Projects are Largely External and Mostly Debt Financed: A New Approach to Testing Capital Structure. Colin Mayer and Oren Sussman

Transcription

1 Projects are Largely External and Mostly Debt Financed: A New Approach to Testing Capital Structure Colin Mayer and Oren Sussman Saïd Business School, University of Oxford 23 June 2002 First Draft We are grateful to Zhangkai Huang for research assistance on the paper and to the Peter Moores Foundation for financial support. We thank participants at conferences in Copenhagen, Oslo and Oxford for comments. We are grateful to Kristian Rydqvist for helpful suggestions.

2 Abstract This paper reports a new test of capital structure theories using a filtering technique to identify large investment projects. Contrary to the results of aggregate studies, firms respond to investment spikes by raising large amounts of externa l finance. Large firms raise debt finance and small firms issue new equity. These results run counter to predictions of the pecking order theory. New equity does not come higher up the pecking order either in relation to specific investments or over the life cycle of firms. We also reject a static version of the trade-off theory by which the financing of new investment is determined by invariant characteristics of firms. However, we find support for a dynamic version of the theory in which firms adjust to target levels of leverage both during and after investment projects. We provide evidence that the reason why a dynamic rather than a static version of the theory prevails is that firms face constraints in their choice of finance. Key words: Capital structure, project financing, pecking order theory, trade -off theory JEL classification: G32

3 1 Introduction There are two widely cited theories of capital structure that are currently the subject of active debate in the finance literature: the pecking order theory and the tradeoff theory. According to the pecking order theory, firms use internal finance in preference to external finance, and, because of an adverse selection problem associated with new equity issues, use debt finance in preference to issues of new equity. According to the trade-off theory, firms have an optimal capital structure that is determined by, for example, setting the tax advantages of debt against the costs of financial distress. Firms choose their finance at any particular point in time to move their capital structure closer towards its optimum. Patterns of corporate financing of corporations are consistent with a pecking order theory. Retained earnings are by far and away the dominant source of finance in all countries. Debt finance is much the most significant source of external finance, primarily from banks but also, in the case of North America, from bond markets. In comparison, new equity finance accounts for a very small proportion of total corporate sector financin g (see, for example, Corbett and Jenkinson (1997), Mayer (1988) and Rajan and Zingales (1995)). These aggregate studies provide broad descriptions of overall patterns of finance across the corporate sector as a whole over extended periods of time. They do not, however, identify how individual firms respond to particular financing requirements nor do they describe the dynamics of the financing process. They do not therefore provide a direct test of the capital structure theories. The existing literature has instead used panels of firms to relate capital structure to individual firms financing needs and characteristics. Shyam-Sunder and Myers (1999) find evidence to support the prediction of the pecking order theory that a regression of debt financing on firms funds flow deficit (real investment and dividend obligations less internal funds) should have a slope coefficient close to one. However, Chirinko and Singha (2000) note that this is neither a necessary nor a sufficient condition for the pecking order theory to be valid: the slope coefficient could fall well short of unity when the pecking order theory holds and be close to unity when it does not. Frank and Goyal (2002) record that Shyam-Sunder and Myers results are anyway not robust to alternative sample selection criterion and that the pecking order theory appears to perform particularly poorly amongst small firms, for which adverse selection problems of raising external equity might have been expected to be most relevant. In relation to trade-off theories, Bradley et al (1984) summarize some of the early evidence in support of an optimal capital structure. However, using a factor-analytic 1

4 approach, Titman and Wessels (1988) find little evidence that such factors as tax shields and the volatility of earnings predicted by the trade-off theory actually influence firms capital structure. Baker and Wurgler (2002) record that, contrary to the trade-off theory, fluctuations in firm s market valuations have long rather than transient impacts on the ir capital structures. Welch (2002) finds that leverage ratios respond mechanistically to firms own stock returns. They show a great deal of inertia and little tendency to return to target levels. These studies reveal two deficiencies of existing appr oaches. As Chirinko and Singha (2000) have noted, the pecking order theory is a non-linear model. As funding requirements increase, firms initially employ retained earnings exclusively and, when these are exhausted, debt finance alone, and, once debt capacity is exhausted, turn to external equity finance. The problem that Chirinko and Singha have revealed with Shyam-Sunder and Myers study and indeed with most panel analyses is that they are not well suited to capturing this non-linearity in financing behaviour. The second deficiency of existing studies is they do not provide a precise description of the dynamics of financing. As we describe below, the pecking order theory has clear predictions about dynamics. However, as Baker and Wurgler, and Welch reveal, the dynamics of capital structure may be dominated by other considerations, such as current and past share price movements. These may swamp evidence of partial adjustment back to long-run equilibria in normal circumstances. However, the dynamics of financing may be more pronounced in the extreme circumstances described in this paper when large misalignments can occur. The fundamental problem with existing approaches is that they do not focus on the issue in which one is primarily interested, name ly how is investment financed. Aggregate studies address this issue but not in an informative way. Most financing activity is associated with routine replacement rather than expansion of the capital stock. The fact that internal funds are the normal source of finance does not mean that they are the most significant in the sense that they meet firms requirements for financing large investment projects. As Harris and Raviv (1991) note, existing empirical studies have identified a large number of potential determinants of capital structure. The empirical work so far has not, however, sorted out which of these are important in various contexts..the empirical work is largely consistent with the theory, although there are a few instances where the evidence seems to contradict certain models. These inconsistencies cannot, however, be regarded as conclusive, because the empirical studies were not designed specifically to test the models and were, therefore not careful about satisfying the ceteris 2

