4 Operating Income Net Assetsi,t -1 i,t -1 +b 5 M arket Assetsi,t -1 + b 6Leverage i,t -1 + e i,t Book Assetsi,t -1 (1) where Financing is the sum of new equity and debt issued during the quarter. Tight Capital is an indicator variable that takes a value of one if the CP-Bill yield spread is below the sample median. Other variables serve as control for the firm-specific determinants of firms‟ 12 decision to raise external debt or equity. For example, the ratio of market-to-book assets controls for the amount and quality of firm growth opportunities. We estimate the regression for the whole sample, and separately for external financing dependent and independent firms The results of the regression are presented in Table 2. The coefficient on Tight Capital is negative for small and research intensive firms, indicating that, consistent with our conjecture, firms with greater information problems find it harder to raise capital during bad times. In contrast, larger firms, which have information problems, appear to increase their issuance of debt and equity during periods of low market liquidity. 4. Empirical Results 4.1. The market value of cash To measure the impact of capital market conditions on the value of cash holdings, we ideally would use the total market value of a firm. However, because we do not have data on the market value of firm debt, we can only examine changes in market value of equity instead. To this end, we use the methodology in Faulkender and Wang (2006) and estimate the relation between an unexpected change in cash holdings and a contemporaneous change in market value of equity over the one-year period, year t=-1 to year t. Because our objective is to determine whether variation in market conditions affects the marginal value of cash holdings, we extend Faulkender and Wang‟s model by including a binary indicator variable for capita market tightness as an additional explanatory variable and interacting change in cash holdings with the capital tightness binary variable. The change in market value of equity is measured by the excess return for firm i during fiscal year t less the return of stock i‟s benchmark portfolio during fiscal year t. The benchmark portfolio is one of the 25 Fama and French portfolios formed on size and 13 book-to-market. We examine excess returns in order to control for time-varying risk factors that may impact a firm‟s rate of return. To quantify the importance of the cyclical variations on cash, we model the excess returns for each firm as a function of change in contemporaneous cash holdings normalized by beginning-of-period equity value, a measure of the market conditions and a number of control variables. The control variables are as in Faulkender and Wang (2006) and include firm profitability, investments, and financing. The following equation describes the regression, which we also estimate separately for financially constrained and unconstrained firms. Cash Excess Return i, t t a + b 1 b4 + b6 Earnings i, t + b5 M Vi, t -1 Dividends i, t M Vi, t -1 + b 10 Leverage i, t -1 i, t M Vi, t -1 + b 2 Re cession i, t b 3 Re cession i, t * Net Assetsi, t + b7 M Vi, t -1 Cash i, t -1 Cash i, t M Vi, t -1 M Vi, t -1 + b 11 + b6 R & Di, t M Vi, t -1 + b7 M Vi, t -1 i, t M Vi, t -1 Interest Expenses i, t + b 8 Leverage i, t -1 + b 9 Cash i, t -1 NewFinancin g i, t Cash M Vi, t -1 (1) Cash i, t M Vi, t -1 + e i, t where ΔXi,t indicates the change in the variable X of firm i from the period from t−1 to t+1. The dependent variable is firm i‟s stock return over the fiscal quarter t minus the Fama and French (1993) size and book-to-market matched portfolio return. The dependent variables include the change in cash plus marketable securities; the change in earnings before extraordinary items; the change in the change in noncash assets; the change in R&D expenses; the change in interest expenses; the change in common dividends; Leverage (total debt over the market value of assets) at the beginning of the year t=-1; cash plus marketable securities at the beginning of the year t=1; net financing (net new equity issues +net new debt issues). When R&D expense is missing, we set its value to zero. Tight Capital is a dummy variable that equals one if the fiscal quarter is in the low liquidity period period, and zero otherwise. All dependent variables except leverage are scaled by market value of equity at the beginning of the quarter. We winsorize all 14 independent variables at the 1% and 99% level to reduce the influence of outliers. The standard errors are corrected for heteroskedasticity