Price stability and financial stability: has there been a link? The case of the US & Eurozone

Transcription

1 Price stability and financial stability: has there been a link? The case of the US & Eurozone Christophe Blot Jérôme Creel Paul Hubert Francesco Saraceno

2 Motivation Conventional wisdom The belief that price stability produces financial stability was seemingly wrong! However, it has been at the core of monetary policy strategy since the mid-1980s The rise of the conventional wisdom raises several questions: Why did it develop? A matter of theory and beliefs, without strong empirical support Can we evidence past/present similarities/differences btw the US and Eurozone? What might have come true for some have not for the others btw the various index of financial stability and price stability? Since there are different types of index of financial stability, are there some that are more related to price stability than others? We extend the literature in two respects: New empirical results about the conventional wisdom Introduction of price stability in the literature about the correlations of (international) financial markets

3 Literature (1) Only a few contributions have led to a conventional wisdom (or Schwartz hypothesis) The conventional wisdom on the links btw monetary and financial stability is nicely summarised by Bordo et al. (2001) who write that a monetary regime that produces aggregate price stability will, as a by-product, tend to promote stability of the financial system (Borio & Lowe, 2002, p. 27)! See also Issing (2003) The literature attributes the motherhood to A.J. Schwartz (1988, 1995) Her 1995 title is explicit: Why financial stability depends on price stability A quote: price level stability is essential for financial stability On the micro side, A.J. Schwartz relates price instability to inflation distortion, growing uncertainty, shortened investment horizons, and governments nominal gains; they produce financial instability On the macro side, she discusses the impact of price instability on the value of collateral and on financial risk (mainly in the banking sector: the fact remains that price instability undermines sound banking. It contributes to financial risk, p.39) The paper by A.J. Schwartz goes beyond the debt-deflation à la Fisher (1933) as she relates the end of price (hence financial) instability to sound monetary policy

4 Literature (2) Empirical evidence? To our knowledge, only one paper specifically dedicated to this issue: Bordo et al. (2001): they conclude that unanticipated movements in the price level and inflation rate have contributed historically to financial instability in the US, ever more so btw 1870 & 1933, or in the 1980s and 1990s

5 Before the GFC, the conventional wisdom has come to be criticized e.g. by Borio & Lowe (2002), White (2006), Leijonhufvud (2007): price stability is not enough Price stability might not be sufficient to ensure macroeconomic stability (e.g. low inflation might foster risk) Major economic and financial crises were not preceded by inflationary pressures Inflation is not a good indicator to predict a crisis Cf. IMF (2009) [next slide]

6

7 The conventional wisdom and the critics Critics urged discussing about the introduction of financial stability as an objective of the central bank Many discussions about the Tinbergen principle N instruments to achieve N objectives Literature is abundant (see Disyatat, 2010) but it does not seriously challenge the conventional wisdom : e.g. Blanchard et al. (2010): no change required in the policy reaction function, except a better cooperation with supervisory body e.g. Woodford (2012): only a marginal change required (application of a flexible inflation targeting strategy with enlarged tasks of the CB)

8 Literature on the dynamic correlations of financial markets Yang et al. (2009): bivariate smooth transition Garch, stock-bond correlations in the US and the UK, , dependent on macro conditions (including inflation environment), Cai et al. (2009): DSTCC-CARR, international stock correlations btw 6 countries, , dependent on inflation fluctuation and market volatility Antonakakis (2012): DCC Garch, co-movements of sovereign bond yield spreads in 9 Eurozone countries, , and the role of credit ratings agencies Silvennoinen & Thorp (2013): DSTCC-Garch, correlations btw stocks, bonds and commodity futures, Creti et al. (2013): DCC Garch, correlations of price returns for 25 commodities (including energy raw materials) and stocks (S&P 500),

9 Use of different techniques Simple correlation IRF from VAR specification To identify the possibly time-varying relationship btw price and financial stability, we estimate a time-varying measure of correlations based on the DCC model of Engle (2002), in which the conditional correlation follows a GARCH(1,1) process By construction, the GARCH model is a specification of both the conditional mean and the conditional variance, where the variance is a function of prior unanticipated innovations ε t 2 and prior conditional variances σ t2. A GARCH(1,1) specification with only a constant corresponds to the modelling of an inflation target with an error term A DCC-GARCH model can be seen as a multivariate representation of univariate GARCH process from which dynamic covariances are computed from conditional variances. The procedure involves 2 steps: estimating the conditional volatility of each individual series and then capturing dynamics in the covariance of the standardized residuals from the first stage used as inputs to estimate a time-varying correlation matrix Possible extensions: introducing a monetary aggregate as an explanatory variable of both (price and financial stability) processes (as in Bordo et al., 2001)

10 Results: simple correlation Samples: US ( ) EZ ( ), monthly data CPI: consumer price index (in difference); FSI: Financial stability index (US: St Louis Fed Financial Stress Index; EZ: Composite Indicator of Systemic Stress) Correlation us_cpi us_fsi ez_cpi ez_fsi us_cpi 1 ez_cpi 1 us_fsi -0,31 1 ez_fsi -0,05 1 N 229 N 168 NB: in the US, strong negative correlation robust to the choice of similar sample as EZ