5 paribus conditions.with regard to further empirical work, it seems essential that empirical studies concentrate on testing particular models or classes of models in an attempt to discover the most important determinants of capital structure in given environments. (Harris and Raviv (1991), p. 3). We take up Harris and Raviv s challenge by following a different procedure from the existing literature. Firstly, we use disaggregate rather than aggregate data. In fact, we operate at the level of projects or collections of projects. Secondly, we look at what happens out of rather than in steady state. We are interested in the financing of unusually large projects - shocks rather than routine investment. This approach is informative about new investment activities in a way in which neither the aggregate nor panel data studies can be. Specifically, we construct a filter for identifying firms that display investment spikes distinct sharp one-off increases in investment. We look at the financing of firms around and during the spikes. We then examine whether there is a relation between financing patterns before, after and during the spike and the characteristics of firms. In effect, we undertake an event study of the financing response of firms to large investment spikes. This methodology allows a variety of questions that are central to corporate finance to be addressed in an informative way. Do firms finance spikes in investment from retentions, debt or new equity? Does debt finance precede new equity? Do firms accumulate cash balances prior to large investment expenditures? Do large firms use more or less bank finance than small firms? Are small firms excluded from new equity sources? The investment spike approach also allows the pecking order and trade-off theories to be tested directly. The pecking order theory has clear implications for the ordering of finance, the relation between types of finance and scale of investment projects, and the forms of finance that different types of firms employ. The trade-off theory predicts that large projects should be funded in accordance with the firm s optimal capital structure and, to the extent that actual capital structure is perturbed from its equilibrium level, there should be evidence of partial adjustment back to equilibrium. By stress testing, large projects are more informative about both the determinants and dynamics of capital structure. 3

6 Some striking results emerge. First, while firms may raise little external financing in steady state, they raise large amounts in response to investment spikes. Large investment expenditures are not financed out of accumulated reserves. Second, debt is the dominant source of external finance. Third, the use of debt finance is predominantly a large company result. Surprisingly, small listed companies rely heavily on new equity sources. Fourth, the use of debt finance comes after new equity issues. Companies do not appear to exhaust debt finance before they go to stock markets; in fact, they use debt after accessing equity markets. Fifth, the fact that large companies finance large projects out of debt and small companies out of new equity means that the life cycle development of firm financing is from stock markets to debt not vice versa. Sixth, firms respond to the financing needs created by investment projects differently from those associated with declines in earnings. These observations are quite contrary to the pecking order theory. Firms do not work up the pecking order from debt to new equity. While debt is an important source of external finance, it does not precede new equity finance either sequentially in the financing of particular projects or over the life cycle of firms. There is a large amount of equity financing, particularly by small firms, which are predicted by the pecking order theory to be particularly discouraged by information asymmetries. We also reject what we term a static version of the trade-off theory, according to which, firms choice of finance should be influenced by invariant characteristics of the firm. We do not find that financing patterns are as stable as this theory predicts. However, we do report support for a dynamic version of the trade-off theory in which firms adjust to target capital structure both during and after spikes in investment. We suggest that the reason why a dynamic rather than a static version of the trade-off theory applies is that there are financing constraints. We provide evidence of such constraints in raising finance in response to earnings losses around the time of investment peaks. But investment projects appear to offer opportunities for adjusting towards as well as being displaced from optimal structures. These opportunities may be created by the collateral and future cash flows that investment projects produce. 4