and clustered at the firm level since individual returns may not be independent. The methodology used in this study essentially represents a long-run event study. In this case, the event is the unexpected change in cash holdings, while the event window is defined as the tow-year period. Our main variable of interest is the interaction between change in cash and Recession. The coefficient on this variable represents the difference in the value shareholders place on an additional dollar of cash in low liquidity periods compared to high liquidity periods, thus indicating the impact of market downturns on the value of cash holdings. We expect the coefficient on change in the cash holdings to be significantly higher in recessions, capturing the effect of worsening economic conditions on the change in firm value resulting from a one dollar change in cash holdings. We also include the indicator variable in the regression to ensure that the estimated coefficient of the interaction term is due to the interaction, and not due to a market condition itself. We report the results of this regression in Table 3. The first column reports the results for the entire sample without including the recession variables. The coefficients and significance on all variables are largely similar to those found in Faulkender and Wang‟s (2006). One exception is the coefficient on change in R&D expenditures, which is not significant in our sample. In contrast, Faulkender and Wang find that changes in R&D are positively associated with contemporaneous stock returns. We then include a Tight Capital dummy and the interaction of a Tight Capital dummy and change in cash holdings. We report the results in the second column for the entire sample. The results do not support our conjecture that cash is more valuable during low market liquidity 15 periods for all firms. The coefficient on the interaction of cash and a Tight Capital dummy for is not statistically different from zero. The coefficient estimates imply that the marginal value of cash is roughly the same in all market conditions. Columns 3-6 shows results of the regression by partitioning the data based on R&D intensity and firm size. The null hypothesis is that firm characteristics should not affect the value of cash. Conversely, the alternative hypothesis predicts that the value of changes in cash holdings of R&D intensive and smaller firms to be more affected by adverse market conditions if these firms face greater difficulties in raising external capital than larger and less R&D intensive firms. only the results are for research-intensive are consistent with our expectations. Specifically, the coefficient on Tight Capital is 0.139 and significant at the1 percent level. This finding indicates that the value of an additional dollar is indeed higher during adverse capital market conditions for more research-intensive firms, who precisely at the same time find it harder to access capital markets. 4.2 Effect of Cash on Corporate Performance We will now investigate whether firms with higher cash holdings are more likely to experience better performance in the economic downturns than other firms. We measure firm performance by sales growth and changes in operating performance over the one-year window. We define operating performance as operating income divided by non-cash assets (ROA). The change in operating performance is the change in operating return on assets from the previous to current fiscal year. We measure sales and ROA relative to its industry rivals by removing the industry means of these variables. We define industries using the Fama and French (1997) 48industry classification. The constructed industry-adjusted performance variables can be thought 16 as a change in a firm‟s sales and profitability relative to its competitors. All independent variables are also measured relative to industry rivals by deducting the industry-means. In effect, we use the industry average cash holding as a benchmark to measure a firm required or optimal cash holdings. The industry-adjustment also serves as a control for any unobserved industry effects. To estimate a reasonable benchmark for industry adjustment, we require industries to have at least ten firms. We estimate the relation between firm performance and corporate liquidity using the following regression: Sales Growthi,t a + b1 + b5Leverage i,t -1 +b 6 Cash i,t -1 Cash i,t -1 + b 2 Tight Capital + b 3 Tight Capital* + b 4 Log of sales i,t -1 Net Assetsi,t -1 Net Assetsi,t -1 Operating Income Net Assetsi,t -1 ROAi,t - ROAi,t -1 a + b1 + b5Leverage