Docstoc

what is loan to value ratio

Document Sample
what is loan to value ratio Powered By Docstoc
					                                       Seminar, Czech National Bank,
                                                 November 24th 2006




  House price inflation and household
  indebtedness: Is there a problem for
       macroeconomic stability?


                      Richard Disney
Centre for Finance and Credit Markets, School of Economics,
University of Nottingham, and Institute for Fiscal Studies, London
  Macroeconomic stability, house prices
          and indebtedness
• Bank of England has independent role in setting
  interest-rates to obtain macroeconomic stability
• Asset price stability is not a goal per se, but asset
  price instability may have implications:
   – For inflation, via impact on consumer spending (especially
     debt-financed consumption)
   – For stability of banking sector where defaults on debt affect
     balance sheets of financial institutions (e.g. Basel)
• Movements in house prices, and in consumer debt
  are closely watched by Bank of England
                     Plan of talk
• Look at aggregate data on house price trends, debt
  (secured & unsecured) and consumption in UK
• Examine a popular model of how increases in asset
  values induce strong effects on aggregate consumption –
  the ‘financial accelerator’ model.
• Use microeconomic (household panel) data to examine
  impact of changes in asset values (housing) on debt-
  financed consumption
• Suggests that the popular model (and using
  macroeconomic rather than microeconomic data)
  overstates impact of changing asset values on
  consumption.
                                            Some aggregate UK trends1:
                                               House price growth
                         Changes in nominal and real house price indices,
                                         1971 to 2003
                         60

                         50
per cent annual change




                         40

                         30
                                                                                                                                                  Update:
                         20                                                                                                                       Further growth 2004,
                                                                                                                                                  levelled out 2005,
                         10
                                                                                                                                                  now growing again:
                          0                                                                                                                       see next slide
                         -10

                         -20
                           1971
                                  1973
                                         1975
                                                1977
                                                       1979
                                                              1981
                                                                     1983
                                                                            1985
                                                                                   1987
                                                                                          1989
                                                                                                 1991
                                                                                                        1993
                                                                                                               1995
                                                                                                                      1997
                                                                                                                             1999
                                                                                                                                    2001
                                                                                                                                           2003
                          All house mix-adjusted index                                            All house index less RPI
                Some aggregate UK trends 2:
                 Mortgage equity withdrawal
                                 Housing equity withdrawal and house price changes
                                            Quarterly changes 1991-2006

              10.0


               8.0

               6.0




                                                                                                                                                                                                             % per quarter
as % of GDP




               4.0


               2.0


               0.0

              -2.0


              -4.0
                     30-Jun-91

                                 30-Jun-92

                                             30-Jun-93

                                                         30-Jun-94

                                                                       30-Jun-95

                                                                                   30-Jun-96

                                                                                               30-Jun-97

                                                                                                           30-Jun-98

                                                                                                                         30-Jun-99

                                                                                                                                     30-Jun-00

                                                                                                                                                 30-Jun-01

                                                                                                                                                             30-Jun-02

                                                                                                                                                                         30-Jun-03

                                                                                                                                                                                     30-Jun-04

                                                                                                                                                                                                 30-Jun-05
                                                                     HEW as % of Y(1-t)                                Change in house prices
  Some aggregate UK trends 3:
Household spending & indebtedness
                                            Household Spending and Indebtedness


                    14

                   13.5

                    13

                   12.5
   ln(£ million)




                    12

                   11.5

                    11

                   10.5

                    10

                    9.5

                     9
                          Jun-93

                                   Jun-94

                                            Jun-95

                                                     Jun-96

                                                                 Jun-97

                                                                          Jun-98

                                                                                   Jun-99

                                                                                            Jun-00

                                                                                                     Jun-01

                                                                                                              Jun-02

                                                                                                                       Jun-03

                                                                                                                                Jun-04

                                                                                                                                         Jun-05
                                                              ln(Household final consumption expenditure)
                                                              ln(Total unsecured household debt)
                                                              ln(Total secured household debt)
                      Some aggregate UK trends 4:
Secured & unsecured debt as % of household income
                      Total Debt as % of Household Disposable Income

