q(f)-

u\C(t))

dr. (8.29)

As Craine (1989) emphasizes, (8.29) Implies that the impact of a projects riskiness on investment in the project depends on the same considerations that determine the impact of assets riskiness on their values in the consumption CAPM. Idiosyncratic risk-that is, randomness in MK) that is uncorrelated with u(C) -has no impact on the market value of capital, and thus no impact on investment. Bul uncertainty that is positively correlated with aggregate risk-that is, positive correlation of ( ) and C, and thus negative correlation of { ) and u(C)-lowers the value of capital and hence reduces investment. And uncertainty that is negatively correlated with aggregate risk raises investment.

8.7 Financial-Market Imperfections

Introduction

When firms and investors are equally weU informed, financial markets function efficiently. Investments are valued according to their expected payoffs and riskiness; as a result, they are undertaken if their value exceeds the cost of acquiring and installing the necessary capital. These are the assumptions underlying our analysis so far. In particular, we have assumed that firms make investments if they raise the present value of profits evaluated usmg the prevailing economy-wide mterest rate; thus we have implicitly assumed that firms can borrow al that interest rate.

In practice, however, firms are much better informed about their investment projects than potential outside investors are. Outside financing must ultimately come from individuals. These individuals usually have little contact with the firm and little expertise concerning the firms activities. In addition, their slakes in the firm are usually low enough that their incentive to acquire relevant information is smaU.

Uncertainty about Discount Factors

Firms are uncertain not only about what their future profits will be, but also about how those payoffs will be valued. To see the effects of this uncertainty, suppose the firm is owned by a representative consumer. As we saw in Section 7.5, the consumer values future payoffs not according to a constant interest rate, but according to the marginal utiUty of consumption. The discounted marginal utility of consumption at time , relative to the marginal utility of consumption al (, is e " ( ( ))/ ( ( )), where p is the consumers discount rate, is the instantaneous utility function, and is consumption. Thus our expression for the value of a unit of capital, (8.26), becomes

Assumptions

An entrepreneur has the opportunity to undertake a project that requires 1 unit of resources. The entrepreneur has wealth of W, which is less than 1; thus he or she must obtain 1 - W units of outside financing to undertake the project. If the project is undertaken, it has an expected output of y, which is positive, is heterogeneous across entrepreneurs, and is pub-ficly observable. Actual output can differ from expected output, however; specifically, the actual output of a project with an expected output of is distributed uniformly on [0,2 ]. Since the entrepreneurs wealth is aU invested in the project, his or her payment to the outside investors cannot exceed the projects output. This limit on the amount that the entrepreneur can pay to outside investors means that the investors must bear some of the projects risk.

Because of these problems, institutions such as banks, mutual funds, and bond-rating agencies that specialize in acquiring and transmitting information play central roles in financial markets. But even they are much less informed than the firms or individuals in whom they are investing their funds. The issuer of a credit card, for example, is usually much less informed than the holder of the card about the holders financial circumstances and spending habits. In addition, the existence of intermediaries between the ultimate investors and Arms means that there is a two-level problem of asymmetric information: there is asymmetric information not just between the intermediaries and the firms, but also between the individuals and the intermediaries (Diamond, 1984).

Asymmetric information creates agency prob/ems between investors and firms. Some of the risk in the payoff to investment is usually borne by the investors rather than by the firm; this occurs, for example, in any situation where there is a possibility that the firm may go bankrupt. When this is the case, the firm can change its behavior to take advantage of its superior information. It can only borrow if it knows that its project is particularly risky, for example, or it can choose a high-risk strategy over a low-risk one even if this reduces expected returns. Thus asymmetric information can distort investment choices away from the most efficient projects. In addition, the presence of asymmetric information can lead the investors to expend resources monitoring the firms activities; thus again it imposes costs.

This section presents a simple model of asymmetric information and the resulting agency problems, and discusses some of their effects. We will find that when there is asymmetric information, investment depends on more than just interest rates and profitability; such factors as investors ability to monitor firms and firms ability to finance their investment using internal funds also matter. We will also see that asymmetric information changes how interest rates and profitability affect investment.

The Equilibrium under Symmetric Information

In the absence of the cost to outside investors of observing the projects output, the equilibrium is straightforward. Entrepreneurs whose projects have an expected payoff that exceeds 1 - obtain financing and undertake their projects; entrepreneurs whose projects have an expected output less than 1 -I- r do not. For the projects that are undertaken, the contract between the entrepreneur and the outside investors provides the investors with expected payments of (1 - W){1 + r). There are many contracts that do this. One example is a contract that gives to investors the fraction (1 - If )(1 - r)/y of whatever output turns out to be; since expected output is y, this yields an expected payment of (1 - W)(l - r). The entrepreneurs expected income is then - (1 - W){1 -b r), which equals W{1 - ) - - (1 - r). Since exceeds 1 - by assumption, this is greater than W(l - r). Thus the entrepreneur is made better off by undertaking the project.

The Form of the Contract under Asymmetric Information

Let us now reintroduce the assumption that it is costly for outside investors to observe a projects output. In addition, assume that each outsiders wealth is greater than 1-W. Thus we can focus on the case where, in

If the entrepreneur does not undertake the project, he or she can invest at the risk-free interest rate, r. The entrepreneur is risk-neutral; thus he or she undertakes the project if the difference between and the expected payments to the outside investors is greater than (1 + r)W.

The outside investors, like the entrepreneur, are risk-neutral and can invest their wealth at the risk-free rate. In addition, the outside investors are competitive; thus in equilibrium their expected rate of return on any financing they provide to entrepreneurs must be r.

The key assumption of the model is that entrepreneurs are better informed than outside investors about their projects actual output. Specifically, an entrepreneur observes his or her output costlessly; an outside investor, however, must pay a cost to observe output, is assumed to be positive; for convenience, it is also assumed to be less than expected output, y.

This type of asymmetric information is known as costly state verification (Townsend, 1979). We focus on this type of asymmetric information between entrepreneurs and investors not because it is the most important type in practice, but because it is relatively straightforward to analyze. Other types of information asymmetries, such as asymmetric information about the riskiness of projects or entrepreneurs actions, have broadly similar effects.