(Some theoretical aspects of) Corporate Finance

(Some theoretical aspects of) Corporate Finance V. Filipe Martins-da-Rocha Department of Economics UC Davis Part 6. Lending Relationships and Investor Activism V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 1 / 34

Outline 1 Basics of Investor Activism: Monitoring 2 Learning by lending V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 2 / 34

The model We start from the fixed-investment model We add a monitor who can intervene in order to reduce the scope for moral hazard V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 3 / 34

The model A risk-neutral entrepreneur with wealth A has to borrow I A in order to implement a project costing I The project yields R when it succeeds 0 when it fails The probability of success is ph if the entrepreneur works pl = p h p if he shirks V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 4 / 34

No monitoring In the absence of monitoring, shirking provides the private benefit B Denote by R b the entrepreneur s reward in the case of success (he receives nothing in the case of failure) Incentive compatibility requires that ( p)r b B Funding requires that the pledgeable income exceeds the investor s investment ( ρ 0 p h R B ) I A p If this condition is satisfied, and given that the investors breakeven condition is binding for an optimal contract, the entrepreneur s net payoff is equal to the project s NPV: Π b = p h R I V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 5 / 34

Monitoring (with fixed intensity) A monitor can reduce the private benefit that can be enjoyed by the entrepreneur from B to b < B The monitor must bear a private non-observable (by the entrepreneur) monitoring cost c > 0 A possible interpretation is that there are three projects good project bad project Bad project Prob(success) p h p l p l Private benefit 0 b B V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 6 / 34

Monitoring (with fixed intensity) The monitor moves first If he incurs effort cost c, he is able to identify the high-private-benefit Bad project and thus prevent the entrepreneur from selecting it he still cannot tell the other two projects apart The entrepreneur can condition his choice of project on the existence or absence of monitoring The entrepreneur can still choose the low-private-benefit bad project if he wishes to If the monitor does not incur the monitoring cost c then the investor can choose among the three projects (the low-private-benefit bad project is then irrelevant) V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 7 / 34

Monitoring (with fixed intensity) Assume that the entrepreneur hires a monitor The monitor should have incentives to incur the monitoring cost provided a reward R m in the case of success satisfying ( p)r m c The entrepreneur s private benefit from shirking is then equal to b The entrepreneur works if and only if the reward R b satisfies ( p)r b b We will assume that ( p)r b < B otherwise the entrepreneur is induced to work even in the absence of monitoring (which is then useless) V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 8 / 34

References For a discussion concerning the structure of incentives within the monitoring entity, we refer to Berger, A. and Miller, N. and Petersen, M. and Rajan, R. and Stein, J. Does function follow organizational form? Evidence from the lending pratices of large and small banks Journal of Financial Economics 76, 2005 V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 9 / 34

Abundance of monitoring capital Assume that there is a large supply of monitors who are willing to invest their capital in the monitoring activity as long as they breakeven They are thus willing to contribute to the firm s investment at level I m such that p h R n c = I m Then the monitor obtains no rent and receives the net payment p h R m I m equal to his monitoring cost c V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 10 / 34

Non-monitoring Investors Non-monitoring investors are willing to fund the project if and only if p h (R R b R m ) I A I m It follows that a necessary and sufficient for the project to be funded is ( p h R b ) I A I m p Monitoring reduces the agency costs from p h B/ p to p h b/ p But it adds monitoring cost c V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 11 / 34

The monitor The monitor s stake R m can be chosen equal to R m = c p The monitor s investment contribution can be chosen equal to I m = p lc p To obtain some potential role for monitoring, let us assume that monitoring increases the pledgeable income: p h b p + c < p B h p V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 12 / 34

The entrepreneur When does the entrepreneur benefit from having a monitor? Given the choice of the monitor s reward and investment, for the optimal contract, the non-monitoring investor must breakeven with equality Neither the monitoring nor the non-monitoring investors obtain a rent, therefore the payoff of the entrepreneur coincides with the NPV, i.e., Π b = p h R I c We assume that the NPV is positive even in the presence of monitoring p h R > I + c V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 13 / 34