                5.0                                                                                                                         0.30
                4.5
                4.0                                                                                                                         0.25




                                                                                                                                                   Unsecured debt
 Secured debt




                3.5                                                                                                                         0.20
                3.0
                2.5                                                                                                                         0.15
                2.0
                1.5                                                                                                                         0.10
                1.0                                                                                                                         0.05
                0.5
                0.0                                                                                                                         0.00
                      Jun-93

                                Jun-94

                                         Jun-95

                                                  Jun-96

                                                           Jun-97

                                                                    Jun-98

                                                                             Jun-99

                                                                                      Jun-00

                                                                                               Jun-01

                                                                                                        Jun-02

                                                                                                                 Jun-03

                                                                                                                          Jun-04

                                                                                                                                   Jun-05
                               Secured debt as % of Y(1-t)                                     Unsecured debt as % of Y(1-t)
    How asset values affect consumption

• Impact of changes in house prices on debt-financed
  consumption by 2 routes:
   – via impact on life cycle wealth
   – via impact on collateralised lending – ‘financial
     accelerator’
• If debt is secured on nominal collateral value, house
  price shocks ‘amplify’ or ‘accelerate’ changes in
  consumer spending during general inflation. This
  termed the ‘financial accelerator’.
• What role for unsecured debt?
   Existing literature on ‘financial accelerator’

• Bernanke, Gertler & Gilchrist (1999) – imperfect information
  in credit markets affects quantity and cost of external
  finance.
• Originally applied to corporate finance (v. Modigliani-Miller),
  now applied to household behaviour
• Examples:
   – Iacoviello (2004, 2005): Aggregate US data, calibration, VARs
   – Aoki, Proudman & Vlieghe (2004): Aggregate UK data,
     calibration, VARs
   – Lamont and Stein (1999): US regions
   – Almeida, Campello & Liu (2005): cross country data
   – And many others….
    The ‘household financial accelerator’
       as described by Aoki et al (2004)
“Houses represent collateral for homeowners, and borrowing on a
secured basis against ample housing collateral is generally
cheaper than borrowing against little collateral or an unsecured
basis (via a personal loan or credit card). So an increase in
house prices makes more collateral available to homeowners,
which may in turn encourage them to borrow more, in the form of
mortgage equity withdrawal, to finance desired levels of
consumption and housing investment. The increase in house
prices may be caused by a variety of shocks, including an
unanticipated reduction in real interest rates, which will lower the
rate at which future housing services are discounted.”
   The collateral constraint illustrated

Ct+1




                      rs




                                       Ct
            Yt    Yt +E(vT+1)Ht−Rs
       The financial accelerator

Ct+1




                    rs




                         ∆E(vt+1)
                         = ∆Ct

                                    Ct
           Yt   Yt +E(vt+1)ht−Rs
        But no role for unsecured debt?

• Financial Accelerator model plays down role of
  unsecured debt.
• Assumes either most borrowing is in practice
  collateralised…
    “Consumers are actually inundated by offers of car
    loans, credit cards, home equity loans, and so
    on…Most of these loans require the borrower to
    post some collateral.” (Iacoviello, 2004)
• …or unsecured depends (indirectly) on having
  collateral
• …or unsecured borrowing is too expensive.
       But unsecured debt is pervasive….

• Although mortgage is largest single household debt,
  and is secured, most smaller debt-financed purchases
  are effectively unsecured.
• Most households use unsecured debt even when they
  have ‘unused’ collateral (Bertaut and Haliassos, 2006).
• Credit bureaux ‘score’ home ownership highly but this
  signals permanent address, secured job etc rather than
  collateral value.
• It may be more costly to change value of collateralised
  loan (admin fee) than value of unsecured debt.
   Relaxing the collateral constraint in a
       model with unsecured debt
• A shock to asset values allows the household to
  substitute secured debt for unsecured debt, reducing
  the average cost of debt
• The wealth effect of this relative price effect may
  increase debt-financed consumption
• But the presence of unsecured debt reduces the
  ‘amplification’ or ‘accelerator’ effect of changes in
  asset values
• And with costs of adjustment of secured debt
  (remortgage fees), some households may not adjust
  at all.
The collateral constraint with unsecured debt:
   consumption is higher, but lower than
           unconstrained consumer
Ct+1




