Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
European repo market The model Empirical analysis Liquidity, Government Bonds and Sovereign Debt Crises Francesco Molteni European University Institute 6th International Research Conference Bangko Sentral ng Pilipinas September 20, 2016 1 / 29 European repo market The model Empirical analysis Motivations Figure: 10-year government bond yield spread 2 / 29 European repo market The model Empirical analysis Motivations Figure: 10-year government bond yield spread Credit risk I Fundamentally driven I Self-fulfilling dynamics 2 / 29 European repo market The model Empirical analysis Motivations Figure: 10-year government bond yield spread Credit risk I Fundamentally driven I Self-fulfilling dynamics Liquidity I Market liquidity I Funding liquidity 2 / 29 European repo market The model Empirical analysis Main ideas Government bonds are the prime collateral in the interbank market Before the crisis government bonds were as liquid as money circulating across banks During the crisis the rise in sovereign risk reduced the funding liquidity of peripheral government bonds 3 / 29 European repo market The model Empirical analysis Repurchase agreement Figure: Bilateral repo contract 4 / 29 European repo market The model Empirical analysis Repurchase agreement Figure: Bilateral repo contract 5 / 29 European repo market The model Empirical analysis Liquidity shock Figure: Yields of 10-year Italian (LHS) and German (RHS) govt bonds ”Italian bonds are in a perfect storm at the moment. Real money investors are running away and those investors using Italian bonds to finance will also be clearing the decks now”. FT 9th November 2011 ”LCH Clearnet SA raises margin on Italian bonds” 6 / 29 European repo market The model Empirical analysis Main contributions 1. Collect data on the European repo market and haircuts on government bonds 2. Study the systemic impact of a liquidity shock in a DSGE model 3. Assess empirically the consequence of a rise in haircuts on sovereign bond yields 7 / 29 European repo market The model Empirical analysis Related literature 1. The US repo market during the liquidity crisis of 2007-09 Adrian and Shin (2009, 2010), Brunnermeier (2009), Copeland et al. (2010), Gorton and Metrick (2012), Krishnamurthy et al. (2013) 2. Negative feedback between haircuts and the value of collateral Brunnermeier and Pedersen (2009), Ashcraft et al. (2010), Gârleanu and Pedersen (2009) 3. Link between banking and sovereign weakness Acharya et al. (2011), Gennaioli et al. (2014), Broner et al. (2014), Coimbra (2014), Bocola (2015), Fahri and Tirole (2015) 4. Liquidity frictions in a DSGE model Kiyotaki and Moore (2012), Del Negro et al. (2012), Jermann and Quadrini (2012), Ajello (2011), Benigno and Nasticò (2013) 8 / 29 European repo market The model Empirical analysis Secured and unsecured borrowing Figure: Repos and unsecured borrowing (total turnover) Figure: Shares of bilateral and tri-party repos (in percent of the total) Source: European Money Market Survey (ECB) 9 / 29 European repo market The model Empirical analysis Collateral Figure: Share of collateral in the european repo market (in percent) Source: European Repo Survey (ICMA) 10 / 29 European repo market The model Empirical analysis Yields and haircuts on 10-year govt bonds LCH Clearnet Ltd Figure: Ireland Figure: Portugal Source: Bloomberg and LCH Clearnet Ltd 11 / 29 European repo market The model Empirical analysis Section 2 The model 12 / 29 European repo market The model Empirical analysis Model environment Household The representative household consists of a continuum of members which are either workers or entrepreuners Firms Intermediaries in the production and labour market (nominal and real rigidities) Government collects taxes sets nominal