The Reversal Rate. Effective Lower Bound on Monetary Policy. Markus K. Brunnermeier & Yann Koby. Princeton University. Brunnermeier & Koby

Transcription

1 The Reversal Rate Effective Lower Bound on Monetary Policy Markus K. Brunnermeier & Yann Koby Princeton University BIS Research Network Meeting Basel, March 14 th, 2016

2 Motivating Questions New Keynesian models: ZLB = Liquidity trap Is zero special? Are negative rates special? Ignoring headline risk No Lower bound or Reversal Rate Rate at which accommodative policy becomes contractionary (possibly due to financial instability) Does strict financial regulation reduce effectiveness or reverse MoPo? What factors determines the Reversal Rate? Market structure Banks equity Interaction with prudential regulation Interaction with QE

3 Motivation Interest rate cut Substitution effect: safe asset risky loans Wealth effect: negative rate = tax Not in representative agent analysis

4 Motivation Interest rate cut Substitution effect: safe asset risky loans Wealth effect: negative rate = tax

5 Banks balance sheet A L Reserves C f Bonds B B Deposits D D Loans L L Net worth E 0 Two-sided market Output: loans, reserves Input: deposits

6 Model Loan market L r L = 0 1 l i r L di L r L = L(r L )/I Deposit market 1 D r D ; r f = 0 d i r D ; r f di D r L ; r f = D(r L ; r f )/I d i r d ; r f = argmax U(W, L c, d ) Liquidity service Bank competition I banks Bertrand competition but house bank advantage

7 Model Loan market L r L = 0 1 l i r L di L r L = L(r L )/I Deposit market 1 D r D ; r f = 0 d i r D ; r f di D r L ; r f = D(r L ; r f )/I d i r d ; r f = argmax U(W, L c, d ) Bank competition I banks Bertrand competition but house bank advantage Liquidity service r f + κ L r f r f κ D Mark-up μ L Mark-down μ D

8 Roadmap Impact on profit/equity Perfect competition policy rate cut perfect pass through House bank driven markups perfect pass through quantity adjustment Local monopolist/monopsonist Impact on lending/credit growth mark-up depends on semi-elasticities

9 Roadmap Impact on profit/equity Perfect competition policy rate cut perfect pass through House bank driven markups perfect pass through quantity adjustment Local monopolist/monopsonist Impact on lending/credit growth Determinants of Reversal Rate Interaction with financial regulation Interaction with QE optimal sequencing mark-up depends on semi-elasticities

10 Roadmap Impact on profit/equity Perfect competition policy rate cut perfect pass through House bank driven markups perfect pass through quantity adjustment Local monopolist/monopsonist mark-up depends on semi-elasticities ε L r L ε D r D, r f ε D,rf log L r L log D r D log D r D ;r f r f

11 Perfect competition A pass through L Reserves C f Bonds B B Deposits D D Loans L L Net worth E 0 r f = r L = r D perfect pass through 1. Profits from ongoing business/interest rate margins = 0 2. Re-evaluation gains Bdr f Funding of bonds B that yield r B is now lower by dr D Interest rate cut = stealth recapitalization

12 κ-mark-ups A pass through L Reserves C f Bonds B B Deposits D D Loans L L Net worth E 0 r L = r f + κ L r D = r f κ D 1. Profits from ongoing business change since loan quantity and deposits adjust 2. Re-evaluation gains Bdr f

13 Monopoly & general case Loan problem is separate from deposit problem Why? Reserve holdings is in between Loan rate after mark-up μ L r L = r f + μ L (r L ), μ L r L min{κ L, Deposit rate after mark-down μ D r D = r f + μ D (r D, r f ), μ D r D, r f min{κ D, 1 ε L (r L ) } 1 ε D (r } D,rf ) where κ L, κ D are new relationship costs outside of house bank κ L, κ D = 0 κ L, κ D = perfect competition segmented markets & monopolies Profit has four parts: Π 1 r f = μ L r L L + μ D r D, r f D + r B r f B π E E 0 Implicit assumption: Price stickiness