                                       Unsecured
                           rs            debt


                 Secured
                  debt



                                                   ru
                                                        Ct
            Yt      Yt +E(vt)Ht+1−Rs
Relaxing the collateral constraint with
           unsecured debt
Ct+1




                    rs



                              ∆Ct


                         ∆E(vt+1)
                                    ru
          Yt    Yt +E(vt+1)ht−Rs         Ct
               Our empirical strategy

• To examine financial accelerator model, and a model
  augmented by unsecured debt, using household data.
• Objects:
   – What % of households face ‘collateral constraints’?
   – Is value of unsecured debt related to value of
     collateral?
   – Do collateral constrained households utilise more
     unsecured debt?
   – If collateral constraint unbinds, do households increase
     secured debt (remortgaging)?
   – And what happens to value of total debt (after
     substitution of secured for unsecured debt)?
    Data: British Household Panel Survey (BHPS)
                 1991 on, 5000 households

Secured debt/housing equity
• Every wave contains information on:
   –   Self-assessed house value
   –   Original date, value and terms of mortgage/remortgage
   –   Current self-assessed value of mortgage
   –   Interest payments
   –   Motives for remortgage

• Calculate loan-to-value ratio:
   – Cross section: self-assessed current mortgage value divided
     by self-assessed housing house value.
   – Panel: predicted value of current mortgage(s) from original
     value & terms (e.g. duration & type) of mortgage divided by
     self-assessed housing value.
                    Data: continued

Unsecured debt
• 1995 and 2000 waves contain information on:
   – Access to types of credit arrangements
   – Whether indebted on different credit instruments
   – Total value of debt (other than mortgage)

• We control for life cycle effects with demographics,
  income, assets etc (polynomials)
• We control for preferences by ‘whether a regular
  saver’, ‘smoker’ and random effects.
          Distribution of the change in the instrumented
                  loan to value ratio 1995 to 2000

900
800
700
600
500
400
300
200
100
  0
      -.8 to - -.6 to - -.4 to - -.2 to - 0 to - 0.1 to 2.1 to 4.1 to 6.1 to 8.1 to
         1       .79      .69      .39     .19     2      4      6      8      1
                          0
                        200
                        400
                        600
                        800
                       1000
                       1200
                       1400
                       1600
                       1800
                       2000
        0         0
            to
    50           50
       1           0
           to
   10       10
      01       00
         to
   15       15
      01       00
         to
   20       20
      01       00
         to
   25       25
      01       00
         to
   30       30
      01       00
         to
   35       35
      01       00
         to
   40       40
      01       00
         to
   45       45
      01       00
50       to
   01       50
      an       00
         d
                              Distribution of self-reported unsecured debt 1995




           ab
              ov
                e
         Benchmark empirical findings

• Campbell & Cocco (Journal of Monetary Economics,
  November 2006) find elasticity of consumption to house
  price changes in UK of 1.3 to 1.9:
• For average values of house, consumption, this translates
  into MPC out of housing wealth of around 0.1 (0.08 to
  0.12).
• Iacoviello (Journal of Health Economics, 2005), from
  calibrated macroeconomic model, gets response of
  consumption to house price shock of 1.3 to 1.9.
• Most studies (e.g. Campbell & Mankiw 1989, Campbell
  and Cocco 2006, Aoki et al 2004, Iacoviello, 2004, 2005)
  assume proportion of collateral constrained households is
  between 0.3 and 0.5.
What % of households are collateral constrained?
                   Distribution of Net Worth 1995, 2000 and Pooled
                                   1-100th Percentile                                            Proportion of sample with 1995 Loan-to-
                                                                                                            Value Ratio >0.x

                   100
                                                                                                 0.50


                   90


                   80                                                                            0.40


                   70


                   60                                                                            0.30




                                                                                   % of sample
                                                                          Pooled
 CDF %




                   50                                                     2000
                                                                          1995
                   40                                                                            0.20