interest rate unconventional policy 13 / 29 European repo market The model Empirical analysis HH structure Continuum of members j ∈ [0,1] At the beginning of each period they receive an idiosyncratic shock determining their profession: ( Entrepreuneurs with probability γ j= Workers with probability 1 − γ At the end of the period asset and consumption sharing Et P∞ s−t s=t β h Cs1−σ 1−σ − ω 1+ν R1 χ Hs (j)1+ν dj i 14 / 29 European repo market The model Empirical analysis Portfolio Table: Household’s balance sheet (financial assets) Assets Capital stock: Others’ equity: Long-term bonds: qt Kt qt NtO L QtL BPt Short-term bonds: QtS BPt Liabilities Equity issued: S Net worth: Nt = NtO qt NtI L S qt Nt + QtL BPt + QtS BPt Kt − N I | {z }t + unmortgaged capital Table: Return stream asset Nt BtL BtS t+1 rtK 1 1 t+2 rtK λ 0 t+3 rtK λ2 0 ... ... ... ... 15 / 29 European repo market The model Empirical analysis HH problem Budget constraint L Ct (j) + ptI It (j) + qt [Nt+1 (j) − It (j) − λNt ] + QtL Bt+1 (j) − BtL S +QtS Bt+1 (j) − BtS = rtk Nt + Wt (j)Ht (j) + Dt + DtI − τt Funding constraints Nt+1 (j) ≥ (1 − θ)It (j) | {z } +λNt borrowing constraint L (j) ≥ (1 − φ )B L Bt+1 t | {z t} liquidity constraint S (j) ≥ 0 Bt+1 16 / 29 European repo market The model Empirical analysis The Government Intertemporal budget constraint L S B BtL S Bt+1 + T = QtL Pt+1 − λ + Q t t P P t t t BtS Pt + BtL Pt Fiscal policy Tt − T = ψT BtL Pt − BL P Conventional monetary policy Rt = max πtψπ , 1 17 / 29 European repo market The model Empirical analysis The Government Intertemporal budget constraint L S B BtL S Bt+1 + T = QtL Pt+1 − λ + Q t t P P t t t BtS Pt + BtL Pt Fiscal policy Tt − T = ψT BtL Pt − BL P Conventional monetary policy Rt = max πtψπ , 1 Unconventional policy S Bt+1 Kt = ψB QtS = φt φ −1 1 Rt 17 / 29 European repo market The model Empirical analysis Liquidity shock Laissez-faire economy Figure: Impulse response function 18 / 29 European repo market The model Empirical analysis Liquidity shock Unconventional policy Figure: Impulse response function 19 / 29 European repo market The model Empirical analysis Liquidity shock Zero lower bound Figure: Impulse response function 20 / 29 European repo market The model Empirical analysis Section 3 Empirical analysis 21 / 29 European repo market The model Empirical analysis The model Let yt = [ht , CDSt , ydt ]0 yt = B(L)yt + t where B(L) = B0 + Pp k k=1 Bk L and t ∼ (0, Σ ) Flat priors for the coefficient matrices and variance-covariance matrix 22 / 29 European repo market The model Empirical analysis Identification of shock I Narrative approach and High Frequency Identification (HFI) I Choleski decomposition I Delay between the announcement and the implementation of changes in haircuts I Shock to haircut unanticipated (liquidity surprises) 23 / 29 European repo market The model Empirical analysis Example of Repo Clear Margin Rate Circular 24 / 29 European repo market The model Empirical analysis Impulse response function Figure: Impulse response function of government bond yields 25 / 29 European repo market The model Empirical analysis Robustness test Impulse responses by local projections (Jorda) Hansen (2000) test: threshold effect on CDS State dependent model: ydt = δL0 Xt −1 + βshock + utL 0 Xt −1 + βshock + utH ydt = δH if CDSt ≤ 600 if CDSt > 600 26 / 29 European repo market The model Empirical analysis Impulse response function Figure: linear model Figure: state-dependent model 27 / 29 European repo market The model Empirical analysis Conclusion I Liquidity channel of the European financial crisis through the European repo market I Effectiveness of unconventional monetary policy in a liquidity crisis I Lack of a liquid asset 28 / 29 European repo market The model Empirical analysis Thank you for your attention ! 29 / 29