4a Defining Leverage
5. Descriptive Statistics
Table 1 presents major prior empirical findings in connection to the relationship between
major capital structure theories and the factor variables for both measures of leverage (i.e.
Book & Market leverage). A detail discussion on the directional relationship of these
variables is reserved for the empirical results section, where an integrative analysis
provides focus on the fluidity and sometimes conflicting predictions among capital
structure theories and the factor variables. Figure 1 graphically presents the over-arching
concept of both optimal leverage and the adjustment process to target leverage. Within
the trade-off scheme, managers weigh the costs and benefits of debt financing, mainly
driven by tax advantages and probability of bankruptcy. Well noted in corporate finance
is that, debt is relatively cheaper than equity14, thus in a broader scheme of financing costs, debt potentially minimizes the weighted average cost of capital (WACC). However,
debt has a diminishing return to scale after reaching a certain threshold. That is, debt
14One of the reasons has been the relatively lower informational asymmetry
16 could become a burden as the possibility of an involuntary bankruptcy rises15. The
attributes in figure 1 are among the theoretical factors that influences optimal leverage
determination. For instance, tangibility minimizes bankruptcy costs, because fixed assets
can be sold in the process of bankruptcy and therefore enables capital recovery by
creditors. Thus, collaterability is expected to be positively related to debt. In another
instance, the pecking order theory predicts negative relationship between ROA (i.e.
profitability) and leverage, because profitable firms can use the excess cash to finance
investment projects instead of issuing debt or equity. Similarly, in the adjustment
process, firms weigh the costs of deviation from target and the costs of adjustment to
target. Some elements of adjustment process are out of management‟s control and others
depend on firm-level characteristics. Prior evidence suggests that, the two main deciding
factors of adjustment to target are financial flexibility and external financing costs. As
these factors or elements vary across firms; speed of adjustment is expected to differ
across firms as well. Although, it appears that certain mitigating factors help minimize
external financing costs, these factors also vary cross firms, implying varying capital structure adjustment speed among firms‟. Table 2, reports the summary statistics of
leverage and factor variables for the pooled sample, MNCs and DCs. As prior researches
on capital structure of MNCs have shown, MNCs on average have lower debt to equity,
lower target leverage, and lower leverage ratio than DCs (also see table (4) panel B, and
figure 2a and 2b).16 Both book and market leverage (i.e. bolded in table 2) are higher for DCs than MNCs. Book leverage and market leverage for DCs are approximately the
same, whilst book leverage is higher than market leverage for MNCs. Cross-sectional
15Certain indicators of financial distress such as operational risks,
financial risks and debt-induced overall higher costs of financing may trigger involuntary bankruptcy.
16 Fatemi, Ali 1984, 1988; Burgman, T.1996; Lee & Kwok 1988; Michel &
Shaked 1986; Doukas & Pantzalis, 2003; Shapiro, 1978. These authors’ have attributed the relatively lower level of leverage of MNCs to various factors including agency costs.
17 dispersion for DCs is roughly 17% for book leverage and 21% for market leverage. On
the other hand, dispersion for MNCs is 13% for book leverage and 16% for market
leverage. The relatively higher dispersion for market leverage likely reflects the volatility
of equity prices. Tobin‟s Q and non-debt tax shield are almost identical for MNCs and
DCs, which is likely due to the lower asset base of DCs. As expected, MNCs are on
average larger than DCs based on log total assets. Financial distress costs (FDC) is
significantly higher for DCs (27.62) relative to MNCs (2.68). MNCs on average have
higher cash flows than DCs, as well as higher return on assets. However, DCs on average
have higher capital expenditure than MNCs for the same sample period. The average
panel length is approximately 11 years, with 5 years of minimum observations and
maximum observation of 19 years. Table 3A to 3D provides the correlation matrix of the
factors and leverage. The relationship between the factors and both measures of leverage
shows the expected correlations. However, the strength of the correlation of factors with
the two measures of leverage is not stable.
Besides Tobin‟s Q, Cash Flows and Investment, the correlation is more pronounced for
the book leverage. As expected, the two highest correlations for both book and market
leverage was between size and financial distress cost (-.6411) and between tangibility and
investment (.5970). In table 4 panel A, I compute the number and percentage of positive
equity returns by year and by group (i.e. DC, MNC, and MNC10). On a yearly basis for
the sample period, higher number of MNC firms experienced higher equity returns than
their domestic counterparts. Similarly, both leverage percentage returns (i.e. % of firms
with positive returns out of the group subsample – 65% vs. 56%) and the overall median
returns are higher for MNCs than DCs (i.e. 12% vs. 8%). These statistics supports the
market-timing theory of inertia or slower relative adjustment speed for MNCs. Hence
18 cumulative effect of past market equity prices where managers issue equity when market
prices are favourable, the likely resulting effect is lower market leverage, and further
deviation from the target, thus, an inertia or a much slower adjustment speed back to
target leverage (Baker & Wurgler,2002; Welch, 2004). In table 5, I obtain by group, the
number and the percentage of firms‟ with debt capacity more than unity (i.e. ppe/debt>1).
The pecking order theory suggests that firms‟ that are under-leveraged adjust relatively
faster to target leverage than over-levered firms‟. In table 5, average percentage of over- levered firms‟ for DCs are 28% (i.e. 72% under-levered), 36% for MNC (i.e. 64% under-
levered), and 38% for MNC10 (i.e. 62% under-levered). Clearly, DC firms‟ are on
average more under-levered than MNCs, suggesting overall relative faster adjustment for
DCs. Finally, in table 6, I compute average percentage of firms‟ above-target leverage by group. The DC group on average have higher percentage of firms‟ above their target
leverage than the MNC and MNC10 group, thus supporting the trade-off theory that
above-target firms‟ adjust faster to target leverage than firms‟ below their target leverage.
The combination of relatively lower equity returns, higher percentage of under-leverage,
higher percentage of above-target leverage for DCs, and the relative lower debt-equity
ratio lower leverage ratio for MNCs provides a strong recipe for overall relative faster
adjustment for DCs.