Financial Structure ECON 40364: Monetary Theory & Policy

Eric Sims

University of Notre Dame

Fall 2018

1 / 34 Readings

I Text: I Mishkin Ch. 8

2 / 34 Financial Structure

I Firms all have balance sheets – they finance assets with some mix of equity and debt (liabilities)

I Non-financial firms: these assets are real assets (capital) I Financial firms: these assets are financial assets (contractual claims)

I How do non-financial firms finance their assets? Does it matter? Why does it matter? I A couple of useful distinctions: I External vs. internal: raise funds externally or use retained earnings I Equity vs. debt: promised share of cash flows from assets (equity) or promised fixed payments (debt) I Direct vs. indirect: raise funds directly from lender/equity investor or indirectly through financial intermediary

I Modigliani-Miller: firm financial structure is irrelevant, but assumptions underlying this don’t seem particularly realistic

3 / 34 How Do Businesses Finance Their Activities?

I A significant fraction of business investment comes from external funds I Two sources of external funds: 1. Direct: get funds directly from a lender or equity investor 2. Indirect: get funds indirectly from a financial intermediary (e.g. a bank)

I Indirect finance is mostly comprised of debt contracts, whereas direct finance could either be debt (e.g. issue corporate bonds) or equity (e.g. issue new stock)

I Fact: indirect finance is more important than direct finance, particularly for all but the very largest firms, and bank loans (use of financial intermediary) are really important

I This is why financial structure is relevant for monetary policy – so much depends on banking

I Financial intermediation is process by which funds get routed from savers to non-financial firms

4 / 34 Sources of External Funding

5 / 34 Why is Financial Intermediation so Important?

I Chiefly for two reasons: 1. Transactions costs 2. Informational asymmetries

I Transaction costs: cheaper to finance projects on a large scale, which means it is efficient to pool lots of small resources and have an intermediary invest it rather than each small saver doing the investment directly

I We will focus mostly on informational asymmetries

6 / 34 Asymmetric Information

I Asymmetric information generally refers to a situation in which different parties to a transaction are not equally informed about characteristics or actions of the other parties to the transaction I Two main kinds of asymmetric information: 1. Adverse Selection: information asymmetry which occurs before a transaction takes place 2. Moral Hazard: information asymmetry which occurs aftera transaction takes place

I Both types of asymmetric information can help us understand the kind of financial structure we observe in the real world – in particular, why indirect finance is so important (and hence why financial intermediation is important)

7 / 34 Adverse Selection

I The buyer of a product (e.g. a car, a stock) doesn’t know the true “type” of the seller of the product (e.g. good or bad, risky or safe)

I Only knows the average type of the seller I Hence, buyer will only be willing to pay the average valuation, which is more than the bad type but less than the good type

I This tends to drive sellers who are a good type away and attract sellers who are a bad type

I But then buyer knows this, and entire market can fall apart I Easy to understand through an example – “lemons” in the market for used cars

8 / 34 Lemons Example

I After Ackerlof (1970) I Suppose there are two types of used cars: lemons (bad) and peaches (good)

I Sellers know whether they have a lemon or a peach, but buyers only know the fraction of lemons and peaches out there

I Suppose each type has the following valuations:

Valuation Peach Lemon Buyer $20,000 $15,000 Seller $18,000 $13,000

I Without informational asymmetry, both kinds of cars would be sold – buyers value each type more than sellers

9 / 34 Asymmetric Information

I Suppose buyer doesn’t know whether she is meeting a seller of a peach or a lemon

I She only knows there is a 50 percent chance it’s a peach, and 50 percent chance it’s a lemon

I The average valuation for the buyer is $17,500, which is the maximum she will pay for a car

I But since this is less than a peach owner’s valuation, peaches will not be sold

I But then the buyer will know only lemons are on the market I Only lemons will sell for a price between $13,000 and $15,000 I The “good” cars get driven out of the market by the presence of bad cars – inefficient!

10 / 34 Alternative Example

I Suppose valuations are now:

Valuation Peach Lemon Buyer $20,000 $12,000 Seller $18,000 $13,000

I Buyer values lemons less than seller. With symmetric information, only peaches would be sold

I Suppose probabilities of peaches and lemons are same as above. Average valuation from buyer’s perspective is now is $16,000

I Since this is less than seller’s valuation, peaches will not be sold

I But then buyer knows she can only buy a lemon, but doesn’t want a lemon

I End result: market breaks down and no cars are sold

11 / 34 Dealing with the Lemons Problem in the Used Car Market

I Most used car deals are through dealerships, not person-to-person transactions

I Sort of easy to understand why I The dealership serves as an intermediary and helps solve the informational problem

I The dealership gets good at determining lemons vs. peaches, and can offer warranties to buyers to ensure that the buyer isn’t dealing with a lemon

I Hence, intermediaries who specialize in resolving informational asymmetry problem naturally arise in the car market