Beneficial monitoring Monitoring reduces the entrepreneur s payoff by the monitoring cost The entrepreneur forgoes monitoring if he can obtain funding in its absence, i.e., if ( p h R B ) I A p We recall the threshold A defined by ( A I p h R B ) p Entrepreneurs with strong balance sheets, i.e., A A borrow cheaply because they can do without monitoring V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 14 / 34

Beneficial monitoring We introduce the threshold A defined by ( A I + c p h R b ) p Assuming that monitoring increases the pledgeable income then we get p h b p + c < p B h p A < A Borrowers with weaker balance sheets, i.e., borrow more expensively A A < A V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 15 / 34

References Empirical papers demonstrating the role of banks in the reduction of agency costs James, C. Some evidence on the uniqueness of bank loans Journal of Financial Economics 19, 1987 Lummer, S. L. and McConnell, J. Further evidence on the bank lending process and the capital-market response to bank loan agreements Journal of Financial Economics 25, 1989 Cantillo, M. and Wright, J. How do firms choose their lenders? An empirical investigation Review of Financial Studies 13, 2000 V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 16 / 34

Optimal Monitoring We extend the monitoring model by introducing uncertainty about the outcome of monitoring The monitor discovers the identity of the Bad project (the one yielding private benefit B) with probability x and learns nothing with probability 1 x The probability x of effective monitoring depends on the unverifiable effort cost or disutility of effort c(x) incurred by the monitor We assume that this disutility of effort is strictly increasing, convex and satisfies c(x) c(0) lim = 0 and lim x 0 + x x 1 c (x) = V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 17 / 34

Optimal Monitoring We assume that the borrower s reward R b in the case of success is smaller than B/ p (otherwise monitoring is useless) Assuming that monitoring capital and non-monitoring capital are abundant, they only require breakeven to participate to the financing of the project If the loan agreement is characterized by the monitoring intensity x then the net payoff of the borrower equals the NPV Π b (x) = xp h R + (1 x)(p l R + B) I c(x) The optimal monitoring intensity x must satisfy x (0, 1) and ( p)r B = c (x ) V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 18 / 34

Optimal Monitoring Let us assume that at this level of monitoring, there is enough pledgeable income to pay back investors, monitoring and non-monitoring: { [x p h + (1 x )p l ] R b } I A + c(x ) p We still assume that unmonitored borrowing is infeasible, i.e., [ ρ 0 p h R B ] < I A p V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 19 / 34

Optimal Monitoring The borrower cannot observe the monitoring intensity x, therefore the reward R m should be chosen such that x argmax{[xp h + (1 x)p l ]R m c(x) : x [0, 1]} and so that is, ( p)r m = c (x ) R m = R B p V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 20 / 34

Optimal Monitoring Because the entrepreneur is unable to borrow in the absence of monitoring, we have implying that R b < B p R m < R R b The monitor should not hold all external shares in the firm V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 21 / 34

Reference Pagano, M. and Roell, A. The choice of stock ownership structure: agency costs, monitoring, and the decision to bo public Quarterly Journal of Economics 113, 1998 V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 22 / 34

Learning by lending Now there are two periods t {1, 2} There is a discount factor β between the two periods Saving is impossible between the two dates V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 23 / 34

Date 1 Consider an entrepreneur without cash A = 0 He has a project requiring investment I at date t = 1 This project is successful with probability p and yields R fails with probability 1 p and yields 0 There is moral hazard p {p h, p l } and two types of misbehavior: low-private-benefit b of the bad project high-private-benefit B of the Bad project V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 24 / 34

Date 1 The private benefit B is large enough such that there is not enough pledgeable income to reimburse the initial investors if there is no monitoring There is no scarcity of monitors who incur cost c for monitoring Monitoring brings down the private benefit to b < B and generates enough pledgeable income to pay back the investment and the monitoring cost, i.e., ( p h R b ) I + c p This implies that the project can be financed V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 25 / 34

