This paper examines the value to a firm from information spillovers arising from its publicly traded stock. “Information spillover” refers to the positive information externalities when different investor groups produce information about the firm. Specifically, information produced by investors in the stock market is noisily reflected in an observable stock price for the firm and this information can be used by other investors in the firm.
The notion of information spillover was explored in Grossman and Stiglitz (1980). They show that information produced by informed investors gets communicated to the uninformed through stock prices, although this is achieved imperfectly. Several papers have since modelled the impact of information spillovers from public stock markets in different contexts.
In this paper, I provide an empirical estimate of the value of having an informative stock price using financing costs to measure the benefits. A testable implication of the information spillover hypothesis is that firms face lower costs of financing on all their securities when they have publicly traded stock which aggregates and reflects information produced by the stockholders. This results in an information spillover across different securities of the same firm resulting in reduction in information asymmetry and consequently lower overall financing costs. Specifically, the study focuses on the savings in the firm’s bank borrowing costs due to information produced in the stock market. Consistent with the hypothesis, I find that controlling for firm and loan characteristics, the cost of bank borrowing is lower in the presence of public equity and it is decreasing in measures of information production in the stock market.
Under the information spillover hypothesis, private firms with no publicly traded equity should face higher costs of bank borrowing relative to publicly traded firms since these private firms do not have a publicly observable stock price as an informative signal of firm value. Sunder (2002) models how all investors, current and future, use the information contained in stock prices to monitor the firm and value its claims. In the model, information spillovers from publicly observable stock prices lowers financing costs by reducing monitoring costs of other investors and decreasing the adverse selection costs faced by future investors in the firm. For instance, if the firm has publicly traded stock, a lending bank can supplement its own monitoring effort with the noisier but cheaper information contained in the firm’s publicly observable stock price.
The empirical tests are conducted on two different samples of private firms: (i) loans by firms prior to the IPO event (pre-IPO sample) and (ii) loans by firms after they go private (post-LBO sample). These loans are compared to loans by public firms to measure the value of publicly traded stock on the cost of bank debt. The cost of bank borrowing, as measured by spreads over LIBOR, is compared for loans of private versus public firms after controlling for firm risk and loan characteristics. To proxy for the information that the bank already has, I use various financial measures of firm risk (such as size, leverage, profitability, liquidity ratios, etc.) and non-price terms of loans (such as maturity, loan size, rating, security, etc.) which capture the residual risk.
I find that the cost of borrowing is statistically and economically significantly lower for firms with observable stock prices (after their IPO) after controlling for firm risk and loan characteristics. For the pre-IPO sample of private firms, the cost of bank debt reduces by about 30 to 75 basis points in different specifications after these firms go public. In other tests, I show that the cost of bank debt for the pre-IPO firms is significantly larger than a sample of public firms matched on characteristics such as industry, size, and leverage, by as much as 120 basis points. In similar tests on the LBO sample, these firms face higher bank spreads of about 40 to over 100 basis points in different specifications after they delist their stock.
These results suggest a regime shift in the financing costs for these sample firms which would carry forward to all future loans. In additional test, I show that the reduction in borrowing costs is related to stock informativeness and the presence of informed traders, measured by the relative bid-ask spreads, analyst coverage and institutional holdings. This suggests that information production in the stock market has an externality in terms of overall financing costs. For instance, a one standard deviation decrease in relative bid-ask spreads reduces the bank spread by over 18 basis points on average. In fact, this is a conservative estimate of the value of informative stock prices since the bank can be considered among the better informed investors of the firms and therefore the benefit to the bank represents a lower-bound of the benefit of having informative stock prices. Finally this paper provides some preliminary evidence on the indirect benefits of information production in stock markets to other market participants such as rating agencies and competing banks which ultimately reduces the firms financing costs.
The results are subjected to several robustness checks and alternative model specifications. The sample used in the study could potentially suffer from some sample endogeneity issues. In order to control for the sample selection biases that may be associated with the sample of firms that go public, I estimate the model with a treatment effects correction and the results continue to be significant. In order to interpret the results as benefits of having publicly traded stock, it is important that changes in firm risk are adequately controlled for.
In the tests where the same firms are studied pre-and post-IPO, the IPO event is such a significant event in the life of the firm that there may be some unobserved changes in the risk profile associated with the IPO event which might not be fully captured in the model specifications. Also for these firms it may be hard to separate out the information effects associated with the marketing of the IPO from the information effects of having a publicly traded stock. To address these issues, I use a matched sample of contemporaneous loans by public firms with similar characteristics as the pre-IPO sample, such as, industry, size and leverage. Further, the tests that document the cross sectional effect of proxies of stock informativeness on the cost of bank debt are not affected by changes in risk around the IPO event.
Overall the findings provide evidence that information production in stock markets is valuable to all investors in the firm. Using a relatively sophisticated investors such as banks, the estimates of savings in the cost of debt represent a lower bound of the value of these spillovers. Consistent with my findings, KMV Corporation of San Francisco uses stock prices as a key input in a default prediction model used by banks to predict loan default rates for all companies with publicly traded equity (Saunders (1998)).
The empirical findings also contribute to the large empirical literature on bank uniqueness which show that stock market investors try to infer a bank’s superior information from actions of the bank. These studies document significant stock market reactions to loan initiations, renewals or sales. The bank is assumed to have superior information due to its lending relationship and therefore its actions convey this information to the market. In contrast to these papers, my tests focus on information that flows from markets to the bank and evaluate the impact of these information spillovers on the efficiency and cost of bank monitoring. The results in this paper therefore suggest complementarities in information production between well-functioning stock markets and the banking system.
The remainder of the paper is structured as follows. Section 2 contains a discussion of the related literature.Section 3 outlines the sample selection, methodology and describes the variables. Section 4 discusses the empirical results on the IPO sample. Section 5 relates the borrowing costs to various measures of information production in stock markets. Section 6 provides the results in the going-private sample and Section 7 concludes the paper.
Download
PDF Ebook Information Production in Stock Markets and Cost of Bank Debt
