Capital Flows, House Prices, and the Macroeconomy Capital Flows, House Prices, and the Evidence from Advanced and Emerging Market Economies Alessandro Cesa Bianchi, Bank of England Luis Céspedes, U. Adolfo Ibáñez Alessandro Rebucci, Johns Hopkins University This paper was presented at Housing, Stability and the Macroeconomy: International Perspectives conference, November 14 15 213. The conference was sponsored by the Federal Reserve Bank of Dallas, the International Monetary Fund, and the Journal of Money, Credit and Banking. The conference was held at Federal Reserve Bank of Dallas (http://dallasfed.org).
Capital Flows, House Prices, and the Macroeconomy Evidence from Advanced and Emerging Market Economies 1 A. Cesa-Bianchi 1 L.F. Cespedes 2 A. Rebucci 3 1 Bank of England 2 U. Adolfo Ibanez 3 JHU Carey Business School 14 November 213 Dallas FED 1 The views expressed in this paper are those of the authors, and not necessarily those of the Bank of England. 1 / 25
Housing quintessential non-tradable asset & non-tradable sector at the core of financial crises... 2 SPAIN 18 IRELAND 1 15 5 147 3 1 1 113 3 5 15 3 6 9 12 8 1 3 6 9 12 16 SOUTH AFRICA 15 25 HONG KONG 1 127 7 183 3 93 2 117 3 6 1 77 8 83 86 Real House Price Index (left ax.) 5 1 89 92 95 98 1 Current Account / GDP (right ax.) 2 / 25
...capital abundant and highly mobile with limited investment opportunities 8 6 Billion US$ 4 2 2 23 26 29 212 US M EMEs Reserves 3 / 25
Contribution New comprehensive, quarterly house price data set comprising 57 advanced and developing economies A new set of house price stylized facts Characteristics of house price booms Transmission of a global liquidity shock 4 / 25
Preview of the results Relative to AEs, house prices in EMEs are Slower and more associated with fundametals, more volatile and less persistent More associated with external variables Relative to AEs, house price booms in EMEs are Larger, more closely associated with loose global liquidity conditions A global liquidity shock has A stronger impact on consumption in EMEs Qualitatively different impact on external variables 5 / 25
Outline House Price Data & Descriptive statistics Event Study Global Liquidity VAR Analysis Conclusion 6 / 25
Data Unbalanced panel of 57 time series with varying coverage from 197.I 212.IV Source: OECD house price database, the BIS new property price data set, national central banks, national statistical offices, and academic publications on housing markets Value added relative to readily available datasets Additional countries: Argentina, Brazil, Chile, Colombia, Croatia, India, Peru, Taiwan, Ukraine and Uruguay Additional historical data: Austria, Czech Republic, Estonia, Hong Kong, Hungary, Indonesia, Malaysia, Philippines, Poland, Serbia, Singapore, Slovakia, Slovenia and Thailand. 7 / 25
Real house price annual returns Summary statistics Real Real Real House Price GDP Consumption Group AEs EMEs AEs EMEs AEs EMEs Mean 2.% 1.2% 2.2% 3.8% 2.3% 4.% Median 2.1% 1.5% 2.5% 5.% 2.4% 4.7% Max 18.3% 27.5% 7.% 13.3% 7.5% 16.7% Min -12.5% -34.5% -5.8% -13.3% -3.9% -16.4% St. Dev. 6.4% 12.5% 2.3% 5.1% 2.2% 5.9% Auto Corr..92.86.83.87.81.79 Skew..1 -.44-1. -1.12 -.31 -.67 Kurt. 3.2 4.34 4.91 6.15 3.88 6.12 Note. The country-specific summary statistics are averaged across each group, namely advanced economies (AEs) and emerging economies (EMEs) and are computed across the common sample 1985.I 212.IV. 8 / 25
Average and the standard deviation of real house price annual returns 1 (a) Moving Average AEs 1 (b) Moving Average EMEs 5 5 Percent 5 Percent 5 1 95 5 1 1 95 5 1 15 (c) Moving Std. Deviation AEs 15 (d) Moving Std. Deviation EMEs Percent 1 5 Percent 1 5 95 5 1 95 5 1 9 / 25