I Similarly in financial markets

12 / 34 Lemons Problem and Indirect Finance

I Lemons problem helps us understand why direct finance is not that important for external funds

I The firms who most want funds are probably those who are the worst type

I But savers know this I And hence will not buy stock or debt directly from firms I Financial intermediaries (e.g. banks) can step in I These financial intermediaries can become experts in learning about firms and can therefore alleviate the informational asymmetry problems

I Because these financial intermediaries make private loans they can avoid the free rider problem that arises when third party firms try to produce information about firms

13 / 34 Example: Risky and Safe Firms

I There are two types of firms who need 1 unit to undertake a project

I Project succeeds or fails with a given probability I Firm types and payoffs are:

Safe Firm Risky Firm Payoff in “good” state 4 8 Payoff in “bad” state 0 0 1 1 Prob. of “good” state 2 4

I Expected return the same for both firms, but lender would prefer to loan to safe firm since it is less risky

14 / 34 One Kind of Debt Contract

I Suppose there is only one kind of debt contract: bank lends firm one unit, firm promises to repay R (gross) units if project succeeds, 0 otherwise (it can’t pay back in event low state occurs)

I Borrower only has to pay back in event good state materializes. Borrower expected payoffs are: 1 Safe = (4 − R) 2 1 Risky = (8 − R) 4

I Borrower will only take a loan if his/her expected payout is non-negative

I Hence, if R > 4, safe firms won’t take loan I If R > 8, neither firm will take loan

15 / 34 Expected Payoffs for Borrower

Borrower Expected Profit

Safe firm 2

Risky firm 1

4 8 𝑅𝑅

16 / 34 Lender Problem

I Lender’s opportunity cost of funds is 1 – if it doesn’t earn at least 1 in expectation, it won’t make a loan

I If lender charges R > 8, it will make no loan and hence “earns” 1 (i.e. keeps its money)

I If it charges R ≤ 4, both types of firms will take the loan I If it charges R > 4, only the risky type firm will take the loan I Suppose fraction q of firms are risky, and 1 − q are safe. Lender expected payout: 1 1 R ≤ 4 E (payout) = (1 − q) R +q R 2 4 |{z} |{z} Safe Risky 1 4 < R ≤ 8 E (payout) = R 4

17 / 34 Pooling vs. Separating Equilibrium

I A pooling equilibrium is a value of R (i.e. a debt contract) in which both types of firms take a loan I A separating equilibrium is a value of R in which only one type of firm gets a loan, and the other type of firm chooses to sit out I Let’s first look for a pooling equilibrium to see if one exists (i.e. focus on R ≤ 4). Can write lender’s expected payout as: 1 1  E (payout) = − q R 2 4

I Suppose q = 0.9, so most firms are risky. Expected payout must be at least 1. Solve for the “break-even” R:

R ≥ 3.6364

I So 3.6364 ≤ R ≤ 4 would be a pooling equilibrium, whereas 4 < R ≤ 8 would be a separating equilibrium.

18 / 34 Lender Expected Payoffs

Lender Expected Profit

0.2750 ×

𝑅𝑅 1.1 1.0 0.25 ×

𝑅𝑅

3.64 4 8 𝑅𝑅

19 / 34 Equilibrium

I A pooling equilibrium exists for 3.64 ≤ R ≤ 4 I A separating equilibrium exists for 4 < R ≤ 8 I Don’t know which equilibria we’ll end up at I But if it’s separating equilibrium, safe firm doesn’t get a loan, which is a bad outcome relative to symmetric information case

I If it’s pooling equilibrium, interest rate charged to safe firm may be “too high” relative to symmetric information case and interest rate charged to risky firm is “too low”

I This will tend to over-attract risky firms and deter safe firms from getting loans

20 / 34 Adverse Selection and Collateral

I Collateral is an important feature of many debt contracts I A firm receiving funds pledges some collateral that can be seized in the event that the firm defaults

I Banks can offer different kinds of contracts – some require posting more collateral and some require less collateral but charge higher interest rates

I This offering different kinds of contracts can get firms to voluntarily reveal their type

I We see exactly same thing in mortgage finance – the more you put down (more collateral), the better the terms on the loan typically

I So collateral can be a useful way for financial contracts to deal with information asymmetry. When collateral loses value (“bubble bursting”) this can exacerbate information asymmetry problems

21 / 34 Loan Contracts with Collateral

I Suppose the lender requires borrower to post collateral, C, which borrower has to pay in event project fails

I Then borrower expected payouts are: 1 1 Safe = (4 − R) − C 2 2 1 3 Risky = (8 − R) − C 4 4

I Observe that R “hurts” the safe firm more, and C “hurts” the risky firm more

I Suppose that the lender posts two contracts, one with (R, C = 0) and the other (RC , C > 0). The idea is to get the safe firm to post collateral and thereby reveal itself