7 In Section 2 we set out the hypotheses on firm financing that we test in the remainder of the paper. In Section 3, we describe the data and the filtering technique that we have employed to identify our sample. In Section 4, we use our sample to describe the way in which large projects are funded and the results of the tests of our hypotheses. Section 5 concludes the article and draws inferences for the pecking order and trade-off theories of finance. 2 Hypotheses 2.1 The Pecking Order Theory According to the pecking order theory (Myers (1984) and Myers and Majluf (1984)), retained earnings are preferred to external finance because of asymmetries of information between investors and firms make external finance expensive. Existing empirical studies appear to support the dominance of retained earnings. The figures in Table 1 are typical of those reported in aggregate studies of corporate financing (e.g Corbett and Jenkinson (1997) and Rajan and Zingales (1995)). They refer to average financing proportions for the sample of firms in this study. 77% of gross investment is financed from internal sources and only 23% externally. INSERT TABLE 1 H1: Project finance is largely internal. However, as Table 2 reveals, aggregate patterns of financing are not informative about the financing of individual firms or projects or about the dynamics of the financing process. Consider two firms A and B that fund a project costing 150 over three periods. Firm A is internally financed and accumulates liquid assets to pay for the capital expenditure. Firm B is largely (two-thirds) externally financed. Over the three periods together (columns 5 and 9 of Panel A of Table 2), the external flow of finance to both firms is zero. Similarly in time period 1 when the only investment activity that is being funded is firm B, two-thirds of which comes from external sources, the aggregate data in Panel B suggest that two-thirds of total investment in firms A and B combined is coming from internal sources. As a description of the financing of A and B together and/ or of the period 1 to 3 combined, the aggregate figures are an accurate description. They therefore correctly record financing at the level of economies over extended period. However, disaggregated figures are required to identify the financing of individual firms or projects and the dynamics of financing around particular investments. INSERT TABLE 2 5

8 According to the pecking order theory, problems of asymmetry of information are particularly acute in new equity issues and therefore, to the extent that firms are externally financed, this predominantly comes from debt. H2: External finance is mainly debt. Since asymmetries of information are likely to be especially pronounced for small firms, they will be particularly afflicted by the adverse selection problems of raising new equity. H3: Small firms are particularly reliant on internal sources and debt. According to the pecking order theory, firms use internal sources of finance ahead of external sources and debt ahead of equity. Equity finance will therefore be associated with particularly large investment requirements. H4: New equity issues are associated with particularly large investment projects. Once a company has issued equity it must have exhausted its debt capacity and, having revealed its low valuation, it will go on issuing equity so long as it requires external finance. Thus if: (1) DEBT 0 = (? +?x t )(I 0 -?OPR 0 ), where DEBT? is the flow of borrowing, I? is investment and OPR? are earnings from operations, and subscripts refer to project time?=-2,...2, with t = 0 being the time at which the investment takes place, then, according to the pecking order theory: H5:? =0,?=1, x?? 0 if EQUITY?? 0,??? 1or??? 1 otherwise The above applies to buy-backs, which are negative values of EQUITY. 2 The pecking order theory does not distinguish between financing requirements created by new investment and by operating losses. Thus it assumes that: H6:? = 1. 6

9 2.2 The Trade -Off Theory According to the trade-off theory, x t is a vector of firm characteristics that is correlated with the underlying determinants of optimal capital structure, for example, the volatility of earnings proxying for risks of default, R&D reflecting asset intangibility, and industry affiliation rela ting to opportunities for over-investment, mature industries having less such opportunities. 1 However, usually only crude measures of the underlying determinants of optimal leverage are available. In the static version of the trade-off theory, past levels of leverage at, for example,?=-2, before the investment shock, are a sufficient statistic of static leverage. Hence, H7: According to a static version of the trade-off theory, in equation 1, x -2 is past leverage,? =?=0, and?=1. If firms do not have complete discretion over the way in which the investment shock is financed then they may be perturbed away from their optimal level of leverage. If so a dynamic version of the trade-off theory predicts partial adjustment back to static levels in the subsequent periods t = +1 and +2. Thus H8: If x? is leverage at? then x 2 x 0 =? ( x 0 x -2 ) where -1 <? < 0. Alternatively, if firms are not at their optimal level of leverage at t = -2 then the financing of the investment spike may itself be an opportunity to adjust towards the optimum. In particular, the collateral associated with the project may allow debt, which is constrained at other times, to be raised or the cash flows generated by the project may permit new equity to be raised. H9: ß is negative in DEBT 0 = (? +?x -2 )(I 0 -?OPR 0 ) where x -2 is leverage at t = -2. We might expect an inability of firms to establish optimal levels of leverage before or during investment spikes to be reflected in constraints on financing capability. For example, if H6 does not hold then we might anticipate that H8 and H9 are associated with differential funding of profits and losses and, in particular, less response of debt to increased losses than reduced profits. H10: Financial constraints are reflected in a low? where OPR 0 <0. 7