Cross-correlations of real house price annual returns (AEs) (a) Advanced Economies.75 Real GDP Real Consumption.75.75 CPI Labor Productivity.75.5.5.5.5.25.25.25.25.25.25.25.25.5 4 3 2 1 +1+2+3+4.5 4 3 2 1 +1+2+3+4.5 4 3 2 1 +1+2+3+4.5 4 3 2 1 +1+2+3+4.75 Real Equity Price Short term Int. Rate.75 Real Eff. Exch. Rate.75 Current Account / GDP.75.5.5.5.5.25.25.25.25.25.25.25.25.5 4 3 2 1 +1+2+3+4.5 4 3 2 1 +1+2+3+4.5 4 3 2 1 +1+2+3+4.5 4 3 2 1 +1+2+3+4 1 / 25
Cross-correlations of real house price annual returns (EMEs) (b) Emerging Economies.75 Real GDP Real Consumption.75.75 CPI Labor Productivity.75.5.5.5.5.25.25.25.25.25.25.25.25.5 4 3 2 1 +1+2+3+4.5 4 3 2 1 +1+2+3+4.5 4 3 2 1 +1+2+3+4.5 4 3 2 1 +1+2+3+4.75 Real Equity Price Short term Int. Rate.75 Real Eff. Exch. Rate.75 Current Account / GDP.75.5.5.5.5.25.25.25.25.25.25.25.25.5 4 3 2 1 +1+2+3+4.5 4 3 2 1 +1+2+3+4.5 4 3 2 1 +1+2+3+4.5 4 3 2 1 +1+2+3+4 11 / 25
Event study We identify 66 real house prices booms (Bordo and Jeanne, 22) g i,t + g i,t 1 + g i,t 2 3 g ± xσ During the identified boom episodes Investigate the behavior of relevant macroeconomic variables (output gap, exch. rates, current acconut, capital inflows, VIX,...) Investigate the role played by country characteristics (fin. market depth, exch. rate flexibility,...) 12 / 25
Event study Results ()R (a) Real lhouse Pi Prices (b) Output Gap (Average increase during episodes, percentage) (Average increase during episodes,,percentage) 8. 6 6. 6. 45 4.5 4. 3 3. 2. 15 1.5.. Advanced Economies Emerging Market Economies Advanced d Economies Emerging Market teconomies (c) Current Account (Average increase during episodes, percentage). 16. (d) Real Exchange Rate (Average increase during episodes, percentage) 1. 12. 2. 8. 3. 4. 4.. Advanced Economies Emerging Market Economies Advanced Economies Emerging Market Economies 13 / 25
Event study Results (cont d) (e) Capital Inflows (Averageincreaseduringepisodes episodes, percentage) 4. 3. (f)global Liquidity (Averageincreaseduringepisodes episodes, percentage) 3. 22.5 2. 15. 1. 7.5.. Advanced Economies Emerging gmarket Economies Advanced Economies Emerging gmarket Economies (g) VIX Index (Average increase during episodes, percentage) 1 1. 5.5 (h) US Real Interest Rate (Average increase during episodes, percentage).. 1. 1.5 5 2. 1. 3. 1.5 4. 4 2. 2 Advanced d Economies Emerging Market Economies Advanced d Economies Emerging Market Economies 14 / 25
Real house price determinants in boom episodes Dependent variable: change in real house price during boom Explanatory variable (1) (2) (3) (4) (5) (6) (7) (8) Capital inflows 2.26 4.23 4.5 4.59 (2.35) (2.47) (1.91) (2.53) Global liquidity.58.88 1.21 1.1 (3.4) (2.87) (3.77) (2.13) Dummy AEs Financial market depth -.9.2 (-.73) -.14 Exchange rate flexibility -.55.66 (-.51) -.44 Dummy AEs -5.27 Capital inflows (-2.73) Financial market depth -.5 Capital inflows (-1.73) Exchange rate flexibility -.38 Capital inflows (-1.9) Dummy AEs Global liquidity -.7 (-2.21) Financial market depth Global liquidity -.1 (-1.95) Exchange rate flexibility -.6 Global liquidity (-1.1) R2.6.16.19.11.14.22.28.16 Number of observations 6. 6. 58. 6. 66. 66. 62. 66. F test 5.51 3.73 1.66 2.3 9.25 4.17 5.89 3.11 Note. All regressions are estimated using a constant, t-test in parenthesis. Significance levels at 1%, 5%, and 1% is denoted by (), (), (), respectively. 15 / 25
