Demographic change, long-run housing demand and the related challenges for the Irish banking sector December 5 th, 2016 esri.ie David Duffy, Daniel Foley, Kieran McQuinn and Niall McInerney
Outline: Addresses two-related mortgage market issues A new model of long-run housing demand Closely related to demographics Using the model generate Forecasts of future housing demand Based on COSMO Assess implications for the banking sector Sizeable increase in credit/balance sheets Policy issues which arise?
Model of Long-Run Housing Demand: Builds on earlier research Duffy, Byrne and FitzGerald (2014) Using micro-data, they look at Household formation and tenure choice They estimate a measure of realised demand Over time (1982 2015) We examine relationship between Activity and structural demand
Structural Demand and Housing Activity Units 100000 90000 80000 70000 60000 50000 40000 30000 20000 10000 0 1982 1986 1990 1994 1998 2002 2006 2010 2014 Demand Activity
Total Credit and Total Deposits ( millions) 300000 250000 200000 150000 100000 50000 0 1995 1999 2003 2007 2011 2015 Total Credit Total Deposits
Annual Net Migration Ireland (000 s) 120.0 100.0 80.0 60.0 40.0 20.0 0.0-20.0 1987 1991 1995 1999 2003 2007 2011 2015-40.0-60.0
Model of Long-Run Housing Demand: Long-Run Model: = f α t 0 α1hht + α 2acred α t 2 where f + = activity levels hh = household formation acred = adjusted credit variable S = total housing stock S t
Actual vs Fundamental Housing Activity Units 100000 90000 80000 70000 60000 50000 40000 30000 20000 10000 0 1990 1994 1998 2002 2006 2010 2014 Activity Structural Demand
Short-Run Models of Housing Demand: Estimate error correction model (ecm): f t Based on long-run model Also include short-run dynamics for Prices and interest rates. Use the ecm To generate forecasts of activity levels Key variable: Forecast of future demographic trends COSMO forecasts
Units Household Formation Forecasts 40000 35000 30000 25000 20000 15000 10000 5000 0 2012 2014 2016 2018 2020 2022 2024
Units Actual vs Structural Demand 40000 35000 30000 25000 20000 15000 10000 5000 0 2012 2014 2016 2018 2020 2022 2024 Structural Demand Activity
REQUIRED HOUSING SUPPLY IN COSMO: Required Housing Stock Residential Investment House Prices Corporate Credit Commercial Property Mortgage Credit Corporate Interest Rate Banks Balance Sheets RWA Capital Deposits Mortgage Interest Rate
Irish Banking Sector - Background Liquidity problems of Irish banks faced Heavy reliance on short-term money market funding, Funding base that is more deposit-based Loan-to-deposit ratios introduced as part of CBI (2011) Basel III greater reliance on stable funding Crucial assumption: COSMO: deposits are assumed to grow with GDP Lending in excess of this growth rate Funded from non-deposit sources?
HOUSING INVESTMENT AND HOUSING STOCK 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 Residential Inv. mn (lhs) H. Stock '000s (rhs) 2350 2300 2250 2200 2150 2100 2050 2000 1950 0 2016 2017 2018 2019 2020 2021 2022 2023 2024 1900
Credit Supply in COSMO New mortgages (per new housing unit) depend on: Fundamentals: income (+), house prices(+), mortgage rate (-) Credit conditions: LTI (+) and LTV (+) NewM t = 5.8 + + 0.6* HP 0.5* NewM t 1 t 1 + 0.7 * LTV 0.01* MRate t t 1 + 0.5* LTI t + 0.8* Income t New mortgages accumulate on previous period s mortgage stock Construction and Real Estate loans are a function of: Economic activity (+), User cost of capital (-) Commercial property values and house prices (+) Non-Property Corporate Credit determined by: Economic activity (+), Corporate profits (-), Cost of capital (+) Commerical property values (+) 15
Required level of credit ( bns)? Mortgages NFC Prop NFC Non-Prop 160 140 120 100 80 60 40 20 0 2016 2017 2018 2019 2020 2021 2022 2023 2024
Funding Requirements of the Banks? Required expansion in lending of 130bn Similar expansion in funding liabilities. This may take the form of Higher levels of retail deposits, short- and long-term debt securities, or Levels of capital. COSMO: Retail deposits assumed to grow in line with GDP while, Capital levels determined by features of the banking sector Size, its exposure to the real estate sector and profitability. Remaining liabilities of the banking sector are treated as a residual.
Projected levels of Deposits and Capital ( bn), and the Loan-to-Deposit Ratio 250 230 210 190 170 150 130 110 90 70 50 Deposits bn (lhs) Capital bn (lhs) LTD (rhs) 2016 2017 2018 2019 2020 2021 2022 2023 2024 1.4 1.2 1 0.8 0.6 0.4 0.2 0
Real Economy Implications Higher level of residential investment Implies a higher level of demand relative to the baseline. Output gap is 0.3 percentage points higher Fiscal intervention in the form of Higher personal/ consumption taxes or Reduced government spending. Irish household sector may not be able Fund the necessary housing investment Greater role for the rental sector? With substantial equity injections, possibly from outside Ireland
Financial Sector Implications Regulatory developments impact on banks funding? Net Stable Funding Ratio (NSFR) A share of liabilities to be of more medium/long-term duration Issuance of covered bonds Debt securities backed by cash flows from the issued loans. Typically cheaper than other types of non-secured funding Bonds are treated as high-quality collateral by the ECB. Required funding achieved through Entry of foreign banks into the Irish retail banking sector. Access to stable medium- to long-term funding via parent
Finally Thoughts How do we stop another credit bubble from occurring? Macroprudential policy crucially important Duffy, McInerney and McQuinn (2016) called for Inclusion of a counter-cyclical component in the policy framework. Key housing market variables examined on a routine basis and Loan-to-value, debt-to-income ratios and CCCB Re-calibrated to reflect the changing circumstances in the market. Housing activity needs to be one such variable Actual housing activity > structural demand Macroprudential measures adjusted accordingly