Credit Market Freezes V. V. Chari & Keyvan Eslami University of - PowerPoint PPT Presentation

Discussion of Benmelech & Bergman’s Credit Market Freezes V. V. Chari & Keyvan Eslami University of Minnesota & Federal Reserve Bank of Minneapolis April 2017

What This Paper Is About Debt issuance fluctuates a lot High for some period of time Abruptly collapses Collapses associated with rise in “illiquidity” Chari & Eslami Credit Market Freezes

Empirical Contributions Impressive Analyze 1873 and 2008–09 crises “Illiquidity” rises a lot for bonds where price fell by a lot Opinion dispersion plays a smaller role Chari & Eslami Credit Market Freezes

What Paper Does Not Do Does not offer a theory of sources of shocks Does not offer a theory of propagation mechanisms No policy implications Chari & Eslami Credit Market Freezes

What Paper Does Not Do Does not offer a theory of sources of shocks Does not offer a theory of propagation mechanisms No policy implications In short, no theory Chari & Eslami Credit Market Freezes

My Discussion Offer a theory Theory consistent with data Offer ideas for sources of shocks Has policy implications See no role for “information sensitivity” or “endogenous adverse selection” as in Dang, Gorton & Holmstr¨ om (2015) Main challenge is to find linkage to macro variables Outline nature of challenges Chari & Eslami Credit Market Freezes

Data on Sudden Collapses

New Issuances of ABSs in 2000s $Bln $Bln 350 350 Other Non-U.S. Residential Mortgages* 300 300 Student Loans Credit Cards 250 250 Autos Commercial Real Estate 200 200 Subprime Home Equity 150 150 100 100 50 50 0 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 *No reliable data for Non-US RMBS after Q3 '08 Source: Morganmarkets, JP Morgan Chase Chari & Eslami Credit Market Freezes

Change in Stock of Real Estate Bonds in 1920s $Bln $Bln 200 200 180 180 160 160 140 140 120 120 100 100 80 80 60 60 40 40 20 20 0 0 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 Note: Data is annual change in real estate bonds divided by Nominal GDP at relevant year multiplied by Nominal GDP 2009. Source: Carter, et. al., Historical Statistics, (2006)Series Dc904 Chari & Eslami Credit Market Freezes

Theory: Reputation, Adverse Selection and Sudden Collapses

Chari, Shourideh & Zetlin-Jones (2014, AER) Large number of buyers Large number of loan originators, or banks Banks endowed with a portfolio of risky loans, size 1 Loan pays ¯ v with probability π and ¯ v with probability 1 − π Let v := ¯ v − ¯ v be spread , ¯ v is collateral value Two types of banks, π ∈ { ¯ π, ¯ π } , ¯ π < ¯ π Chari & Eslami Credit Market Freezes

C-S-ZJ Buyers have comparative advantage in holding loans c > 0 Each bank chooses how much of its loan portfolio to sell, x Let t denote payment bank receives for selling x loans, p is price per loan Bank π payoff from selling x loans for payment t (up to a constant): t + ( 1 − x ) ( π v − c ) Buyer profits from ( x , t ) (up to a constant): x π v − t Chari & Eslami Credit Market Freezes

C-S-ZJ Adverse selection: bank knows type of loans, potential buyers do not Buyers believe given bank is high-quality with probability µ Call µ the reputation of the bank Focus on sales of individual bank with reputation µ Focus on 2 buyers (Bertand-style price competition) Chari & Eslami Credit Market Freezes

Timing in Static Model Buyers simultaneously propose contracts to the bank: z = ( x h , t h , x l , t l ) ∈ Z Bank chooses whether to accept a contract or reject both If bank accepts a contract, then chooses which offer to accept Focus on incentive compatible z Restrict to pure strategies for banks, possibly mixed strategies for buyers, F ( z ) for z ∈ Z Equilibrium is standard Chari & Eslami Credit Market Freezes

Equilibrium Characterization in Static Model Proposition 1 The static model has a (unique) separating equilibrium. When reputation is low, the equilibrium has no cross-subsidization. When reputation is high, the equilibrium has cross-subsidization Low reputation → least-cost separating outcome High reputation → mixed strategies by buyers Chari & Eslami Credit Market Freezes

Equilibrium Characterization in Static Model Three general properties: for all z ∈ supp ( F ) x l = 1 Buyers make zero profits Incentive constraint for low-quality bank holds with equality: t l = t h + ( 1 − x h ) (¯ π v − c ) For reputation below a threshold, ˜ µ , least cost separating outcome has t l = ¯ π v and t h = x h ¯ π v Chari & Eslami Credit Market Freezes