10 3. Data 3.1 The Sample The data used in this paper are the flow-of-funds accounts of non-financial North American companies reported in COMPUSTAT, for the years All the companies are publicly traded, although some are traded over-the-counter rather than listed on one of the exchanges. The data have been deflated to constant price values using the consumer price index. The basic data set comprises more than eleven thousand companies. We have gone to considerable lengths to clean and check the data before using them. We performed consistency checks on the data, such as sources of funds equalling uses, and deleted data records (company-years) that failed these tests. We deleted firms that failed to report key variables, such as after-tax income, depreciation, equity finance or debt finance. Approximately four hundred companies were deleted as a consequence, leaving 10,667 companies. There is a high turnover rate of firms caused by births and deaths. 6,293 of the 10,667 companies were still alive in 1998 and 4,253 of them were in existence by Only 5,568 had five consecutive records and the rest were discarded as the investment spike filter described below was programmed to detect an investment spike relative to the two prior and two subsequent years. The next step was to aggregate the data into the following categories: (2) I ti =OPR ti + EQUITY ti + LTDEBT ti + OTHER ti where, I is fixed investment, OPR is cash flow from operations (after tax), EQUITY is equity finance (net), LTDEBT is long-term debt finance (net), OTHER is the sum of all other variables, t is a time index and i is a company index. The data appendix provides details of the items included in each aggregate (with their COMPUSTAT labels). A positive (negative) sign on the right hand side means 1 See Harris and Raviv (1991). 8

11 source of funds ( use of funds ); for example, LTDEBT is positive (negative) when the firm borrows (repays debt). Missing variables are an endemic problem in COMPUSTAT. Since all records add-up, missing at this point does not mean unaccounted for but rather aggregated in to some other item in an unspecified manner. COMPUSTAT does not automatically allocate a missing item x to item y as against z. Hence, we have followed the only procedure possible of replacing missing values with zeros within variables I, OPR, EQUITY and LTDEBT. For example, EQUITY is equal to SSTK+ PRSTKC (sale of equity and purchase of equity, see Appendix A). If both SSTK and PRSTKC are missing, the whole record, i.e. EQUITY, will have been deleted in the previous stage. If, however, only one of SSTK and PRSTKC is missing, then the missing item has been replaced with zero, on the assumption that EQUITY is already reported on a net basis. We have checked this procedure against a sample of original company accounts and believe the resulting measurement error to be small. All other items are aggregated in OTHER. OTHER therefore includes changes in liquid assets as well as measurement errors and genuinely other items. Given its economic significance, we went to considerable lengths to try to identify changes in liquid assets separately but we were unable to do this. However, we believe that quality of the remaining data employed in this study to be high. 3.2 The Filter The filter is designed to identify companies with spikes of investment. We define a spike as a five-year string of investment that approximates the following pattern: (3) (1, 1, (2 or more), 1, 1), 1 represents the off-spike base level of investment and 2 as the minimum ratio of spike to base level investment. The filter scans the raw data record by record searching for an investment pattern approximating (3). First, it computes a base-level of investment, (4) b i, t? I? I? I? i, t? 2 i, t? 1 i, t? 1 4 I i, t? 2 Given the base-level of investment, p i,t+j, is defined as the base level off spike and double the base level on spike: 9

12 ? 2bi, t if j? 0 (5) pi, t? j??, j?? 2,...,? 2? bi, t otherwise The next step is to calculate the sum of squared errors from the pattern p i,t+j (normalized by the base-level of investment): (6) ER i, t? 2 2? u j?? 2 i, t? j b i, t 5 However, since we define the spike as double or more the base level, we treat positive spike deviations as a perfect fit, i.e. (7) u? 0, I? p?? min i, t? j i, t? j if j? 0? j??, j?? 2,,? 2? Ii, t? j? pi, t? j otherwise i, t? The result is a mapping of a five-year time series of each company s data (excluding its first and last two years) to the pattern defined in equation (5), where quality of fit is measured in terms of the (relative) sum of squared errors from the pattern. The level of fit required to qualify for spike investment is judgemental. This is assisted by plotting typical strings in Figure 1, which shows a sample of strings by decreasing order of measure of fit (best fit is at the upper-left corner). Each string contains five yearly observations along the time index? = -2,...,2 with the spike at?=0. On the basis of this we have chosen 0.25 as a reasonable cut-off level for ER. INSERT FIGURE 1 We delete 17 records with extreme values and obtain our working sample of 535 companies with 5 complete records and a spike in the middle. The criterion for excluding extreme values and its effect are described in the data appendix. The appendix also provides an industry breakdown of the sample and leverage levels by industry. We have examined the possible sample selection involved in the filtering process. Table 3 reports sample statistics of both the raw panel and the filtered sample. The filtered sample is biased towards mature companies, in terms of size and low levels of leverage and towards NYSE and away from NASDAQ firms. However, a probit regression of whether a firm is filtered (dependent variable equals 1) or not reveals that neither these differences nor industry composition are statistically significant (Panel A of Table 3). 10