14 Impact on PROFIT unconstrained case Proposition (general case): dπ 1 dr f = ε D ε D,rf μ D ε L r f μ L L Net interest margin business ณB reevaluation Perfect competition = B κ mark-ups (set ε D,rf = 0) D = κ D 1/ε κ L L D 1/ε B L Local monopoly (set ε D,rf = 0) = D L B = C E 0

15 Impact on PROFIT constrained case Economic or regulatory constraint γ(l r L + φb) E 0 + Π 1 =E 1 If constraint binds: interest rate cut can t lead to a substitution from C to L Loan mark-up even larger than in monopoly case Ongoing business vs. re-evaluation effect Deposit margin is not affected Since constraint only binds L & loan and deposit decisions separable

16 Impact on PROFIT constrained case Amplification/spiral r f Π 1 L, R dπ 1 dr f = γ γ λ C E 0 ε D,r f ε D D where λ = r L o r L = L 1 ( E 0+Π 1 γ φb) r L

17 Impact on LENDING Constraint γ L r L + φb E + Π 1 dl dr f = 1 γ dπ 1 dr f Sum up: Interest rate cut can lead to more or less lending (depending how large B is) Need data on banks interest rate sensitivity (Sraer et al. 2015, Piazzesi et al. 2015)

18 Numerical example Constant ε L, ε D = 1 α+βr D, κ L = κ D =, for different B

19 QE: Optimal sequencing 1. Induce banks to hold more long-run assets B 2. Interest rate cut stealth recapitalization 3. QE: banks sell now highly priced long-run assets to CB 4. Further interest rate cut is less effective/contractionary Reloading strategy 1. if banks suffer losses (e.g. delinquencies) & RR rises > r f 2. Raise policy rate (to increase banks interest margin) 3. Reverse QE or another LTRO

20 Dec-97 Dec-98 Dec-99 Dec-00 Dec-01 Dec-02 Dec-03 Dec-04 Dec-05 Dec-06 Dec-07 Dec-08 Dec-09 Dec-10 Dec-11 Dec-12 Dec-13 Dec-14 Dec-15 Sep-01 Sep-02 Sep-03 Sep-04 Sep-05 Sep-06 Sep-07 Sep-08 Sep-09 Sep-10 Sep-11 Sep-12 Sep-13 Sep-14 Sep-15 Interaction with QE and VLRTO Re-evaluation effect depends on B QE lowers (aggregate) B and increases R 100% 90% 80% 70% 60% 50% Resident banks Nonresidents ITALY VLTRO QE 100% 90% 80% 70% 60% 50% IRELAND Resident banks Non-residents QE 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% One bullet reload with interest rate rise + 2 nd QE + cut

21 Literature Theory Oligopoly: Business margin: Monti-Klein model (B = 0) Competitive: Re-evaluation: BruSan I theory of money Interest rate sensitivity of banks Stock price: Flannery & James (1984), Begenau et al. (2015) Lending: Landier et al. (2015) Deposits: Drechsler et al. (2015), Deposit rate pass through Competition: Maudos & de Guevarra(2005) Delay: DeBondt (2005)

22 Conclusion Zero/negative interest rates are not special! Interest rate cut Substitution effect: safe asset risky loans Wealth effect: tax + prudential regulation

23 Conclusion Zero/negative interest rates are not special! Interest rate cut Substitution effect: safe asset risky loans Wealth effect: tax + prudential regulation Reverses substitution effect + amplification

24 Conclusion Zero/negative interest rates are not special! Interest rate cut Substitution effect: safe asset risky loans Wealth effect: tax + prudential regulation Reverses substitution effect + amplification What determines the Reversal Rate? Market structure and pass through of rates Interaction with prudential regulation Banks equity capitalization countercyclical regulation Duration risk of banks (long-dated assets) Interaction with QE (correct sequencing)