i,t -1 +b 6 i,t +b 5 Capital Expenditure i,t + e i,t Net Assetsi,t -1 (3) Cash i,t -1 Cash i,t -1 + b 2 Tight Capital + b 3 Tight Capital* + b 4 Log of sales i,t -1 Net Assetsi,t -1 Net Assetsi,t -1 Operating Income Net Assetsi,t -1 i,t +b 5 Capital Expenditure i,t + e i,t Net Assetsi,t -1 All independent variables are winsorized at the 1% and 99% level to reduce the influence of outliers. The estimations correct the error structure for heteroskedasticity and within-firm error clustering. We report the estimates for sales growth in Table 4 for the whole sample, and separately for firm sub-samples. Panel A displays results for the next four quarters and Panel B for the next eight quarters. We use two different time windows because it is not clear when the effect of adverse market conditions will be reflected in the data. We find that the estimate of the coefficient on industry-adjusted cash is positive and statistically significant for larger and less research intensive. This result indicates that during 17 normal capital market conditions larger firms are able to use their cash holdings to gain market share at the expense of their industry rivals. However, the coefficient on industry-cash adjusted cash holdings and its interaction with the tight capital dummy for larger and less researchintensive firms is not significant, indicating that having a larger level of cash during adverse market conditions does not lead to greater short-term sales gains. In contrast, for smaller and research- intensive firms the coefficient on the interaction of cash and adverse capital market conditions is negative and significant at the five percent level. This results indicates that smaller and research intensive firms with greater cash reserves relative to industry rivals tend to lose market share during adverse capital market conditions. We report the estimates for change in operating performance in Table 5 for the whole sample, and separately for firm sub-samples. Panel A displays results for the next four quarters and Panel B for the next eight quarters. The main result here is that the coefficient on industrycash adjusted cash holdings and its interaction with the tight capital dummy are not significant in all specifications, indicating that having a larger level of cash in market downturns does not lead to improvement in firm operating performance at the expense of industry rivals. In contrast, the estimate of the coefficient on industry-adjusted cash is either slightly negative or zero across all specifications. This result indicates that even during normal capital market conditions cash holdings does not lead to better operating performance relative to industry rivals. 4.3 Effect of Cash on Corporate Investment The sales and change in operating performance regressions do not allow us to identify the specific channels through which cash holdings can affect firm behavior during difference phases of the market-wide liquidity cycle. In this section we examine one of the potential channels, the 18 investment channel. As was previously discussed, a potential value-enhancing benefit of cash holdings is the possibility to finance profitable projects that may not get financed through external capital markets. Therefore, by examining the effect of a firm's cash reserves on investment decisions, we can better understand the results reported in the previous sections. We estimate the relation between investment and corporate liquidity using the standard investment equation: Investmenti,t Cash i,t -1 Cash i,t -1 a + b1 + b 2 Tight Capital + b 3 Tight Capital* + b 4 Log of sales i,t -1 Net Assetsi,t -1 Net Assetsi,t -1 Net Assetsi,t -1 +b 6 Operating Income Net Assetsi,t -1 i,t -1 +b 5 M arket Assetsi,t -1 + b 7Leverage i,t -1 + e i,t Book Assetsi,t -1 (4) where Investment represents either capital expenditures or investment in inventory. Table 6 reports the results of the regressions using capital expenditures and Table 7 using inventory expenditures as dependent variables. We present results for the investment over the next four quarters and eight quarters. The results are similar to those found for the impact of cash reserves on firm performance over different capital market conditions. Specifically, we find that the estimate of the coefficient on industry-adjusted cash is positive and statistically significant. The estimate in Panel B of Table 6 suggests that every dollar of abnormal reserves enhances capital investment over the next eight quarters by 15 cents. By the same token, the estimate in Panel B of Table 7 suggests that every dollar of abnormal reserves enhances investment in inventory over the next eight quarters by 3 cents. However, the estimate of the coefficient on cash and Tight Capital dummy is insignificant for both types of investment at both time intervals. The results thus indicate corporate investments are increasing in firm cash reserves by the same amount during regular and unfavorable capital market conditions. 