13 Panel B of Table 3 shows that the main determinant of whether a firm is included in the filtered sample is how many consecutive observations are available for that firm. The conditional probability of being included in the filtered sample rises from 2.4% for a firm with 5 consecutive years observations to 16.6% for a firm with 11 years of consecutive observations. INSERT TABLE 3 4. Results 4.1 Flow of Funds Table 4 provides statistics on the characteristics of the sample of filtered firms categorized by three equal size groups measured by base level investment. The large size group have average total assets of $3 billion and the small of $25 million. Average annual growth in assets is between 4.7% and 5.9% over the five year strings and between 4.8% and 6.4% over the two years to peak investment. There is little difference in growth rates between the three size groups. Market to book ratios average just under 2 and leverage ratios (debt over total assets) around 25%. Again these show little systematic difference across the size groups. Where the main differences are observed are in profit (after tax income before extraordinary items normalized by base level investment) and cash flow volatility. Whereas large and medium sized firms are on average making profits, small firms are on average making losses and the volatility of small firms cash flows are appreciably greater than those of medium and large sized firms. INSERT TABLE 4 Table 5 describes the flow of funds around the investment spike of the filtered firms, partitioned by the three size groups of base level investment. The average size of the spike is 2.71 times base investment for large firms, 2.85 times base investment for medium-sized firms and 3.81 times base investment for small firms. For large firms, there is little accumulation of internal resources prior to or during the investment spike as evidenced by internal earnings being close to base investment. There is virtually no new equity on average issued prior to or during the investment spike and there is a large increase in debt finance during the investment spike. Panel B of Table 5 computes implied project leverage as debt (for example, 1.03 in the case of large firms) divided by the increase in investment above its base level ( in the case of large firms). 60.2% of the investment spike is funded from debt by large firms on this basis. INSERT TABLE 5 11

14 The picture for medium sized firms is similar but they issue more new equity prior to the investment spike than large firms. 63.2% of the investment spike is financed from debt and 10.8% from new equity in the case of medium sized firms. Small companies issue the largest amounts of new equity prior to and during the investment spike and only modest amounts of debt. 91.1% of the investment spike is funded from new equity and only 22.8% from debt. Table 6 records flow of funds for firms that engage in substantial equity operations new equity issues and buy-backs of shares. A substantial operation is defined as one that is in excess of base level of investment, i.e. outside the range 1 to 1 relative to base investment, b. There were 297 substantial new equity issues in total over the period 2 to +2, 46 by large firms, 64 by medium sized firms and 187 by small firms. There were 125 substantial buy-backs, 51 by large firms, 37 by medium sized firms and 37 by small firms. Therefore around the time of large investment projects, large firms are actually engaging in share buy-backs and small firms in substantial new equity issue. INSERT TABLE 6 The new equity issue activity of small firms is associated with exceptionally large investment projects, 5.58 times base investment on average and large operating losses ( times base investment). The buy-backs of large firms are associated with large operating profits (2.44 times base investment) of a magnitude about equal to their spike investment (2.54 times base investment). In summary, large projects are predominantly externally financed. External finance primarily takes the form of debt for large and medium sized firms and new equity for small firms. There is a substantial amount of new equity activity but it is primarily associated with small not large companies. It is particularly closely associated with small firms that encounter operating losses as well as large investment requirements. Large firms in contrast are engaging in as much repurchase of equity as new issues around a large investment, in particular where there are operating surpluses. In relation to the first three hypotheses, we strongly reject H1 and H3. Project finance is not primarily internally financed. There is a substantial amount of external finance in all size categories in our sample. Consistent with H2, we find that large firms do mainly use debt finance but this is not true for small companies. Contrary to H3, far from small firms being dependent on internal and debt finance, they are particularly reliant on external equity. 12

15 4.2 The Pecking Order Theory According to the pecking order theory, modest external financing requirements should be met from debt and only once debt capacity has been exhausted will firms issue new equity. A concave relation between debt financing and investment is therefore expected as debt capacity becomes exhausted. Debt 45 Investment Table 7 tests this by examining the relation between debt finance and investment. Column 2 reports the predicted positive relation with approximately 60% of a marginal increase in the investment spike being financed from debt for all three size categories. In column 3, a spline regression is reported in which the variable investment is set equal to zero for investment expenditures in excess of 4 (i.e. 4 times base investment) and a second variable investment>4 is set equal to investment minus 4 for investment expenditures in excess of 4 and zero otherwise. Concavity would therefore lead to the prediction that the coefficient on investment>4 should be less than that on investment. There is no clear evidence of this for any size category and strong rejection for small firms. Column 4 tests for concavity by including a term in investment squared as well as investment. The positive coefficient in large and small firms suggests a convex rather than concave relation. There is therefore no support for H4 that debt finance gives way to new equity issues in large investment projects as debt capacity becomes exhausted. INSERT TABLE 7 The pecking order theory also has implications for the sequencing of financing. Firms will only be willing to reveal their type in a new equity issue once they have exhausted their debt capacity. Having done this they will continue to issue new equity if required rather than raise debt finance. Conversely, having revealed their type in a buyback, firms will issue debt rather than new equity. Column 2 of Table 8 reports regressions of debt finance on cash flow from operations, investment and interactive terms of two dummies with investment for the three size categories of firms. The first dummy (D-EQ) is equal to one if there are substantial new equity issues (greater than the 13

