Natural Expectations and Home Equity Extraction

Transcription

1 Natural Expectations and Home Equity Extraction Roberto Pancrazi 1 Mario Pietrunti 2 1 University of Warwick 2 Toulouse School of Economics, Banca d Italia 4 December 2013 AMSE Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

2 Motivation Introduction Spark of the recent global crisis: the 2007 mortgage crisis Boom of collateralized credit market related to housing wealth Economy excessively exposed to an unpredicted fall of house prices Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

3 Introduction Why Economy Excessively Exposed? 1 Financial instruments made housing a liquid asset 2 House prices are hard to predict These two ingredients likely to create distortions ancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

4 Intuition Introduction Simply from the Permanent Income Hyphotesis... Households have biased expectations about future housing prices Incentive to borrow today to consume perceived higher future housing wealth Transform future housing wealth in present consumption with Home Equity Loans Banks optimism allows households borrowing. ancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

5 Home Equity Extraction Introduction Housing wealth becoming a liquid asset Home Equity Line of Credit (HELOC): loan in which lender lend a maximum amount within an agreed period (term), where the collateral is the borrower s equity in her house. Ms. Kennedy says that with the $70,000 from the cash-out refinancing she got this year, she is investing in the stock market and considering buying another home for investment purposes. Her current home surged in value by $134,000 over the past 16 years. Now, her goal is to turn her cash-out windfall into $250,000 over the next 10 to 15 years. I just didn t want to let $70,000 sit in my home Wall Street Journal, July 26, 2001 ancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

6 Home Equity Extraction Introduction Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

7 What we do: Empirically Introduction Parsimoniously model biased expectation for house price 1 Natural agents fail to estimate the medium- long-run properties of house prices 2 Document that financial expert had biased expectations ancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

8 Introduction What we do: Theoretically Modelling Natural Expectations and Home Equity Extraction Households and Banks meet in the credit market Households pledge their house as collateral to obtain credit. Allow for both agents being natural Assess the effect of biased expectations during ancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

9 Findings Introduction Quantifying the Role of Natural Expections Debt to Income Ratio: 16% higher for natural households during boom Interest Rate: lower for natural agents Vulnerability: higher for natural agents when house prices decline Biased expectation contributed to Economy Excessively Exposed ancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

10 Outline House Price and HEE House Prices and Home Equity Exctraction Modelling Home Equity Extraction Natural Expectations for House Price Quantifying the Role of Optimism in the Credit Market Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

11 Home Equity Loans House Price and HEE Figure: Revolving Home Equity Loans in the U.S.: Trillion of dollars Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

12 House Price and HEE Home Equity Loans and House Prices Figure: Home Equity Loans-Disposable Income ratio and U.S. House Price index Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

13 House Price and HEE Home Equity Extraction and Education HELOC Exposure HELOC Participation Education Category Average Table: Heloc for Education Categories: Survey of Consumer Finance Duca and Kumar (2014): Financial illiteracy affected Home Equity Extraction Agurwar and Mazumder (2013): Cognitive ability affects House price forecast precision Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

14 Outline Model 1 House Prices and Home Equity Exctraction 2 Modelling Home Equity Extraction 3 Natural Expectations for House Price 4 Quantifying the Role of Optimism in the Credit Market Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

15 Model Elements of the model Infinitely lived economy Two agents: Household and Bank Two goods: consumption and housing Household owns house and pledge it as collateral Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

16 Model Household Problem max {ct,d t,def t} T t=0 EH 0 T β t u(c t, h), t=0 def t =0: def t =1: c t + (1 + r t 1 )d t 1 = y + d t c t + γp t h = y Expectation Bias p t+1 = p t ( 1 + r h t+1 ), r h t+1 = Θ p (L)r h t + σε t+1,, Default Probability { Pr {def t+1 = 1} = Pr ε t+1 1 [ ]} 1 + r t σ p t (1 + γ)h d t 1 Θ p (L)rt h Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

17 Bank Problem Model (r t 1 i t 1 )d t 1 + δe B t π t+1 (r t, d t, p t+1 ) if the household does not default (and did not default in the past) max d t π t (r t 1, d t 1, p t ) = κp t h (1 + i t 1 )d t 1 if the household defaults (but did not in the past) 0 if the household defaulted in the past. Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

18 Outline Natural Expectations for House Price 1 House Prices and Home Equity Exctraction 2 Modelling Home Equity Extraction 3 Natural Expectations for House Price 4 Quantifying the Role of Optimism in the Credit Market Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

19 Question for you Natural Expectations for House Price What was the real appreciation of the average U.S. house price (index) from 1955 to 2000? ancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

20 Question for you Natural Expectations for House Price What was the real appreciation of the average U.S. house price (index) from 1955 to 2000? Answer: 0%! ancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

