Corporate Finance Theory – Asymmetric Information & Signaling (Comprehensive Study Notes)

The Lemon Market & Adverse Selection

• Classic contribution: George A. Akerlof, 1970
  • "Lemon market" shows how asymmetric information (AI) between sellers and buyers produces adverse selection (AS).
  • Intuition:
  • Consumers cannot distinguish high- and low-quality goods.
  • They therefore offer only the average price.
  • Average price attractive to sellers of low quality → proportion of lemons on the market rises.
  • Rational consumers anticipate this → further reduce willingness to pay.
  • Downward spiral leads to market breakdown (no trade).
• Key logical chain (negative spiral):
  1. AI → buyers pay mean price.
  2. Mean price → good-quality sellers exit.
  3. Quality of goods offered falls.
  4. Buyers adjust beliefs → lower price again.
  5. Repeat until price → $0$ & market collapses.
• Broader relevance: insurance, labor, credit, IPOs–wherever one side knows more than the other.

Corporate-Finance Version of the Lemon Problem

• Replace "cars" with firms.
• Investors cannot observe firm value ex-ante when purchasing shares.
  • Offer only average price for equity.
  • Owners of bad firms find that price attractive; owners of good firms do not.
• Consequence:
  • Equity that actually reaches the market is below average quality.
  • Rational investors discount price further; potential for total market shutdown identical to car market.

Uniform-Distribution Thought Experiment

• Population of firms: future cash flows $X \sim U(0,100)$.
• Risk-neutral investors cannot observe $X$ → pay at most $E[X]=50$.
• Firms with X>50 refuse to sell; only $X\le 50$ remain.
  • Conditional distribution becomes $U(0,50)$ with mean $25$.
• Investors refine price to $25$ → trigger next round.
• Iteration converges to equilibrium price $0$ → only worst firms trade → market crashes.

Myers & Majluf (1984): Asymmetric Information & Pecking Order

• Observation: Equity issuance may signal that managers want to sell over-valued shares (a lemon).
• Equity becomes under-priced in equilibrium.
  • Severe underpricing → firm may reject positive-NPV project.
• Resulting pecking-order hierarchy:
  1. Internal funds (retained earnings/slack $S$).
  2. Debt (less informationally sensitive).
  3. External equity (most sensitive).

Basic Myers–Majluf Model (Set-up)

• Existing asset (value random variable a>0, mean $\bar A$).
• Investment opportunity b>0, mean $\bar B$.
• Needs funds $I$; has cash $S$ → external equity needed $E = I-S$.
• Managers know $(a,b)$ at $t=0$; markets do not.
• Objective: maximize old (incumbent) shareholders' value.

Valuations With & Without Equity Issue

• If equity is issued at price $P<em>e$ (determined by market beliefs): $V^{old}</em>1 = \frac{P<em>e}{P</em>e+E}\,(E+S+a+b)$
• If no issue/investment:
  $V^{old}_2 = S + a$
• Equity issued when:
  $V^{old}<em>1 \ge V^{old}</em>2 \;\Longrightarrow\; E+b \ge \frac{E}{P_e}(S+a)$
• Market‐clearing price (rational expectations):
  $P<em>e = S + E\big[ A\mid E+b \ge \tfrac{E}{P</em>e}(S+a) \big] + E\big[ B\mid E+b \ge \tfrac{E}{P_e}(S+a) \big]$

Illustration 1: Under-investment Equilibrium

• Two equally likely states:
  • State 1: $(a,b)=(150,20)$
  • State 2: $(a,b)=(50,10)$
• Parameters: $I=100,\;S=0\Rightarrow E=100$.
• If firm always issues equity & invests: $P_e=115$.
  • State-wise firm value: $V<em>1=270,\;V</em>2=160$.
  • Old shareholders get (using above formula):
  • State 1: $144.42$ vs 150 if they do nothing.
  • State 2: $85.58$ vs 50 if do nothing.
• Optimal policy for incumbents: issue only in bad state.
  • Market understands; price falls to $P_e=60$.
  • Payoffs then: State 1 do nothing = 150; State 2 issue = 60.
  • Expected incumbent payoff =\tfrac{150+60}{2}=105<115 ⇒ society loses 10 value.
• Lesson: AI causes firm to abandon good projects → Deadweight loss.
• Internal cash ($S=100$) would avoid mispricing → expected value back to $115$.