22 / 34 Requirements

I Suppose that, if lender seizes collateral, a fraction d goes bad, so lender only recovers (1 − d)C. Think of this as a “bankruptcy cost” I For this to work, we must have the following hold: 1. Risky firm prefers no collateral contract 2. Safe firm prefers posting collateral 3. Lender breaks even (or better) on both contracts

23 / 34 Risky Firm Prefers No Collateral

I This requires that: 1 3 1 (8 − R ) − C ≤ (8 − R) 4 C 4 4

I This requires: R − RC ≤ 3C

I We can see: 1. R > RC (you get a lower interest rate if you post collateral) 2. Collateral must be sufficiently big to induce risky firm to take a higher interest rate

24 / 34 Safe Firm Prefers Posting Collateral

I This requires that: 1 1 1 (4 − R) ≤ (4 − R ) − C 2 2 C 2

I Which works out to:

R − RC ≥ C

I We can see: 1. R > RC again 2. Collateral can’t be too big, otherwise safe firm won’t post it

25 / 34 Lender Must at Least Break Even on Both Contracts

I This requires: 1 R ≥ 1 4 1 1 R + (1 − d)C ≥ 1 2 C 2

26 / 34 All Conditions

I All together, we have:

R − RC ≤ 3C

R − RC ≥ C R ≥ 4

RC + (1 − d)C ≥ 2

I Will be multiple solutions. One possibility: lender just breaks even on both contracts and risky firm is indifferent between contracts (weakly prefers the no-collateral contract) 1 I Suppose d = 4

27 / 34 Lender Just Breaks Even

3 I Would mean: R = 4 and RC + 4 C = 2 which means:

3C = 8 − 4RC

I Risky firm just breaks even:

4 − RC = 3C

= 8 − 4RC

4 8 I Implies RC = 3 and C = 9 I The firms voluntarily separate

28 / 34 Collateral, Adverse Selection, and Business Cycles: Financial Accelerator

I Posting of collateral allows safe firms to reveal their type. Allows them to get loans (depending on market structure) and results in more efficient allocation

I Since collateral consists of assets, asset price fluctuations can affect ability of firms to get loans I Financial accelerator: 1. Decline in economic activity (e.g. a recession) causes assets to lose value 2. Declining asset values makes it harder for firms to post collateral 3. Inability to post collateral exacerbates adverse selection problem, resulting in less investment and/or a worse allocation of investment between safe and risky firms 4. Less investment causes more declines in economic activity, and further falls in collateral values 5. An adverse feedback loop!

29 / 34 Moral Hazard

I Moral hazard: information asymmetry which occurs after a transaction takes place

I For example, someone lends you money, but then can’t perfectly monitor what you do with the money

I Because of limited liability, you have an incentive to “gamble” with someone else’s money

I In other words, moral hazard can encourage excessive risk taking once a loan has been made

I Can be applied to insurance markets too – once you have insurance, you have less incentive to behave safely

I But insurer will know that, and may not sell you insurance in first place

I Just like adverse selection, markets can break down

30 / 34 Moral Hazard and Financial Structure

I Moral hazard can also help us understand why indirect finance is more important than direct finance

I Intermediaries (e.g. banks) become experts in monitoring the behavior of borrowers in a way that wouldn’t be possible with direct finance: leads to less “gambling” and explains why loan contracts often include covenants restricting behavior

I Also helps make sense of preference of debt over equity I With equity, lender needs to monitor profits all the time (since equity owner is due his/her share of profits)

I With debt, since payments are fixed, only need to monitor behavior of firm in event of default

I So lower monitoring costs (what is called “costly state verification”) with debt over equity

31 / 34 Moral Hazard and Collateral

I Collateral also plays a role in mitigating moral hazard I Requiring firms to post collateral gives them some “skin in the game” and encourages good behavior

I Without collateral, lenders may be reluctant to lend because they can’t perfectly control what borrowers do

I Similarly to adverse selection, this importance of collateral can give rise to a financial accelerator mechanism

I If assets decline in value, harder to post collateral, harder for firms to get loans

32 / 34 Economic Development and Financial Structure

I Developing countries often have poorly developed legal systems and ill-defined property rights

I This makes it difficult for collateral to serve the role it does in developed economies like the US

I It is also more difficult for lenders to monitor the behavior of borrowers

I As a result, the financial system is underdeveloped I This makes it difficult to funnel savings to investments (particularly the most profitable investments), which results in weak economic growth

33 / 34 Relevance for Monetary Policy

I Financial intermediation (indirect finance) is much more important than direct finance for external funding for firms

I Asymmetric information (adverse selection and moral hazard) can help explain this phenomenon

I Asymmetric information also affords an important role to collateral in indirect finance

I Possibility of financial accelerator mechanism I Relevance for Monetary Policy: 1. Banking system (over which central banks have some control) really important for functioning of economy 2. Fluctuating asset prices (e.g. bubbles) can impact ability of banking system to funnel savings into productive investments

34 / 34