19 5. Conclusion We investigate whether the market value of firm cash reserves differs across different capital market conditions and the economic source for the time-variation in the value of cash. Managers justify sizable holdings of cash reserves as a precaution against a decrease in cash flow that would interfere with undertaking profitable investment. Evidence in support of managers‟ justification is important because studies that do not condition on the state of the economy. We examine the contemporaneous relation between change in cash reserves and the firm value across high and low market liquidity states. We find evidence suggesting that for an average firm the value of an additional dollar in cash holdings is the same regardless of the state of capital markets. However, when partition the data according to firms ex-ante needs for external financing, we find that more cash is more valuable for research intensive firms during periods of low market liquidity. We also examine the hypothesis that larger cash reserves support investment in the recession and thereby enhance subsequent performance. Performance is measured by the changes in operating return on assets and annual sales growth. We find that cash reserves during the high market liquidity periods lead to higher sales growth and slightly lower operating performance relative to industry rivals. However, we do not find that larger cash reserves can a firm to increase and boost its performance at the expense of industry rivals during a recession. Our findings are important in that they provide evidence of the effect of accumulating cash reserves in anticipation of an economic downturn. Our results indicate that even during a recession, firms need to guard against the potential for excess cash reserves leading to investment that harms future performance. 20 References Almeida, Heitor, Murillo Campello, and Michael S. Weisbach, 2004, The cash flow sensitivity of cash, Journal of Finance 59, 1777–1804. Bernanke, B., and A. Blinder, 1988, Credit, Money, and Aggregate Demand, American Economic Review 78, 435-439. Bernanke, B., and M. Gertler, 1989, Agency Costs, Net Worth, and Business Fluctuations, American Economic Review 79, 14-31. Bernanke, B., and M. Gertler, 1995, Inside the Black Box: The Credit Channel of Monetary Policy Transmission, Journal of Economic Perspectives 9, 27-48. Blanchard, Lopez-de-Silanes, and A. Shleifer, 1994, What do firms do with cash windfalls?, Journal of Financial Economics 36, 337-360. Campello, Murillo, 2003, Capital structure and product markets interactions: Evidence from the business cycles, Journal of Financial Economics 68, 353-378. Dimitrov, V., and S. Tice, 2006, Corporate Diversification and Credit Constraints: Real Effects across the Business Cycle, Review of Financial Studies 4, 1465-1498. Dittmar, A., & Mahrt-Smith, J. (2007). Corporate Governance and the Value of Cash Holdings. Journal of Financial Economics 83, 599-634. Faulkender, M., & Wang, R. (2006). Corporate Financial Policy and the Value of Cash. Journal of Finance 61, 1957-1990. Fazzari, S., G. Hubbard, and B. Petersen, 1988, Financing Constraints and Corporate Investment, Brookings Papers on Economic Activity 1, 141-195, 30 Gertler, M., and S. Gilchrist, 1994, Monetary Policy, Business Cycles, and the Behavior of Small Manufacturing Firms, Quarterly Journal of Economics 2, 309-340. Harford, J., 1999. Corporate cash reserves and acquisitions. Journal of Finance 54, 1969–1997. Jensen, M., 1986, Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers, American Economic Review 76, 323-329. Kashyap, A. K., O. Lamont, and J. Stein, 1994, “Credit Conditions and the Cyclical Behavior of Inventories,” Quarterly Journal of Economics 109, 565-592. Kashyap, A., J. Stein, and D. Wilcox, 1993, Monetary Policy and Credit Conditions: Evidence from the Composition of External Finance, American Economic Review 83, 78-98. Korajczyk, Robert A., and Amnon Levy, 2003, Capital structure choice: Macroeconomic conditions and financial constraints, Journal of Financial Economics 68, 75–109. Krishnamurthy, Arvind, 2002, The Bond/Old-Bond Spread, Journal of Financial Economics 66, 463-506. 