Equilibrium Volume of Trade When reputation is low, expected volume for a given bank is µ T = 1 + d + 1 − µ d =(¯ π − ¯ π ) v , adverse selection discount c When reputation is high, expected volume is � � − 1 � c 2 1 − 1 − µ � c T = µ π ) v + π ) v ] 2 (¯ π − ¯ [(¯ π − ¯ µ In both cases, if v = (¯ π − ¯ π ) increases, then T falls That is, if adverse selection worsens, trade falls Problem: No reason for “discontinuous” fall in volume Chari & Eslami Credit Market Freezes

C-S-ZJ Dynamic extension Bank’s type permanent Investors learn over time Collateral values fluctuate over time Can get pooling equilibrium New banks enter with exogenous initial reputation Banks exit with given probability Chari & Eslami Credit Market Freezes

Dynamic Extension Chari & Eslami Credit Market Freezes

Dynamic Simulation Suppose v is such at µ h both types selling Collateral value fall, increases spread v Moves all banks at µ h from selling to holding Results in fall in trading volume Discontinuous because of mass point at µ h Chari & Eslami Credit Market Freezes

Dynamic Simulation Chari & Eslami Credit Market Freezes

Dynamic Simulation 0.7 10 0.65 8 Collateral Value Collateral Value 0.6 6 Volume 0.55 4 Volume 0.5 2 0.45 0 0 0 5 5 10 10 15 15 20 20 25 25 30 30 35 35 40 40 45 45 50 50 Period Chari & Eslami Credit Market Freezes

Punchline of C-S-ZJ Adverse selection model with reputation can produce large changes in volume associated with small changes in fundamentals Equilibrium is efficient (for large range of parameter values) No role for policy Chari & Eslami Credit Market Freezes

Punchline of C-S-ZJ Why did collateral values fall? Housing prices start leveling off in 2006 TFP growth slows starting in 2005 Demographics: In 2006–07, aging and retiring of baby boomers creates anticipated slowdown of GDP growth Chari & Eslami Credit Market Freezes

Bid-Ask Spreads in C-S-ZJ Consider following decentralization All sellers (banks) ask price ¯ π ¯ v − ( 1 − ¯ π )¯ v In pooling (both types selling) region p ( µ ) = µ [¯ π ¯ v + ( 1 − ¯ π )¯ v ] + ( 1 − µ ) [¯ π ¯ v + ( 1 − ¯ π )¯ v ] Spread is ( 1 − µ ) (¯ π − ¯ π ) (¯ v − ¯ v ) Fall in ¯ v leads to only low quality banks selling p ( µ ) = ¯ π ¯ v + ( 1 − ¯ π )¯ v Spread becomes (¯ π − ¯ π ) (¯ v − ¯ v ) Rise in spread with fall in trade volume Chari & Eslami Credit Market Freezes

Extending C-S-ZJ to Allow for Real Effects Suppose bank incurs cost c ∼ F ( c ) to originate loan Will originate loan as long as profits ≥ c ∗ , where c ∗ is cut-off Fall in collateral values reduces loans originated Note, fall still occurs even if secondary markets absent No obvious “multiplier” effect from presence of secondary markets Chari & Eslami Credit Market Freezes

Risk versus Illiquidity B&B use cov (∆ p t , ∆ p t + 1 ) to measure “illiquidity” Or use idea prices are mean-reverting cov (∆ p t , ∆ p t + 1 ) can be negative if pricing pricing kernel fluctuates, and markets are perfectly liquid Price of a claim that pays off 1 in period 2: p 0 = 1, p 1 ( good ) = 1, p 1 ( bad ) = 0 . 2, and p 2 = 1 When ∆ p 1 < 0, ∆ p 2 > 0, when ∆ p 1 = 0, ∆ p 2 = 0 Chari & Eslami Credit Market Freezes

Challenges for Links Between Financial Market Distress and Real Outcomes In data, most investment is “self-financed” Financial market disturbances do not directly reduce “available funds” Financial markets play some role in reallocating funds from cash-rich firms to project-rich firms Geting high “multiplier” effect challenge See Shourideh & Zetlin-Jones for resolution Chari & Eslami Credit Market Freezes

Linkage with Macro Outcomes: Challenges

Big Issue Financial friction models working through investment channel Pipes get clogged Chari & Eslami Credit Market Freezes

Big Issue Financial friction models working through investment channel Problem: In data, flows go other way Chari & Eslami Credit Market Freezes

Does Typical Firm Use External Funds to Finance Investment? Use data from From Flow of Funds for all non-financial corporations Available Funds, AF = Revenues - Wages - Materials - Interest Payments - Taxes In Flow of Funds, AF = Internal Funds + Dividends Alternatively, AF = Retained Earnings + Dividends + Depreciation In Flow of Funds use Capital Expenditures Chari & Eslami Credit Market Freezes