16 base level of investment) in either periods t = -2 or 1 and the second (D-BK) is equal to one if there are substantial buy-backs in either t= 2 or 1. Contrary to the predictions of the pecking order theory, the interactive term in (D-EQ) is significantly positive for small firms, the largest size category of equity issuers. The coefficient on the interactive term is (D-BK) is insignificant rather than positive in all size categories including large firms, which are the largest size category of repurchasers, and in two out of three of the size categories, the coefficient on the interactive term with (D-EQ) is larger than that with (D- BK). We therefore reject H5. INSERT TABLE 8 The pecking order theory predicts that external financing is driven by net financing requirements. Marginal increases in earnings should therefore have an equal and opposite sign to that on investment. Table 7 and Table 8 reveal that the coefficients on cash flow from operations are negative and of opposite sign to those on investment. However, they are frequently insignificantly different from zero and not of equal magnitude to the coefficients on investment. Column 3 of Table 8 pursues this in greater depth by distinguishing between firms that make profits at t = 0 from those that make losses. Negative coefficients mean that debt financing increases when earnings decline. The striking observation from this table is that debt financing increases with declining earnings so long as cash flow from operations are positive. However, if cash flows are negative then increases in losses are not associated with increased debt financing and in the case of small firms with declining debt financing. This is very suggestive of credit constraints operating on small firms subject to losses. We therefore reject H6. 2 Columns 4 and 5 of Table 8 evaluate whether the results may be distorted by firms anticipating financing requirements in t=0 by accumulating reserves in previous periods. Column 4 includes funds from other operations and column 5 flow of funds in the previous period (t=-1). There is some evidence that cash flow from operations at t=-1 increases debt financing in period 0. Again this is suggestive of debt financing being primarily associated with profitable firms. Neither of the two additional variables affect the results about the sequencing of finance. In summary, none of the hypotheses H1 to H6 provide support for the pecking order theory and some are strongly rejected. 2 The fact that debt is not in general associated with the financing of losses, and, as was noted above for small firms, loss makers tend to issue equity may help to reconcile our findings with one of the main observations made by supporters of the pecking-order theory that announcements of new debt tend to have a positive share price effect, while announcements of new equity tend to have a negative price effect. 14

17 4.3 The Trade -Off Theory According to the trade-off theory firms have target levels of leverage that are determined by such considerations as taxation, volatility of earnings and the composition of firms assets. Faced with a new investment, firms should finance their activities in line with their target leverage. Financing proportions in period 0 should therefore be proportional to levels of leverage in previous periods prior to the investment spike. Table 9 examines this by including a term in leverage at t=-2 interacted with investment at t=0 in a regression of debt finance at t=0 on cash flow from operations and investment. Some of the regressions also control for industry by including industry dummies interactively with investment. In none of the size groups is there any evidence of a relation between debt finance and leverage at t=-2, let alone a coefficient equal to unity. Coefficients on cash flow from operations and investment are similar to those in Column 2 of Table 7 and, with the exception of the insignificant term on cash flow from operations for large firms, industry dummies do not substantially alter estimated coefficients on these two variables. INSERT TABLE 9 Figure 2 provides a pictorial representation between the leverage associated with the investment spike ( dynamic leverage ) and leverage levels at t=-2 ( static leverage ). If the static trade-off theory applied then observations would be expected to lie along a 45% line through the origin. There is no evidence of this. The figure also distinguishes between firms that have previously issued new equity, bought back shares or done neither. According to the pecking order theory, firms that have previously issued equity would be less reluctant to issue new equity and might therefore be less constrained in the their financing choices at t=0. Firms that have previously bought back shares will be particularly reluctant to issue new equity and may thus be more constrained in their financing choices. The figure does not suggest that previous new equity activity affects the relation between dynamic and static leverage. We therefore reject the static leverage hypothesis, H7. INSERT FIGURE 2 If firms are constrained in the way in which they can finance the investment spike in period 0 then they may be perturbed from their optimal level of leverage. If so, the dynamic aspects of the trade-off theory predict a subsequent correction of leverage levels and partial adjustment back to previous levels. Table 10 tests this by regressing changes in leverage between t=0 and 2 on changes in leverage between t=-2 and 0. Partial adjustment is indicated by a coefficient between 0 and 1. Table 10 reports significantly 15