21 Mikkelson, W., Partch, M., 2003. Do persistent large cash reserves hinder performance? Journal of Financial and Quantitative Analysis 38, 275–294. Myers, S. C., and N. S. Majluf, 1984, Corporate Financing and Investment Decisions when Firms have Information that Investors do Not Have, Journal of Financial Economics 13, 187-221. Opler, T., Pinkowitz, L., Stulz, R., Williamson, R., 1999, The Determinants and Implications of Corporate Cash Holdings, Journal of Financial Economics 52, 3-46. Opler, T., and S. Titman, 1994, Financial Distress and Corporate Performance, Journal of Finance, 49, 1015–1040. Passov, R., 2003, How Much Cash Does your Company Need?, Harvard Business Review, November, 1-8. Perez-Quiros, G., and A. Timmermann, 2000, „Firm Size and Cyclical Variations in Stock Returns”, Journal of Finance, 55, 1229–1262. Pinkowitz, L., & Williamson, R. (2002). What is a Dollar Worth? The Market Value of Cash Holdings. Working paper, Georgetown University. Pulvino, T., 1998, Do asset fire-sales exist? An empirical investigation of commercial aircraft ransactions, Journal of Finance 53, 939-978. Shleifer A. and R. Vishny, 1992 Liquidation values and debt capacity: A market equilibrium approach, Journal of Finance 47, 1343-1366. Sterngold, J., 1995, Facing the next recession without fear; newly lean corporations expect to do fine when the tough times come, New York Times, May 9, p. 1. Stiglitz and Weiss, 1981, Credit Rationing in Markets with Imperfect Information, American Economic Review 71, 393-410 22 Figure 1 Quarterly Commercial Paper-Treasury Bill yield spread and percentage change in real GDP 23 Table 1 Summary Statistics This table provides summary statistics for the variables in our sample over the period 1972 to 2008. All accounting numbers are from Compustat. Cash is cash plus marketable securities. Earnings is earnings before extraordinary items plus interest, deferred tax credits, and investment tax credits. Net assets is total assets minus cash holdings. Market-to-Book is market value of equity scaled by total assets. Sales Growth is current net sales minus prior year net sales, scaled by prior year net sales. Leverage is total debt scaled by total market assets. Operating income is earnings before depreciation and amortization expenses to total assets. Changes are measured as the value of the variable is at the end of fiscal year minus its value at the beginning of fiscal year. Stock returns are from CRSP. Excess return is raw return minus the Fama and French (1993) size and book-to-market matched portfolio return. ` CP-Bill spread Below sample median: Above Sample median: Capital Abundance Tight Capital mean median mean median Book Assets 530.36 51.37 444.16 43.2 Sale 497.24 53.6 437.42 47.32 Cash/Net Assets 0.41 0.08 0.33 0.07 Cash/Market equity 0.16 0.08 0.15 0.08 4-quarter sales growth 0.43 0.11 0.39 0.1 8-quarter sales growth 0.84 0.22 0.81 0.19 4-quarter change in ROA 0 0 0 0 8-quarter change in ROA 0 0 0 0 Market to book assets 1.97 1.42 1.85 1.29 Raw stock return 0.06 0.03 0.01 0 Excess return 0 -0.02 -0.01 -0.02 Change in cash/market equity (t-1) 0 0 0 0 0.03 0.01 0.04 0.02 Change in net assets/market equity (t-1) Change in earnings/market equity (t-1) 0 0 0 0 New debt/net assets 0.05 0.01 0.06 0.01 New equity/net assets 0.13 0 0.08 0 Market leverage 0.22 0.16 0.25 0.2 Capital exp./ net assets 0.13 0.04 0.14 0.05 0.1 0 0.09 0 0.29 0.3 0.78 0.69 133,410 133,410 117,430 117,430 R&D exp./ net assets CP-Bill spread Number of obs. 24 Table 2 External financing and capital market conditions: 1975-2008 This table examines the impact of capital market conditions on firm issuance of external financing. The independent variable of the sum of quarterly debt and equity issuance scaled by lagged non-cash assets. Tight Capital is a binary variable that takes a value of one if the CP-Bill yield spread is above the sample median. All the independent variables are winsorized at the 1 and 99 percentiles. The sample period is between 1975 and 2008. We report standard errors in brackets. The standard errors are computed adjusting for heteroskedasticity and within-firm error clustering. ** and * denote statistical significance at the 1% and 5% level, respectively. Tight Capital Dummy Log of Sale (t-1) Oper. income/Net Assets Market-to-book (t-1) Market leverage (t-1) Constant All firms -0.023 [0.010]** -0.054 [0.005]*** -0.012 [0.011] 0.149 [0.020]*** 0.12 [0.061]** 0.042 [0.060] 250840 0.01 Observations R-squared Robust standard errors in brackets * significant at 10%; ** significant at 5%; *** significant