Illustration 2: Project "Too Good to Miss"

• Update b in good state: $b=100$ (rest same).
  • Means $\bar B =55$.
• Market price if always issue: $P_e=155$.
• Payoffs to old shareholders (issue vs not):
  • State 1: $212.75$ vs 150 (gain).
  • State 2: $97.25$ vs 50.
• Because gains dominate, firm issues in both states → equilibrium.
• Moral: very high-NPV projects survive AI cost; marginal ones do not.

General Properties of the Myers–Majluf Framework

• If AI affects only investment opportunity b (asset-in-place value a observable):
  • Price satisfies $P_e \ge S+a$ → issuance always value-adding → no loss.
  • Firm can separate claims: sell asset-in-place, spin-offs, carve-outs to eliminate AI.
• If no project exists ($b=0$):
  • Equity never issued unless $a$ at known minimum $a_{min}$.
  • Market price forced to $P<em>e = a</em>{min}+S$ (Akerlof result) → only lemons issue.

Debt vs Equity under Asymmetric Information

Introducing Debt Choice (t=0 chosen)

• Finance $I-S$ via debt $D$ or equity $E$.
• Define capital gains at $t=2$ when truth revealed:
  $\Delta E = E<em>1 - E,\; \Delta D = D</em>1 - D$.
• Issue & invest if additional payoff positive:
  • Equity: invest when $b \ge \Delta E$.
  • Debt: invest when $b \ge \Delta D$.
• Assume debt value less information-sensitive: |\Delta E| > |\Delta D| (Galai & Masulis, 1976).
  • More states satisfy $b \ge \Delta D$ than $b \ge \Delta E$.
  • ⇒ Ex-ante firm value higher under debt → mitigates under-investment.

Financing Choice After Private Information Known (t=1)

• Manager decides between debt/equity after observing $(a,b)$.
• Equity issue signals: b-\Delta E > b-\Delta D \Rightarrow \Delta E < \Delta D.
• In any rational‐expectations equilibrium $\Delta E = \Delta D = 0$, but equity would be chosen only if \Delta E < 0 → contradiction.
• Therefore no equilibrium with equity issuance; firm always prefers debt (or internal funds).

Pecking-Order Restated

1. Internal capital (no mispricing risk).
2. Debt (low sensitivity to AI).
3. External equity (high sensitivity).

Can Good Firms Signal Their Quality?

Leland & Pyle (1977) Ownership-Retention Signal

• Good owners credibly signal by keeping a large equity stake (α).
• Market reasoning:
  • Entrepreneurs are risk-averse.
  • Holding risky equity costly unless expected cash flow high.
  • High retention therefore implies high mean cash flow.
• Empirical prediction: managerial ownership + firm quality positively related.

Model Structure

• Period 0: Entrepreneur sells fraction $1-α$ of firm.
• Period 1: Market observes sold fraction & updates belief.
• Period 2: Cash flow $X \sim N(μ,σ^2)$ where $μ ∈{X<em>L, X</em>H},\; X<em>H>X</em>L$.
• Entrepreneur utility (CARA):
  $U(W<em>t) = E[W</em>t] - \frac{b}{2}Var(W<em>t),\; b>0$ where $W</em>t = αX_t + (1-α)V,\; t∈{L,H}$.
• Variance term: $Var(W_t)=α^2σ^2$.

Full-Information Benchmark

• Market knows type → $V = X_t$.
• Utility: $U = X_t - \tfrac{b}{2}α^2σ^2$.
• Risk-averse entrepreneur sets $α=0$ (sell 100 %).

