Betting the House Oscar Jorda Moritz Schularick and

“Betting the House” Oscar Jorda, Moritz Schularick, and Alan M. Taylor Discussion by Jim Wilcox Haas School of Business University of California, Berkeley September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 1

Outline Big Data n Big Questions n Time-Varying Results n Time-Varying Banks and Mortgage Markets n Concluding Observations n September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 2

No Real Complaint About the Paper, Until … the Title n Like home owners, banks also bet • Banks’ losses can lead to credit crunches and crises n Rather than “Betting the House”, maybe… • • • Effects of External Monetary Shocks on Housing Banking on Housing Betting on Houses Betting the Bank on Houses Playing with House Money September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 3

Big Data: Across Countries and Decades n n Huge sunk costs: time, effort, money, etc. Data only for commercial banks • Nonbanks’ shares of credit also vary across countries and decades n Data for mortgage and nonmortgage loans • Mortgages for owner-occupied, multi-family, plus commercial buildings n Their shares also vary, a lot, across time and space September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 4

Big Patterns in the Data n n Large, pervasive, recent rise in mortgages Large declines in nominal mortgage rates permitted more borrowing and buying by the previously-cash-payment-constrained • Starting in mid-1980 s in U. S. • Enough “Eu-rope” to hang Spain, Portugal, Ireland, … n Explains some of the big increases in ratios • • Mortgages (relative to GDP) House prices (relative to per-capita incomes) September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 5

Applying Big Data to Big Questions n n Effects of monetary policy on long rates, mortgages and house prices (and crises) were large and pervasive, but mostly after WWII Fortunately, all that pre-WWII data still useful • With commercial banks’ ignoring residential mortgages, I would expect weaker responses then of their mortgages and of house prices to rates n n Results pass this placebo test, however inadvertently Booms, busts, bubbles: mentioned, not tested September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 6

Notes for a New Instrument for Monetary Policy n IV for domestic, short-term, interest rates based on external rates • With adjustments for exchange rate regimes and for capital-flow regimes n n Domestic rates had detectable, but very low, correlation with, and response to, the IV, zit The much-stronger correlation of domestic rates with the domestic variables, given zit, is worrisome September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 7

Asides About Estimation n Flexibility of Jorda’s local-projection method Specification restricts effects’ magnitudes and lag patterns to be the same across countries First-differencing of all variables • Implies that the fixed-effects absorb constant, country-specific trends in any of the variables • Effectively raises weight on higher-frequencies • Absent co-integrating equations, untethers levels September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 8

After WW II, Mortgages Trend Up (percent of potential GDP, 1896 -1999) September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 9

After WW II, Residential, Not Commercial, Mortgages Trend Upward (percent of potential GDP, 1896 -1999) September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 10

Commercial Mortgages Were Large, But Falling, Share of All Mortgages (percent of all mortgages, 1896 -1999) September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 11

Some Residential Mortgages Were in Commercial Banks (percent of potential GDP, 1896 -1999) September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 12

Banks’ Share of Residential Mortgages: Low and Slowly-Rising (percent of all residential mortgages, 1896 -1999) September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 13

Relative to Banks, Nonbanks Held Vastly More Residential Mortgages (percent of potential GDP, 1896 -1999) September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 14

Concluding Observations n n n Current PTI seems better predictor of recent defaults than current LTV (e. g. , underwater) Commercial real estate had greater volatility of construction, prices, mortgage defaults, and roles in depositories’ losses and failures Deflated role of inflation • Mortgages/GDP correlated with inflation n Positively but modestly, and likely reversed after 1970 s • Irving Fisher contended deflation was crucial September 5, 2014 Jim Wilcox Conference on Housing and Monetary Policy UCLA Ziman Center for Real Estate and FRB San Francisco 15