at 1% Small -0.085 [0.029]*** -0.098 [0.021]*** -0.983 [0.197]*** 0.164 [0.033]*** 0.124 [0.127] 0.076 [0.117] 83481 0.01 Large 0.013 [0.004]*** -0.01 [0.001]*** -0.835 [0.171]*** 0.062 [0.018]*** 0.017 [0.036] 0.025 [0.033] 83610 0.05 R&D/sales>5% -0.075 [0.043]* -0.129 [0.014]*** -0.01 [0.009] 0.164 [0.026]*** 0.167 [0.186] 0.258 [0.121]** 57511 0.01 R&D/sales<5% -0.009 [0.004]** -0.023 [0.003]*** -0.321 [0.096]*** 0.074 [0.010]*** -0.002 [0.029] 0.065 [0.025]*** 193329 0.01 25 Table 3 The value of cash holdings and capital market conditions This table examines the value of cash changes using Faulkender and Wang‟s (2006) return regressions. All the independent variables are winsorized at the 1 and 99 percentiles. Tight Capital is a binary variable that takes a value of one if the CP-Bill yield spread is above the sample median. The sample period is between 1975 and 2008. We report standard errors in brackets. The standard errors are computed adjusting for heteroskedasticity and within-firm error clustering. ** and * denote statistical significance at the 1% and 5% level, respectively. Panel A: Annual Observations Change in cash Full Sample 0.337 [0.020]*** Tight Capital x Change in cash Tight Capital dummy Change in earnings Change in net assets Change in R&D Change in interest Change in dividends Lagged cash Market leverage Net Financing Change in cash*lagged cash Change in cash*leverage Constant Observations R-squared 0.124 [0.013]*** -0.001 [0.009] -0.136 [0.046]*** -0.307 [0.115]*** 0.223 [0.047]*** 0.297 [0.007]*** -0.263 [0.005]*** 0.036 [0.009]*** -0.057 [0.026]** -0.201 [0.037]*** 0.006 [0.002]*** 250816 0.11 Full Sample 0.335 [0.021]*** 0.007 [0.020] -0.001 [0.001] 0.124 [0.013]*** -0.001 [0.009] -0.136 [0.046]*** -0.305 [0.115]*** 0.223 [0.047]*** 0.297 [0.007]*** -0.263 [0.005]*** 0.036 [0.009]*** -0.056 [0.026]** -0.202 [0.037]*** 0.006 [0.002]*** 250816 0.11 Small 0.415 [0.033]*** 0.026 [0.040] -0.003 [0.004] 0.144 [0.027]*** 0.031 [0.014]** -0.16 [0.068]** -0.288 [0.249] 0.571 [0.108]*** 0.367 [0.011]*** -0.344 [0.012]*** 0.026 [0.016] -0.104 [0.040]*** -0.214 [0.059]*** -0.03 [0.003]*** 83474 0.13 Large 0.259 [0.033]*** -0.002 [0.029] -0.002 [0.002] 0.101 [0.016]*** -0.011 [0.011] -0.098 [0.090] -0.189 [0.164] 0.079 [0.066] 0.242 [0.012]*** -0.228 [0.007]*** 0.031 [0.012]*** -0.025 [0.046] -0.17 [0.061]*** 0.035 [0.002]*** 83602 0.13 R&D >5% 0.49 [0.047]*** 0.139 [0.063]** 0.022 [0.004]*** 0.172 [0.033]*** -0.02 [0.024] -0.096 [0.064] 0.303 [0.692] 0.068 [0.192] 0.441 [0.016]*** -0.448 [0.017]*** 0.056 [0.024]** -0.108 [0.076] -0.331 [0.142]** -0.086 [0.003]*** 57511 0.14 R&D >5% 0.268 [0.023]*** -0.01 [0.021] -0.006 [0.001]*** 0.108 [0.014]*** 0.008 [0.009] 0.026 [0.066] -0.307 [0.114]*** 0.213 [0.047]*** 0.253 [0.007]*** -0.243 [0.005]*** 0.026 [0.010]*** -0.032 [0.027] -0.132 [0.040]*** 0.03 [0.002]*** 193305 0.11 26 Table 4 Capital Market Conditions and the impact of cash on industry adjusted sales growth This table presents results of regressions examining the effect of industry-adjusted cash holdings on industryadjusted sales growth The dependent variable is the annual industry-adjusted sales growth at time t [(Salest Salest−1)/Salest−1]. Cash is the ratio of cash and marketable securities divided by non-cash assets. We report standard errors in brackets. The standard errors are computed adjusting for heteroskedasticity and within-firm error clustering. ** and * denote statistical significance at the 1% and 5% level, respectively. Panel A: 4-quarter growth Industry-adjusted cash (t-1) Recession*lagged adjusted cash Recession dummy Log of Sale (t-1) Industry-adjusted ROA (t-1) Industry-adjusted capital exp.(t-1) Industry-adjusted leverage(t-1) Constant Observations R-squared All firms 0.044 [0.019]** -0.067 [0.039]* -0.046 [0.015]*** -0.094 [0.008]*** -1.954 [0.213]*** 0.091 [0.021]*** -0.403 [0.062]*** 0.561 [0.043]*** 203577 0.01 Small -0.006 [0.022] -0.122 [0.035]*** -0.098 [0.047]** -0.423 [0.052]*** -1.449 [0.209]*** 0.098 [0.054]* -0.506 [0.161]*** 1.071 [0.119]*** 61537 0.01 Large 0.113 [0.029]*** -0.006 [0.052] -0.01 [0.007] -0.047 [0.005]*** -0.643 [0.139]*** 0.084 [0.024]*** -0.174 [0.019]*** 0.352 [0.033]*** 73960 0.02 R&D >5% 0.026 [0.022] -0.077 [0.039]** -0.008 [0.065] -0.133 [0.022]*** -1.785 [0.232]*** 0.121 [0.047]** -1.027 [0.480]** 0.658 [0.104]*** 44211 0.01 R&D <5% 0.121 [0.067]* 0.027 [0.089] -0.062 [0.011]*** -0.079 [0.006]*** -2.825 [0.626]*** 0.083 [0.022]*** -0.349 [0.050]*** 0.505 [0.035]*** 159366 0.01 All firms 0.067 [0.037]* 0.017 [0.106] -0.093 [0.032]*** -0.195 [0.015]*** -3.919 [0.463]*** 0.18 [0.049]*** -0.897 [0.110]*** 1.157 [0.080]*** 174485 0.01 Small -0.045 [0.035] -0.165 [0.099]* -0.2 [0.106]* -0.917 [0.104]*** -2.665 [0.416]*** 0.134 [0.105] -1.253 [0.278]*** 2.297 [0.225]*** 50523 0.02 Large 0.188 [0.048]*** -0.084 [0.062] -0.01 [0.013] -0.104 [0.010]*** -0.667 [0.340]** 0.19 [0.027]*** -0.31 [0.038]*** 0.748 [0.064]*** 65889 0.02 R&D >5% 0.017 [0.047] -0.019 [0.109] 0.016 [0.165] -0.299 [0.047]*** -3.447 [0.495]*** 0.296 [0.127]** -2.789 [0.815]*** 1.492 [0.203]*** 36618 0.02 R&D<5% 0.214 [0.115]* 0.133 [0.175] -0.122 [0.020]*** -0.154 [0.010]*** -5.603 [1.354]*** 0.154 [0.049]*** -0.726 [0.104]*** 0.988 [0.064]*** 137867 0.01 Panel B: 8-quarter growth Industry-adjusted cash (t-1) Recession*lagged adjusted cash Recession dummy Log of Sale (t-1) Industry-adjusted ROA (t-1) Industry-adjusted capital exp.(t-1) Industry-adjusted leverage(t-1) Constant Observations R-squared Robust standard errors in brackets * significant at 10%; ** significant at 5%; *** significant at 1% 27 Table 5 Capital Market Conditions and the impact of cash on industry adjusted operating performance This table presents results of regressions examining the effect of industry-adjusted cash holdings on industryadjusted change in operating performance. The dependent variable is the change in industry-adjusted operating income over non-cash assets ROA t −ROAt−1. Cash is the ratio of cash and marketable securities divided by non-cash assets. We report standard errors in brackets. The standard errors are computed adjusting for heteroskedasticity and within-firm error clustering. ** and * denote statistical significance at the 1% and 5% level, respectively. Panel A: 4-quarter growth Industry-adjusted cash (t-1) Tight Capital*lagged adjusted cash Tight Capital dummy Log of Sale (t-1) Industry-adjusted ROA (t-1) Industry-adjusted capital exp.(t-1) Industry-adjusted leverage(t-1) Constant Observations R-squared All firms -0.012 [0.002]*** -0.004 [0.006] 0.003 [0.001]*** 0.005 [0.000]*** -0.515 [0.026]*** 0.001 [0.001] -0.012 [0.002]*** -0.023 [0.002]*** 203577 0.2 Small -0.013 [0.002]*** -0.011 [0.010] 0.012 [0.003]*** 0.014 [0.002]*** -0.547 [0.032]*** 0.001 [0.001] -0.01 [0.004]*** -0.044 [0.004]*** 61537 0.2 Large 0 [0.003] 0.005 [0.004] -0.001 [0.000]** 0.001 [0.000]*** -0.384 [0.012]*** 0 [0.000] -0.011 [0.001]*** -0.002 [0.001]*** 73960 0.17 R&D>5% -0.013 [0.002]*** -0.006 [0.008] 0.01 [0.003]*** 0.014 [0.001]*** -0.56 [0.033]*** 0.005 [0.004] 0 [0.006] -0.065 [0.005]*** 44211 0.25 R&D<5% -0.005 [0.004] 0.006 [0.004] 0 [0.000] 0.002 [0.000]*** -0.51 [0.025]*** 0 [0.000] -0.012 [0.002]*** -0.007 [0.001]*** 159366 0.11 All firms -0.01 [0.002]*** 0.007 [0.005] 0.002 [0.001]*** 0.005 [0.000]*** -0.58 [0.020]*** 0 [0.000] -0.007 [0.002]*** -0.022 [0.001]*** 174485 -0.01 Small -0.009 [0.002]*** 0.009 [0.007] 0.009 [0.002]*** 0.014 [0.001]*** -0.602 [0.025]*** 0 [0.001] -0.005 [0.003] -0.042 [0.003]*** 50523 -0.009 Large 0.001 [0.003] 0.002 [0.004] -0.001 [0.000]*** 0.001 [0.000]*** -0.487 [0.013]*** 0 [0.000] -0.01 [0.001]*** 0 [0.001] 65889 0.001 R&D>5% -0.01 [0.002]*** 0.008 [0.006] 0.007 [0.003]*** 0.014 [0.001]*** -0.62 [0.026]*** -0.002 [0.002] 0.008 [0.006] -0.061 [0.004]*** 36618 -0.01 R&D<5% -0.003 [0.002] 0.008 [0.003]*** 0 [0.000] 0.002 [0.000]*** -0.605 [0.020]*** 0 [0.000] -0.009 [0.001]*** -0.006 [0.001]*** 137867 -0.003 Panel B: 8-quarter growth Industry-adjusted cash (t-1) Tight Capital *lagged adjusted cash Tight Capital dummy Log of Sale (t-1) Industry-adjusted ROA (t-1) Industry-adjusted capital exp.(t-1) Industry-adjusted leverage(t-1) Constant Observations R-squared Robust standard errors in brackets * significant at 10%; ** significant at 5%; *** significant at 1% 28 Table 6 Capital Market Conditions and the impact of cash on industry adjusted capital expenditures This table presents results of regressions examining the effect of industry-adjusted cash holdings on industryadjusted capital expenditures. The dependent variable is the industry-adjusted capital expenditures over non-cash assets. Cash is the ratio of cash and marketable securities divided by non-cash assets. We report standard errors in brackets. The standard errors are computed adjusting for heteroskedasticity and within-firm error clustering. ** and * denote statistical significance at the 1% and 5% level, respectively. Panel A: 4-quarter growth Industry-adjusted cash (t-1) Tight Capital*lagged adjusted cash Tight Capital dummy Log of Sale (t-1) Industry-adjusted ROA (t-1) Market-to-book (t-1) Industry-adjusted