Asymmetric Information & Separating Equilibrium

• Market initial beliefs: $Pr(H)=1-q,\;Pr(L)=q$.
• Goal: Design $α_H$ so L-type will not mimic.
• Incentive constraints:
  • L-type must prefer his own strategy (α=0):
    $U<em>L(0,X</em>L) \ge U<em>L(α,X</em>H)$ →
    $αX<em>L + (1-α)X</em>H - \tfrac{b}{2}α^2σ^2 \le X_L$.
  • Rearrange → quadratic inequality:
    $\frac{b}{2}σ^2α^2 + ΔX α - ΔX ≥ 0\,,\; ΔX ≡ X<em>H - X</em>L$.
  • Minimal α satisfying:
    $α_1 = \frac{\sqrt{ΔX^2 + 2bσ^2ΔX} - ΔX}{bσ^2}$.
• H-type willing if utility ≥ what he'd get by pooling (sell all):
  $X<em>H - \tfrac{b}{2}α^2σ^2 ≥ X</em>L$ ⇒
  0<α ≤ α_2 = \sqrt{\frac{2ΔX}{bσ^2}}.
• Because $α<em>2 > α</em>1$ (proved via (\sqrt X+\sqrt Y)^2-(\sqrt{X+Y})^2 = 2\sqrt{XY}>0), intersection non-empty.
• Optimal for H-type: choose minimal retention that separates ⇒ $α^*<em>H = α</em>1$.
• Equilibrium strategies:
  $α^<em><em>L = 0,\; α^</em></em>H = α_1.$

Comparative Statics

• $α_1 \uparrow$ with $ΔX \uparrow$ (more AI → need stronger signal).
• $α_1 \downarrow$ with $σ^2 \uparrow$ or $b \uparrow$ (riskier cash flow or more risk-averse entrepreneur lowers retention needed because costlier to fake).
• Industry implications:
  • High-tech / growth (high AI, high volatility) → larger managerial stakes.
  • Mature/transparent industries → lower required stakes.
• Event study: manager share sales perceived as negative signal (price drop).

Ethical, Philosophical & Practical Takeaways

• Market failures from AI destroy social surplus even when profitable trades exist.
• Signaling & financial structure choices are second-best remedies; they consume resources (e.g., risk bearing, debt overhang).
• Managerial incentives & corporate transparency can mitigate, but cannot fully eliminate, deadweight losses.
• Regulators may mandate disclosure, auditing, or certification to reduce AI, but must weigh costs.

Connections & Applications

• Links to previous topics: efficient markets, CAPM, capital-structure irrelevance (MM) break down under AI.
• Real-world:
  • IPO underpricing & lock-up agreements (insider retention).
  • Credit rating agencies (attempt to bridge AI for debt).
  • ICO/crypto markets often extreme AI; founders’ token vesting imitates $α_H$.
  • Used-car warranties and CPO programs are non-financial analogues.

Key Equations (Cheat-Sheet)

• Expected cash flow uniform example: $E[X]=\frac{0+100}{2}=50$.
• Equity issue payoff to incumbents:
  $V^{old}<em>1 = \frac{P</em>e}{P_e+E}(E+S+a+b).$
• Myers–Majluf issuance cutoff: $E+b \ge \frac{E}{P_e}(S+a).$
• Signal retention bounds:
  $α<em>1 = \frac{\sqrt{ΔX^2 + 2bσ^2ΔX} - ΔX}{bσ^2},\quad α</em>2 = \sqrt{\frac{2ΔX}{bσ^2}}.$
• Separating-equilibrium retention: $α<em>H^<em>=α</em>1,\; α_L^</em>=0.$

Summary Bullet List

• Asymmetric information produces adverse selection, potential market collapse.
• In corporate finance, AI drives a pecking order: internal funds → debt → equity.
• Equity underpricing can cause under-investment; extremely high-NPV projects survive.
• Debt less information-sensitive than equity; ex-ante value higher with debt financing.
• Signaling via insider ownership (Leland & Pyle) allows good firms to separate.
• Minimum retention increases with information gap and decreases with volatility & risk aversion.
• Empirical patterns: larger managerial stakes in high-AI industries; share sales interpreted negatively.