Global liquidity: a push factor for capital flows Empirical models of international capital flows typically include push (i.e., global) and pull (i.e., local) drivers Global liquidity is a proxy for the monetary policy stance in whole world economy, as opposed to any individual economy pulling in capital flows or the rest of the world economy pushing them to a particular country We measure global liquidity in three different ways 1 Official global liquidity 2 Private global liquidity 3 VIX Index 16 / 25
Global liquidity measures & VIX index 2 (a) Official and private global liquidity (level) & VIX Index (level) 5 15 4 1 3 5 2 1 79 82 85 88 91 94 97 3 6 9 12.3 (b) Official and private global liquidity (log change) & VIX Index (level) 5.2 38.1 25. 13.1 79 82 85 88 91 94 97 3 6 9 12 Off. liquidity (left ax.) Priv. liquidity (left ax.) VIX Index (right ax.) 17 / 25
Correlation between global liquidity measures Off. Liquidity (level) Priv. Liquidity (level) Off. Liquidity (level) Priv. Liquidity (level) VIX index (level) VIX index (level) Full Sample.92 -.5.1 Pre-Crisis.99 -.3 -.28 Post-Crisis -.12. -.41 Off. Liquidity (log diff.) Priv. Liquidity (log diff.) Off. Liquidity (log diff.) Priv. Liquidity (log diff.) VIX index (level) VIX index (level) Full Sample.29 -.18 -.6 Pre-Crisis.38 -.13 -.23 Post-Crisis.43.12.32 Note. Note here. 18 / 25
A panel VAR with pull and push factors Vector autoregression (VAR) model for country i includes X = GLOBAL LIQUIDITY CONSUMPTION REAL HOUSE PRICE SHORT-TERM INT. RATE REAL EFF. EXCH. RATE CURRENT ACC. / GDP We identify only exogenous changes to one particular push factor: global liquidity Identification assumption: no individual country is large enough to affect it significantly within a given quarter Mean group estimator (dynamic panel data models with heterogenous slope coefficients) 19 / 25
Checking our identification assumption: FEVD to a global liquidity shock 1 (a) AEs PRIV. LIQ 1 (b) EMEs PRIV. LIQ 9 9 8 8 7 7 Percent Explained 6 5 4 Percent Explained 6 5 4 3 3 2 2 1 1 5 1 15 2 25 3 35 4 5 1 15 2 25 3 35 4 2 / 25
Impulse response function to a global liquidity shock (AEs) (a) Advanced Economies 3.5 PRIV. LIQ..3 CONS.6 RHP Percent Deviation 3 2.5 2 1.5 1.5.5 1 2 3 4 Percent Deviation.2.1.1.2 1 2 3 4 Percent Deviation.4.2.2.4.6.8 1 2 3 4 7 IRS.6 REER 15 CA Percent Deviation 6 5 4 3 2 1 1 1 2 3 4 Percent Deviation.5.4.3.2.1.1.2 1 2 3 4 Percent Deviation 1 5 5 1 15 1 2 3 4 21 / 25
Impulse response function to a global liquidity shock (EMEs) (b) Emerging Economies 3.5 PRIV. LIQ..6 CONS 2 RHP Percent Deviation 3 2.5 2 1.5 1.5.5 1 2 3 4 Percent Deviation.4.2.2.4 1 2 3 4 Percent Deviation 1.5 1.5.5 1 2 3 4 6 IRS.6 REER 2 CA 4.4 1 Percent Deviation 2 Percent Deviation.2 Percent Deviation 1 2 2.2 3 4 1 2 3 4.4 1 2 3 4 4 1 2 3 4 22 / 25
Conclusions In this paper we explore empirically the relation among capital flows, house prices, and the broader macroeconomy We find that: House prices in EMEs are slower, more associated with fundametals and external variables, more volatile and less persistent House price booms in EMEs are larger, more closely associated with loose global liquidity conditions A global liquidity shock has a stronger impact on consumption in EMEs with qualitatively different impact on external variables 23 / 25
Conclusions (cont d) We interpret this evidence as suggesting that while global imbalances may have played a lesser role in the housing boom in AEs, the increase in global liquidity in response to it may be playing an important role for house price dynamics in EMEs Work to do Better understanding of the mechanisms Exploring the distribution around the means 24 / 25
THANK YOU 25 / 25