leverage(t-1) Constant Observations R-squared All firms 0.076 [0.012]*** -0.022 [0.012]* -0.001 [0.004] -0.018 [0.001]*** 0.16 [0.063]** 0.024 [0.002]*** -0.213 [0.011]*** 0.112 [0.007]*** 228666 0.04 Small 0.062 [0.014]*** -0.024 [0.014]* -0.012 [0.009] -0.046 [0.006]*** 0.098 [0.076] 0.022 [0.002]*** -0.29 [0.023]*** 0.158 [0.016]*** 73723 0.03 Large 0.135 [0.030]*** -0.021 [0.032] 0.008 [0.003]*** -0.015 [0.001]*** 1.027 [0.145]*** 0.006 [0.002]*** -0.11 [0.008]*** 0.115 [0.008]*** 78334 0.05 R&D>5% 0.068 [0.014]*** -0.017 [0.013] 0.042 [0.011]*** -0.006 [0.003]** 0.046 [0.082] 0.022 [0.002]*** -0.26 [0.030]*** 0.018 [0.014] 51328 0.06 R&D<5% 0.137 [0.035]*** -0.079 [0.036]** -0.01 [0.004]** -0.024 [0.001]*** 0.178 [0.079]** 0.043 [0.004]*** -0.169 [0.014]*** 0.116 [0.010]*** 177338 0.03 All firms 0.149 [0.016]*** 0.03 [0.026] -0.002 [0.009] -0.05 [0.002]*** 0.468 [0.085]*** 0.046 [0.003]*** -0.441 [0.021]*** 0.288 [0.015]*** 194808 0.04 Small 0.121 [0.020]*** 0.017 [0.035] -0.006 [0.027] -0.132 [0.013]*** 0.442 [0.107]*** 0.041 [0.005]*** -0.619 [0.053]*** 0.428 [0.035]*** 60294 0.02 Large 0.151 [0.026]*** 0.098 [0.034]*** 0.024 [0.005]*** -0.034 [0.002]*** 2.268 [0.352]*** 0.018 [0.004]*** -0.226 [0.015]*** 0.227 [0.014]*** 69513 0.06 R&D>5% 0.122 [0.017]*** 0.073 [0.029]** 0.056 [0.021]*** -0.029 [0.004]*** 0.235 [0.100]** 0.04 [0.003]*** -0.64 [0.082]*** 0.132 [0.022]*** 42195 0.07 R&D<5% 0.29 [0.031]*** -0.219 [0.047]*** -0.001 [0.010] -0.059 [0.003]*** 0.323 [0.188]* 0.091 [0.011]*** -0.341 [0.024]*** 0.273 [0.022]*** 152613 0.03 Panel B: 8-quarter growth Industry-adjusted cash (t-1) Tight Capital *lagged adjusted cash Tight Capital dummy Log of Sale (t-1) Industry-adjusted ROA (t-1) Market-to-book (t-1) Industry-adjusted leverage(t-1) Constant Observations R-squared Robust standard errors in brackets * significant at 10%; ** significant at 5%; *** significant at 1% 29 Table 7 Capital Market Conditions and the impact of cash on industry adjusted inventory expenditure This table presents results of regressions examining the effect of industry-adjusted cash holdings on industryadjusted invesntory expenditures. The dependent variable is the industry-adjusted inventory expenditures over noncash assets. Cash is the ratio of cash and marketable securities divided by non-cash assets. We report standard errors in brackets. The standard errors are computed adjusting for heteroskedasticity and within-firm error clustering. ** and * denote statistical significance at the 1% and 5% level, respectively. Panel A: 4-quarter growth Industry-adjusted cash (t-1) Tight Capital *lagged adjusted cash Tight Capital dummy Log of Sale (t-1) Industry-adjusted ROA (t-1) Market-to-book (t-1) Industry-adjusted leverage(t-1) Constant Observations R-squared All firms 0.014 [0.004]*** 0.001 [0.005] -0.001 [0.002] -0.006 [0.000]*** 0.021 [0.025] 0.006 [0.001]*** -0.066 [0.004]*** 0.041 [0.002]*** 224122 0.01 Small 0.013 [0.005]** 0.001 [0.007] 0.003 [0.005] -0.009 [0.002]*** 0.015 [0.031] 0.007 [0.001]*** -0.078 [0.008]*** 0.049 [0.005]*** 72742 0.01 Large 0.016 [0.003]*** 0.002 [0.006] -0.003 [0.001]** -0.002 [0.000]*** 0.296 [0.046]*** 0.001 [0.001] -0.036 [0.006]*** 0.023 [0.004]*** 76295 0.01 R&D>5% 0.014 [0.005]*** 0 [0.006] 0.011 [0.005]** -0.007 [0.001]*** 0.025 [0.027] 0.005 [0.001]*** -0.157 [0.012]*** 0.036 [0.004]*** 50850 0.03 R&D<5% 0.018 [0.008]** -0.002 [0.011] -0.003 [0.002] -0.005 [0.000]*** -0.041 [0.092] 0.011 [0.002]*** -0.051 [0.005]*** 0.035 [0.004]*** 173272 0.01 All firms 0.026 [0.006]*** 0.036 [0.022]* 0.01 [0.006]* -0.012 [0.001]*** 0.01 [0.075] 0.011 [0.001]*** -0.106 [0.009]*** 0.085 [0.004]*** 190564 0.01 Small 0.02 [0.007]*** 0.031 [0.032] 0.029 [0.017]* -0.021 [0.004]*** 0.009 [0.097] 0.015 [0.003]*** -0.116 [0.019]*** 0.1 [0.009]*** 59384 0.01 Large 0.026 [0.006]*** 0.02 [0.019] 0 [0.002] -0.005 [0.001]*** 0.456 [0.071]*** 0.002 [0.001] -0.061 [0.005]*** 0.046 [0.005]*** 67566 0.01 R&D>5% 0.028 [0.007]*** 0.044 [0.027] 0.014 [0.011] -0.017 [0.002]*** 0.078 [0.079] 0.008 [0.002]*** -0.291 [0.021]*** 0.087 [0.008]*** 41761 0.04 R&D <5% 0.029 [0.012]** 0 [0.019] 0.012 [0.007]* -0.011 [0.001]*** -0.303 [0.286] 0.026 [0.007]*** -0.083 [0.010]*** 0.063 [0.012]*** 148803 0.01 Panel B: 8-quarter growth Industry-adjusted cash (t-1) Tight Capital *lagged adjusted cash Tight Capital dummy Log of Sale (t-1) Industry-adjusted ROA (t-1) Market-to-book (t-1) Industry-adjusted leverage(t-1) Constant Observations R-squared Robust standard errors in brackets * significant at 10%; ** significant at 5%; *** significant at 1% 30