MONETARY POLICY AND INDIVIDUAL HETEROGENEITY: THE EXPERIENCE OF THE BANCO DE ESPAÑA Óscar Arce Associate Director General Economics and Research 18-19 May 2017 Bank of England CCBS Central Bank Chief Economists ADG ECONOMICS AND RESEARCH DG ECONOMICS AND STATISTICS
This talk: Two issues - Optimal monetary policy (MP) with heterogeneous agents: A theoretical perspective How does individual heterogeneity shape the optimal response of MP in the short-run vs medium run? - The investment channel in the aftermath of the Spanish financial crisis: The key role of firms heterogeneity Squaring strong investment with strong deleveraging forces at the aggregate The transmission of the ECB s CSPP in a dual economy featuring many small firms and a few (very) large ones 2
Why should MP-makers care about agent heterogeneity and the distributional effects of their policies? - In standard NK models intertemporal substitution effects are centerpiece for understanding MP transmission - But there is growing evidence that this channel is not the full story (may not be even the main one ) The intertemporal allocation effects of changes in interest rates (IR) on consumption and investment are typically low Indirect (general equilibrium) effects of changes in IR are more relevant (Kaplan et al. 2016) Redistributive channels (heterogeneous earnings and IR exposures, Fisherian effects, etc.) may amplify the effect of IR (Auclert, 2016) Some key aspects of non-conventional MP including Forward Guidance and QE-transmission- are bettter understood from a heterogenous-agent perspective (McKay et al. 2016) 3
Individual heterogeneity seems relevant for MP: How should the CB incorporate it into its optimal policy? - Work in this area has been mainly positive, little progress on the normative side - Recent research by Nuño and Thomas (BdE 2016) analyzes the fully optimal monetary policy in a heterogeneous-households (HHs) economy Key ingredients: - HHs face uninsurable idiosyncratic risk à la Huggett (1993) - HHs hold nominal non-contingent assets Fisherian channel - Utility costs of inflation (due to costly price adjustment) 4
Main results - Discretionary MP features a redistribution-driven inflationary bias: With incomplete markets (and concave preferences), low-wealth agents have higher marginal utility than high-wealth ones - Under commitment, central bank promises to reduce inflation gradually over time (inflation front-loading): to avoid high inflation expectations being priced into new bond issuances: optimal long run inflation is zero (under general conditions) both debtors and creditors gain relative to discretion 5
Discretion vs Commitment - Discretionary: inflation starts and remains high, i.e. inflationary bias - Commitment: inflation starts high (no pre-commitments), but falls gradually towards long-run target ( zero) discretion commitment Source: Nuño & Thomas (2016) 6
What do we learn in practical terms? - Medium-long run: Anti-inflation commitment is optimal - Short run: There is margin to optimally exploit the redistribution channel By how much? It depends on the initial wealth distribution.hence, it is an empirical question. Source: Nuño & Thomas (2016) 7
Aggregate consequences of individual heterogeneity: An empirical view on the corporate sector - The Nuño-Thomas (2016) theory offers a stylized framework to think about optimal MP with heterogeneous households / distributional issues. - But the empirical and firm dimensions are relevant in practice too. - Two specific questions on this (focused on the BdE s experience): Squaring strong investment with strong deleveraging forces at the aggregate The transmission of the ECB s CSPP in a dual economy featuring many small firms and a few (very) large ones 8