18 negative coefficients for two out of three of the size groups. Interestingly, the one size group that does not demonstrate significant partial adjustment is the large firms, which are the firms that raise the most debt. This suggests that small and medium sized firms may be constrained in the amount of debt finance that they can raise at t=0 while large firms are not so constrained in the amount of new equity that they can raise. In any event, we find support for the dynamic version of the trade-off theory as set out in H8 and some evidence of small and medium sized firms being constrained in their financ ing choices. INSERT TABLE 10 To date we have assumed that leverage at t=-2 corresponds to static levels. If, however, firms are not then at their optimum, financing choices around the investment spike should be associated with partial adjustment back to the optimum. Table 11 examines this by repeating the procedure described in Table 9 of including a term in the interaction of leverage at t=-2 with investment. Since adjustment could take place around as well as during the investment spike, the table differs from previous ones by aggregating flows over the period t=-1 to t=1. Panel A reveals that flow of funds from operations over the three years are substantial. Debt financing is a more modest proportion of the investment spike of large firms than Table 5 suggested. However, new equity issues remain a very substantial proportion of financing of small company investment. INSERT TABLE 11 Panel B reports the results of regressing debt finance over the period t=-1 to +1 on investment, cash flow from operations and the interaction of leverage at t=-2 with investment over the period t=-1 to +1. Funds from operations are separated between firms reporting positive profits and those reporting losses. Panel C repeats this regression for new equity finance over the period t=-1 to +1. The first observation is that there is significant partial adjustment in all the size groups for both the financing sources. High levels of leverage are associated with low levels of debt finance and high levels of equity issues. It is possible to derive optimal levels of leverage from the coefficients on investment and leverage interacted with investment. 3 From Panel B these are around 30% for the large and small firms in the sample. Levels of leverage in excess of these are associated with subsequent declines back to these target levels. We therefore accept H9. 16

19 Secondly, the coefficients on the cash flow from operations terms suggest, as in Table 8, that debt financing rises in response to declining profits so long as cash flow from operations are positive. However, if they are negative, in at least the case of small and large firms, there is declining debt finance. Panel C shows that declining earnings are associated with increased equity issuance by large firms so long as cash flow from operations are positive but not if they are negative. This points to a limitation on the ability of large firms to raise equity in financial distress. In contrast, there is very strong increase in equity issuance by small firms in response to decline in earnings when they are making losses. As noted before, equity issuance is therefore a source of financing of small firms in financial distress. We therefore accept H10 and note that equity rather than debt appears to meet the funding requirements of small firms in financial distress. In summary, we have accepted the dynamic version of the trade-off theory. We have noted that a possible reason for why the dynamic rather than static version holds is that firms are constrained in raising external finance. In partic ular we have noted that they are constrained in raising debt when they encounter financial distress. On the other hand, large investment projects may present other firms with an opportunity to rebalance their capital structure by issuing debt against the collateral associated and equity against the future cash flows of the project. We therefore observe a mixture of dynamic adjustment during and subsequent to the investment spike. 5. Conclusions This paper has reported a new approach to evaluating capita l structure and the financing decisions of firms. It describes a method for filtering large investment projects and performing event studies of the financing of the projects. We believe that this approach goes to the heart of issues in corporate finance and capital structure debates. The results to emerge are striking. There are large amounts of external finance of large investment projects and considerable use /of new equity issues as well as debt, in particular by small companies. How do these observations square with aggregate statistics that report a preponderance of internal sources and little new equity finance? We believe quite simply. Aggregate statistics are dominated by the routine financing of, in particular, replacement investment. Most of this is internally funded and swamps the large project financing of this paper. New equity is mainly used by small companies and fails to make an impact in 3 From equation (1) if x is leverage and DEBT 0 = (? +?x)(i 0 -?OPR 0 ) then leverage is constant, i.e. DEBT 0 - xi 0 is zero if a = x(1-ß). Optimal levels of leverage are therefore a/(1-ß). 17

20 aggregate statistics on the much larger debt finance raised by large companies. There is therefore no inconsistency between the two sets of results. However, the aggregate statistics cannot be employed in evaluating the validity of different theories of capital structure. We have been concerned with the two most widely cited theories in this paper: the pecking order and the trade-off theories. We have argued that they have clear predictions for patterns of financing of large projects. We have concluded that the evidence is in clear violation of virtually every prediction of the pecking order theory. Finance is not predominantly internal, external finance does not primarily take the form of debt, debt is not of particular importance for companies most afflicted by information asymmetries, equity finance does not kick in for particularly large investment requirements, companies do not move sequentially up the pecking order from debt to equity, and the response of firms to funding needs created by reductions in earnings are quite different from those of large investments. We also reject a static version of the trade-off theory that the financing of large projects is simply proportional to firms target capital structure. However, we believe that there is evidence in support of a dynamic version of the trade-off theory by which firms adjust to optimal capital structures. Adjustments are observed during and after the investment spike. We have suggested the reason why a dynamic rather than a static version of the trade-off theory applies is that there are constraints on firms choice of finance. We have provided some evidence of these around investment spikes in relation to the financing of losses. On the other hand, investment projects appear to offer opportunities for adjustments towards as well as displacements from optimal capital structures. While we have not provided evidence in this paper, we believe that these opportunities may come from the collateral that investment projects offer against debt finance and the future cash flows that can service equity capital. 18

