Firm Financing and Investment Behavior: Fazzari, Hubbard, Petersen
Financing Constraints and Corporate Investment (FHP, 1988)
Introduction
When capital structure is irrelevant, investment (INV) decisions are independent of a firm’s financial condition. However, if internal and external capital are not perfect substitutes, INV may become dependent on financial factors. This study connects conventional investment models with the literature on capital market imperfections and disparities in firms’ access to capital markets.
Key Predictions and Findings
- If the cost disadvantage of external finance is small, earnings retention practices should reveal little about investment, and vice versa.
- The study finds differences across firms in the sensitivity of investment to balance sheet variables measuring liquidity.
- Financial effects on investment are most pronounced when capital market information problems are severe, particularly for high-retention firms (high-dividend firms are generally large).
- Tax Implications: Investment is sensitive to both average tax burdens and marginal tax rates.
Finance and Investment Studies
Previous studies, often based on the Modigliani-Miller (MM) model, asserted the irrelevance of financial factors on real firm decisions. Empirical support, such as that by Jorgenson and Calvin Siebert, appeared consistent with the MM framework. However, the authors identify two key problems:
- Econometric Issues: Concerns regarding time-series data and serial correlation.
- Representative Firm Assumption: The assumption that all firms are identical, regardless of model specifications.
To address these issues, the authors analyze groups of firms with distinct financial characteristics.
Sources and Cost of Finance
Firms of varying sizes exhibit significant heterogeneity in financing practices. For instance, the retention ratio for large firms is typically the smallest among different firm groups.
Cost of Internal Versus External Finance
Internal financing is generally cheaper due to factors such as transaction costs, tax advantages, agency problems, and asymmetric information. The authors particularly emphasize the role of information asymmetry.
- Seasoned Equity Offerings (SEO): New equity issues are often discounted by underwriters. Effective tax rates on capital gains are typically lower than those on dividends. In the context of Q theory, internal financing implies a lower threshold Q value for investment. This aligns with the Pecking Order Theory suggested by Myers and Majluf.
- Debt: The marginal cost of new debt increases due to potential financial distress and agency costs, leading to phenomena like debt overhang. Firms of different sizes also show heterogeneity in accessing the debt market. Large firms can often secure loans from primary banks, whereas small and medium-sized firms frequently rely on private channels that impose significant restrictions on their operations.
Financing Hierarchies and Investment
The study introduces a model for firm valuation. A value-maximizing firm will issue new shares only after exhausting internal finance and when its Q value is greater than 1 (q > 1).
A modified version of the model, which reflects the cost of equity, is also presented. The breakeven Q value for investment financed by new share issues is 1 + Omega (Ω).
Intermediate levels of investment demand are financed by a mix of internal funds and debt. When investment demand is very high, the firm is expected to issue more equity.
Empirically, if information asymmetry is significant, the observed Q should be high relative to historical values before new share issues for information-constrained firms.
Differences in Firm Financing Practices
If the cost disadvantage of external financing is substantial, it should most significantly impact firms that retain the majority of their income. Conversely, if the cost is negligible, retention practices would reveal little about financing practices, Q, or investment behavior.
- Class I firms exhibit an investment demand schedule similar to D2 or D3, while Class III firms resemble D1.
- Table 3 Analysis:
- Class 1 and 2 firms tend to borrow up to their debt capacity and then proceed with Seasoned Equity Offerings (SEO).
- Table 3 also indicates that firms experience varying degrees of information asymmetry over time, reflected in differences in Q values during periods with and without SEOs.
Financial Constraints in Empirical Investment Models
The authors conduct reduced-form investment regressions. To incorporate the information asymmetry argument, they utilize the previously defined firm groups.
Internal Funds in a Q Model of Investment
A significant difference is observed in the estimated coefficients across firm classes. Class 1 firms demonstrate a high reliance on cash flow for investment, whereas Class 3 firms show minimal dependence.
Alternative Estimation Methods and Q Model Specifications
Concerns exist that if the stock market is excessively volatile, Q may not accurately reflect market fundamentals, and the replacement capital stock in Q might be measured with error. Table 5 presents results with various adjustments.
Regardless of the specifications and measures employed, the significant difference in investment behavior between Class 1 and Class 3 firms remains robust.
An alternative interpretation for the effect of cash flow on investment is that cash flow movements reflect productivity shocks not captured by Q (i.e., cash flow may be correlated with disturbances in the adjustment cost function). To address this, the authors use instrumental variables (IV), but the qualitative results persist.
Sales Accelerator Investment Demand Models
An alternative empirical method to the Q model is based on the traditional accelerator principle, which links the demand for capital goods to a firm’s output or sales level/change. Table 7 shows that the explanatory power of cash flow generally decreases with the inclusion of sales. This suggests discrepancies between average Q (AQ) and marginal Q (MQ), or the presence of accelerator effects. However, the qualitative results remain unchanged.
Internal Finance in the Neoclassical Investment Model
Since the accelerator model does not incorporate the relative price of capital or capital services in its empirical specification, the neoclassical investment model suggested by Jorgenson is employed. In perfectly competitive markets, a firm’s optimal demand for capital services ultimately depends on output price and the relative prices of various inputs, including the cost of capital. Following Jorgenson, the authors include the cost of capital alongside sales variables, as they collectively modify the accelerator model into a neoclassical model with partial adjustment assumptions.
Despite this, the qualitative results do not change, even though the model reduces the explanatory power of cash flow on investment.
Investment Equations at the Industry Level
Investment behavior may vary across industry categories. Therefore, the authors segment the samples by both the specified firm classes and SIC 2-digit codes. The results demonstrate robustness across industries.
Balance Sheets, Internal Finance, and Investment
Precautionary Savings
If managers anticipate paying a premium for external funds, they should accumulate a stock of liquid assets when cash flow is high. This strategy helps to:
- Smooth investment during economic downturns.
- Reduce the need for potentially costly external financing.
- Provide necessary collateral for obtaining new debt.
Conversely, Class 3 firms may not perceive the same benefits from precautionary savings (i.e., accumulating liquid assets).
Using cash and equivalents as proxies for liquid assets, the authors demonstrate that changes in balance sheet positions and liquidity significantly affect investment for Class 1 firms. This effect is not observed for Class 3 firms.
Internal Finance and Investment in High-Payout Firms
The study addresses the agency cost of free cash flow, which often leads large firms to distribute earnings. For these firms, if external financing were genuinely costly, they would have already reduced their dividends.