The recovery of the Spanish economy: Deleveraging-cum-investment INVESTMENT (MACHINERY AND EQUIPMENT) 2008 q1 = 100 105 100 95 90 85 80 75 70 65 130 120 110 100 90 80 70 EQUIPMENT INVESTMENT AND NFCS DEBT 2008q1=100 60 2008 2010 2012 2014 2016 60 2012 2014 2016 SPAIN Euro Area INVESTMENT (SPAIN) INVESTMENT (EURO AREA) NFCs DEBT (SPAIN) NFCs DEBT (EURO AREA) Source: Banco de España, Eurostat 9
Aggregate deleveraging has been compatible with a significant share of investing firms raising their indebtedness Only a fraction of the firms distribution is relevant to understand aggregate investment dynamics during the current recovery - see Ottonello and Winberry (2017) for related evidence on the US economy ASSETS AND LIABILITIES OF NFCs WITH POSITIVE OR ZERO NET INVESTMENT. FLOWS ASSETS AND LIABILITIES OF NFCs WITH NEGATIVE NET INVESTMENT. FLOWS % GVA 40 % 80 % GVA 40 % 80 Resources 30 20 10 60 40 20 Resources 30 20 10 60 40 20 0 0 0 0 Uses -10-20 Uses -10-20 -20-40 -20-40 -30-60 -30-60 -40-80 2008 2009 2010 2011 2012 2013 2014 2015-40 2008 2009 2010 2011 2012 2013 2014 2015 Source: INE, Banco de España -28-27 -26-25 -24-23 -22-21 -20-19 -18-17 -16-15 -14-13 -12-11 -9-8 -7-6 -5-4 -3-10 NET CAPITAL INCREASE INTERNAL FINANCE FINANCIAL DEBT GROSS FIXED CAPITAL FORMATION FINANCIAL INVESTMENT NET TRADE CREDIT (ASSETS - LIABILITIES) OTHER NET FLOWS (ASSETS - LIABILITIES) FIRMS AS PERCENTAGE OF TOTAL (Right-hand scale) -80 10
But, which part of the distribution? Following the burst of the crisis, credit has been allocated towards firms that on average are more productive PROBABILITY OF OBTAINING A CREDIT: TIME- VARYING COEFFICIENT ON TFP 2.5 2 1.5 1 0.5 0 AVERAGE DIFFERENCE DIFFRENCE OF OF TOTAL TFP BETWEEN FACTOR PRODUCTIVITY FIRMS BETWEEN WITH NON-NEGATIVE FIRMS WHOSE CREDIT GROWTH DIMINISH AND THOSE WITH FIRMS NEGATIVE WHOSE CREDIT CREDIT GROWTH DO NOT DIMINISH (T-1) (T-1) % 14 12 10 8 6 4-0.5-1 -1.5-2 -2.5 2004 2005 2006 2007 2008 2009 2010 2011 2012 2 0-2 2004-2007 2008-2013 2014-2016 INDUSTRY TRADE AND ACCOMMODATION INFORMATION AND COMMUNICATION CONSTRUCTION AND REAL ESTATE ACT. OTHER SECTORS Source: INE, Banco de España Source: Banco de España 11
and with sounder financial fundamentals PROBABILITY OF GETTING A LOAN FROM ANY BANK 80 % 70 60 50 MEDIAN FIRM AVERAGE FINANCIALLY WEAK FIRM AVERAGE FINANCIALLY STRONG FIRM 40 30 05 06 07 08 09 10 11 12 13 14 15 16 Source: Banco de España 12
The transmission of the Eurosystem's corporate sector purchase programme (CSPP) in a dual economy CSPP has accelerated the trend towards less banking intermediation in Spanish corporates financing INTEREST RATES. CORPORATIONS LOANS AND FIXED INCOME GROSS ISSUANCE 12 month average CSPP Mar 16 CSPP Mar 16 3.0 % 155 b b 19 2.5 153 151 18 17 2.0 1.5 1.0 149 147 145 143 141 16 15 14 13 0.5 139 137 12 11 0.0 2014 2015 2016 135 2014 2015 2016 10 LOANS OVER 1 EUR MILLION NON FINANCIAL CORPORATIONS BONDS (BBB) BANK BONDS (BBB) NEW LOANS OVER 1 EUR MILLION LONG TERM FIXED INCOME BONDS (RHS) Source: Banco de España 13
Beside substitution effects, the CSPP is unchaining positive spillovers to non CSPP-elegible firms through banks credit reallocation Heterogeneity in the distribution of firms across sizes allows for a stronger transmission of monetary instrument (Arce, Gimeno, Mayordomo 2017) b 60 50 40 Loans to bond issuing corporations CSPP Announcement 20% 15% Changes in outstanding loans by corporation and type of bank (feb 2016 vs aug 2016) 30 10% 20 5% 10 0% 0-5% 201602 201603 201604 201605 201606 201607 201608 201609-10% micro and small firms medium firms large firms Banks with negative outflow with issuing firms Banks with positive outflow with issuing firms Fuente: Arce, O., Gimeno, R., y Mayordomo, S. (2017). Making room for the needy: The effects of the CSPP program. Work in progress. 14