21 References Baker, M. and J. Wurgler (2002), Market timing and capital structure, Journal of Finance, 57, Bradley, M, G. Jarrell and E. Kim (1984), On the existence of an optimal capital structure: Theory and evidence, Journal of Finance, 39, Chirinko, R. and A. Singha (2000), Testing static tradeoff against pecking order models of capital structure: A critical comment, Journal of Financial Economics, 58, Corbett, J. and T. Jenkinson (1997), How is investment financed? A study of Germany, Japan, the United Kingdom and the United States, Manchester School, 65, Fama, E. and K. French (2002), Testing trade-off and pecking order theories predictions about dividends and debt, Review of Financial Studies, 15, Fischer, E. R. Heinkel and J. Zechner (1989), Dynamic capital structure choice: Theory and tests, Journal of Finance, 44, Frank, M. and V. Goyal (2002), Testing the pecking order theory of capital structure, Journal of Financial Economics, forthcoming. Harris, M. and A. Raviv (1991), Theory of capital structure, Journal of Finance, 46, Mayer, C. (1988), New issues in corporate finance, European Economic Review, 32, Myers, S. (1984), The capital structure puzzle, Journal of Finance, 39, Myers, S. and N. Majluf ((1984), Corporate financing and investment decisions when firms have information investors do not have, Journal of Financial Economics, 13, Rajan, R. and L. Zingales (1995), What do we know about capital structure? Some evidence from international data, Journal of Finance, 50, Shyam-Sunder, L. and S. Myers (1999), Testing static tradeoff against pecking order models of capital structure, Journal of Financial Economics, 51, Titman, S. and R. Wessels (1988), The determinants of capital structure choice, Journal of Finance, 43, Welch, I. (2002), Columbus egg: Stock returns are the main determinants of capital structure, NBER Working Paper

22 Table 1 Average Sources of Finance This table reports average sources of finance used by all the firms in the sample as a percentage of gross investment averaged over all the time periods in the study. Sources of Finance Percentage of Gross Investment Internal 77 Debt 15 Equity 1 Other 7 Source: the raw sample of firms in the study (see Table 3). Table 2 Aggregation Issues Raised by Table 1 Consider two firms A and B that fund a project costing 150. Firm A is internally financed, while firm B is largely externally financed. (+ is a source and - is a use of funds). Over the three periods together (columns 5 and 9 of Panel A), the external flow of finance to both firms is zero. Similarly time period 1 when the only investment activity that is being funded is firm B two-thirds from external sources, the aggregate data in Panel B suggests that two -thirds of total investment is coming from internal sources. Panel A Firm A Firm B Time Aggr Aggr. Investment Internal Sources Liquid Assets External Sources Panel B Firms A + B Time Aggr. Investment Internal Sources Liquid Assets External Sources

23 Figure 1 Investment Strings and Goodness of Fit The figure provides an illustration of the relationship between the ER measure of the goodness of fit (see equation (6)) and different investment strings. We have sampled strings with ERs between 0.1 and 0.4. at ticks of (approximately) Note that the quality of fit is decreasing across strings (i.e. ER is increasing from upper-left to bottom-right). Investment, I, is deflated by the base-level of investment, b, which equals 1 in these figures.? is time index for project year with the spike at?=0. 4 ER==0.104 ER==0.107 ER==0.137 ER== ER==0.180 ER==0.199 ER==0.220 ER==0.240 investment/base level ER==0.260 ER==0.280 ER==0.301 ER== ER==0.341 ER==0.360 ER==0.380 ER== ? 21

24 Table 3 Testing for Selection Bias In Panel A, we look for systematic differences between the raw panel and the filtered sample. Mean refers to averages of company means. In the last column we report the results of a probit regression where the dependent variable gets a value of one if the firm is filtered out, and zero if it is not (z- statistics in brackets). Panel B presents the results of a probit regression on duration variables (z stats). Duration refers to the number of consecutive observations available for each firm. Since there is no reason to believe that the probability of being filtered out is linear in duration, there is a separate dummy variable for each duration (from five to eleven). No other variables are included. The last column presents the results in terms of conditional probabilities Panel A Relation Between Raw Panel and Filtered Sample Mean Probit Raw Panel Filtered Sample Total Assets ($m) (-0.43) Earning/ Assets* (1.63) Market-Book Ratio (-0.03) Debt/Asset (-1.13) Industry Dummies Insignificant Duration Dummies Significant NYSE (0.96) AMEX (-0.37) NASDAQ (0.14) OTHER N R Panel B Probability of Being Filtered Out, Conditional on Duration Duration Dummies Coefficient Conditional Probability (%) 5 years 4.15 (25.99) years 4.27(.) years 4.64 (32.24) years 4.63 (30.07) years 4.79 (34.28) years 4.94 (37.53) years 5.15 (43.57) 16.6 N 7543 R