21 Question for you Natural Expectations for House Price What was the real appreciation of the average U.S. house price (index) from 1955 to 2000? Answer: 0%! Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

22 U.S. House Prices Natural Expectations for House Price Figure: Real U.S. House Price index (upper panel) and its growth rate (lower panel) Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

23 Natural Expectations for House Price Natural and Rational Agents (1 Φ p (L)) r t = µ + ε t Natural agents estimate and forecast using AR(1) [Fuster et al. (2012), Hommes and Sorger (1998), Hommes and Zha (2013)] Rational (Sophisticated) agents estimate and forecast using BIC/AIC BIC: p=5 // AIC: p=16 Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

24 Natural Expectations for House Price Natural and Rational Estimates Whole Sample: 1953:1-2010:4 Subsample: 1953:1-1996:4 Natural Sophisticated BIC Sophisticated AIC Natural Sophisticated BIC Sophisticated AIC p φ [0.02] [0.06] [0.07] φ [0.10] [0.11] φ [0.10] [0.12] φ [0.10] [0.12] φ [0.06] [0.12] φ [0.13] φ [0.13] φ [0.13] φ [0.13] φ [0.13] φ [0.14] φ [0.13] φ [0.12] φ [0.13] [0.00] [0.08] [0.11] [0.11] [0.11] [0.12] [0.13] [0.13] [0.13] [0.12] [0.11] [0.11] [0.11] [0.08] [0.08] [0.13] [0.12] [0.12] [0.14] [0.14] [0.14] [0.14] [0.14] [0.14] [0.14] [014] [0.14] [0.12] [0.12] [0.08] φ [0.13] φ [0.08] φ [0.08] σ R Pancrazi, LRPietrunti ( 23.7 University of Warwick, 18.3Toulouse School of10.4 Economics, Natural Banca d Italia) HEE AMSE, Dec / 32

25 Natural Expectations for House Price Natural and Rational Impulse Response (1 Φ p (L)) r t = µ + ε t Naive: p=1 Soph.BIC: p= 5 Soph.AIC: p= Naive: p=1 Soph.BIC: p= 5 Soph.AIC: p= Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

26 Natural Expectations for House Price Natural Financial Experts We gather and analyze a unique data set. Source: Moody s Analytics Out-of-sample forecasts by using a rich demand-supply model (income trends, demographics cyclical factors) Quarterly forecasts from years horizon (quarterly) Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

27 Natural Expectations for House Price Financial Experts Forecast Realized Figure: Financial Experts Real House Price Out-of-Sample Forecasts Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

28 Natural Expectations for House Price Natural Financial Experts AR1-Forecast AR13-Forecast Actual Forecast Realized Prices Figure: Natural Forecast, Rational Forecast, and Financial Expert s Forecast Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

29 Outline Quantifying the Role of Optimism in the Credit Market 1 House Prices and Home Equity Exctraction 2 Modelling Home Equity Extraction 3 Natural Expectations for House Price 4 Quantifying the Role of Optimism in the Credit Market Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

30 Quantifying the Role of Optimism in the Credit Market Empirical Strategy Discretize house price growth rate (3 nodes) Solve the model by backward induction Natural Agents: AR(1) Rational Agents (DGP): AR(2) Natural - AR(1) Rational - AR(2) Annual Data ρ ρ σ Long-Run Persistence (LRP) Standard Deviation First-order Autocorrelation Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

31 Calibration Quantifying the Role of Optimism in the Credit Market Parameter Value Description β=δ 0.98 Discount rate for household and banks h 1.5 Housing stock η 2 CRRA coefficient y 1 Income per year γ 0.05 Rental rate as a fraction of house value κ 0.2 Collateral value for the bank as a fraction of house value housing price Figure: Simulated house price Dynamics ( ) Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

32 Quantifying the Role of Optimism in the Credit Market During House Price Boom (1) Both Rational, (2) Both Natural, 1 Natural household extracts more than 2 times than rational 2 At a lower interest rate 3 Natural household consumes more 4 Natural household 4 times more vulnerable Boom Scenario (1) (2) Debt to income (%) Cons to Income (%) Loan to Value (%) Interest Rate (%) Debt Service Ratio (%) Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

33 Quantifying the Role of Optimism in the Credit Market During House Price Bust 1 Abrupt deleveraging for natural household 2 Similar level of vulnerability 3 Rational household keeps constant DRS Bust Scenario (1) (2) Debt to income (%) Cons to Income (%) Loan to Value (%) Interest Rate (%) Debt Service Ratio (%) Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32

34 Conclusions Conclusions Housing: asset with large value and hard to forecast Financial market for housing: exposure to biased expectations Using a quantitative model, we find large effects due to biased expectations Pancrazi, Pietrunti ( University of Warwick, Toulouse School of Economics, Natural Banca d Italia) HEE AMSE, 4 Dec / 32