Date 2 Regardless of the first period profit, the entrepreneur has a new idea This second project (which can be thought of as a continuation of the first) is identical to the first project, except for one thing With probability α, the date-2 probability of success has increased uniformly by τ > 0 With probability 1 α the probabilities of success and failure are still p h and p l It is assumed that even with this improved profitability, the high private benefit B is so large that a monitor is still needed The profit realizations (success, failure) are statistically independent across periods V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 26 / 34

Date 2 We assume that there is no commitment (short-lived claims) First-period investor s return only comes from the first-period profit The firm issues new claims in the second period to finance continuation V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 27 / 34

Symmetric information Assume there is symmetric information at date t = 2: No one learns the realized date-2 profitability All agents (entrepreneur, monitors and uninformed investors) expect a date-2 profitability p h + τ with probability α and p h with probability 1 α The market for active monitors is competitive at all dates The (expected) net payoff of the entrepreneur at date t coincides with the NPV of the project where Π b (t) = p h (t)r (I + c) p h (1) = p h and p h (2) = p h + ατ The overall (expected) net payoff of the entrepreneur is then Π b = [p h R (I + c)] + β [(p h + ατ)r (I + c)] V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 28 / 34

Asymmetric information Assume now that the date-1 active monitor (called incumbent ) learns the date-2 profitability and is the only one to learn it The entrepreneur defines the active monitor s stake R 2 m in the case of success (0 in the case of failure) as follows R 2 m = c p The entrepreneur announces that the active monitor will be the bidder offering the highest investment contribution Im 2 We don t assume that the incumbent and the other potential active monitor/investors (called entrants ) make simultaneous offers It is complex The equilibrium of this bidding game is in general in mixed strategies V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 29 / 34

Asymmetric information New active monitors (entrants) offer investment contributions The incumbent active monitor then makes his offer He either matches the entrants top offer (plus an arbitrary small amount) and then remains the firm s monitor Or does not match it and is replaced The residual date-2 investment, I I 2 m, where I 2 m is the highest bid, is then contributed by non-active monitors (called uninformed ) investors V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 30 / 34

The winner s curse In the bidding game, the entrants optimally bid as if the probability of success were always the lowest possible one Im 2 = p h Rm 2 b c = p h p c Our timing assumption takes the adverse-selection problem to its extreme and maximizes the incumbency rent The uninformed investors contribute to the investment shortfall: I 2 u = I I 2 m The stake R 2 u is such that uninformed investors break even: (p h + ατ)r 2 u = I 2 u It is easy to verify that the entrepreneur s stake exceeds b/ p V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 31 / 34

Monitor s rent The entrepreneur s date-2 net payoff is then Π 2 b = (p h + ατ)(r R 2 u R 2 m) = [(p h + ατ)r I c] ατ ( ) c p The entrepreneur s net expected payoff is the equal to the expected NPV minus the incumbent monitor s expected rent ( ) c R 2 m = ατ p V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 32 / 34

Date-1 competition At date t = 1 potential active monitors are symmetrically informed We assume that the entrepreneur chooses the minimal incentive compatible reward Rm 1 Rm 1 = c p There is perfect competition of monitors for the block share R 1 m Monitors perfectly anticipate the future incumbency rent They are willing to make a generous offer by contributing up to ( ) c Im 1 = p h + βr 2 m p V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 33 / 34

Informational advantage and switching cost One can view the information advantage of the incumbent monitor as a switching cost that tends to lock the firm in with the monitor The anticipated ex post market power enjoyed by the incumbent is competed away at the ex ante stage through a short-term loss-making offer The blockholding is initially acquired at a premium, and is later maintained at a discount V. F. Martins-da-Rocha (UC Davis) Corporate Finance Spring quarter, 2011 34 / 34