11 Results: VAR IRF US: asymmetric reactions to shocks EZ: no reaction to shocks

12 Results: DCC (time-varying correlations), CPI-FSI US EZ

13 Results: DCC (time-varying correlations), CPI-Stocks Mkt US EZ

14 Results: DCC (time-varying correlations), CPI-Loan Mkt US EZ

15 Results: DCC (time-varying correlations), CPI-Housing Mkt US EZ

16 Results: DCC, covariance DCC with CPI Variable Obs Coef SE p-value dcc_us_fsi 229-0,27 0,16 0,10 dcc_us_hous 229 0,10 0,13 0,44 dcc_us_stock 229-0,17 0,07 0,02 dcc_us_loan 229-0,21 0,15 0,18 dcc_ez_fsi 168 0,18 0,14 0,22 dcc_ez_hous 168 0,18 0,18 0,34 dcc_ez_stock 168-0,29 0,12 0,02 dcc_ez_loan 168 0,41 0,18 0,03

17 Conclusions The financial stability index matters! Highly cyclical correlations btw the PSI & the FSI (or the loan market index) More frequent negative correlations btw the PSI & stock markets index More frequent positive correlations btw the PSI & the housing market index Significant covariance PSI-Stock Mkt (<0, US & EZ) & PSI-Housing Mkt (>0, EZ) No strong difference btw US & EZ

18 Further results: DCC determinants, US Determinants of DCC in the US dcc_us_fsi_cpi dcc_us_hous_cpi dcc_us_stock_cpi dcc_us_loan_cpi us_fsi 0.20*** [0.05] us_hous 0.01*** [0.00] us_stock -0.00* [0.00] us_loan -0,03 [0.02] us_cpi 0.16*** -0.04* *** [0.05] [0.02] [0.01] [0.05] us_cbrate 0.11*** 0.07*** 0 0 [0.03] [0.01] [0.00] [0.05] us_m 0.02* -0.01*** ** [0.01] [0.00] [0.00] [0.01] us_indpro 0.03*** -0.02*** 0.00*** -0,01 [0.01] [0.00] [0.00] [0.01] us_bonds -0, ** 0-0,08 [0.05] [0.02] [0.01] [0.07] oil * * [0.00] [0.00] [0.00] [0.00] crisis 0, * 0.04** -0.83*** [0.17] [0.07] [0.02] [0.20] _cons -0.96*** 0, *** 1.24*** [0.32] [0.13] [0.04] [0.42] N r2 0,24 0,21 0,06 0,13 Robust standard errors in brackets. * p < 0.10, ** p < 0.05, *** p < High policy rates increase corr. btw PSI & FSI (US monetary policy, achieving price stability, would also achieve financial stability) High liquidity decreases corr. btw PSI & housing or loan markets (under price stability, more liquidity would foster financial instability) Unconclusive implications of prod. or crisis on corr.

19 Further results: DCC determinants, EZ Determinants of DCC in the EZ dcc_ez_fsi_cpi dcc_ez_hous_cpi dcc_ez_stock_cpi dcc_ez_loan_cpi ez_fsi 1.18** [0.46] ez_hous 0.19*** [0.04] ez_stock 0.01** [0.00] ez_loan -0,02 [0.02] ez_cpi 0, *** 0, *** [0.07] [0.07] [0.08] [0.05] ez_cbrate 0, ** -0,12 0,05 [0.08] [0.10] [0.10] [0.09] ez_m 0, *** ** [0.03] [0.03] [0.03] [0.03] ez_indpro 0.02* -0.04** -0.04*** -0.02** [0.01] [0.01] [0.01] [0.01] ez_bonds 0,04 0,07-0, * [0.09] [0.11] [0.10] [0.09] oil 0.00* 0.01*** -0.00*** 0.00*** [0.00] [0.00] [0.00] [0.00] crisis -0, *** -0,27-0,3 [0.26] [0.36] [0.16] [0.18] _cons -0.66* -0,18 0, * [0.40] [0.51] [0.41] [0.37] N r2 0,23 0,41 0,22 0,32 Robust standard errors in brackets. * p < 0.10, ** p < 0.05, *** p < No clear-cut conclusion about the cons. of policy rates on corr. (EZ monetary policy, achieving price stability, does not seem to foster achievement of financial stability) No clear-cut conclusion about the cons. of liquidity on corr. (under price stability, more liquidity would foster (deter) loan mkt (housing mkt) instability) The rise of prod. reduces corr. (under price stability, high activity would foster financial instability)

20 Policy implications In the US, extending the objectives of the Fed to target financial stability is not required, but we evidence possible financial risks from the QE policy In the EZ, the ECB targeting financial stability is required; it would mitigate the use of policy rates; we also evidence the impact of real activity on financial (in)stability: there are risks that recovery generates financial instability