                   30


                   20                                                                            0.10



                   10

                                                                                                 0.00
                    0
                                                                                                        0.95   0.9   0.85   0.8   0.75    0.7   0.65
         -100000         0      100000     200000       300000   400000
                                                                                                                      LTV ratio 1995
                             Nom inal Value Net Worth

                                                                                                                            % of sample
                               Unsecured debt and housing equity
          (Method: Random effects probit/tobit, pooled over 1995 & 2000 waves)
   Variable           Prob.          Prob. owing      Prob. owing       Prob. owing        Value of
                     owning a         money on         money on         money on a        unsecured
                     credit or         credit or       mail order        personal         debt (tobit)
                    store card        store card       purchase            loan
Homeowner=1           0.82**            0.54**            −0.08             0.16*            0.63**
                      (0.08)            (0.08)            (0.08)            (0.06)           (0.18)
                       0.15              0.09             −0.01              0.03            (0.18)
  Value of           0.002**          −0.002**          −0.002**          −0.002**          −0.01**
housing equity       (0.0004)          (0.001)           (0.001)           (0.001)          (0.001)
                      0.0003         −0.0002            −0.0003           −0.0003           (0.002)
   N (obs)             7418              7418             7418               7418          N=0 4057
  N (groups)           3709              3709             3709               3709          N>0 3341
    LogL             −3196.0           −3318.0          −2677.3           −3087.9          −11902.5

Notes: Value of housing equity = self-reported value of home minus predicted value of current mortgage.
Each cell contains, respectively, coefficient, standard error, marginal effect. **=1% significance, *=5%
significance. Controls include: quadratic in household income, in value of financial assets, and in age of
head of household, gender of head of household, number of children, employment and retirement status,
total number of social security benefits received, educational qualifications and whether head of household
saves regularly and is/is not a smoker.
                Loan-to-Value Ratio and Value of Unsecured Debt

         Specification: Tobit              (1)              (2)               (3)
         LHS Variable: Value              1995             2000            Pooled
         of unsecured debt                                                random
         (£000)                                                            effects
         Loan-to-value ratio            2.50**               5.76**       3.41**
                                         (0.42)            (0.63)          (0.41)
         H income (£)                    0.05**             0.001         0.06**
                                         (0.01)            (0.001)         (0.01)
         Age of H of H                  0.13*               0.07          0.14**
                                        (0.06)             (0.08)          (0.05)
         Age of H of H2                −0.002*            −0.001         −0.002**
                                       (0.001)            (0.001)         (0.001)
         Employed=1                     1.06**              1.27*         0.96**
                                         (0.38)             (0.58)         (0.36)
         Fin. Assets (£)                −0.008               −0.01       −0.016**
                                         (0.006)             (0.008)      (0.005)
         N of obs                            2369                2569           4938
         Log L                            −3689.5            −4284.5       −8043.4
         Wald/LR χ2 (19)                   4627.3               721.2          840.6
         Prob>χ2                           0.0000             0.0000         0.0000

Notes: Specifications also include constant term and dummies for highest educational qualifications,
for whether smoker, regular saver, gender of HofH, other labour market status, household income &
financial assets squared. Coefficient (standard errors in parentheses) are quoted, not conditional
marginal effects. **=1% level of significance; *=5% level of significance.
Unbinding Collateral Constraints I: Remortgaging, Changes in Loan-to-Value Ratio
        and Value of Unsecured Debt 1995-2000 (Probits; marginal effects)

     Specification: Probit          (1)          (2)           (3)            (3)
     LHS Variable: Prob.       ∆LTV ratio      ∆LTV        ∆LTV ratio     ∆LTV ratio
       of remortgaging         + unsecured   ratio*LTV     *unsecured     *unsecured
       1995-2000                   debt      ratio in 95    debt in 95   debt >£500 in
                                                                               95
     ∆LTV ratio1995-2000          −0.07**        -             -               -
                                   (0.02)
     Unsecured debt1995             0.01**      0.01**         -               -
       in £                       (0.002)      (0.003)
     ∆LTV ratio1995-2000            -         −0.006           -               -
     *LTVratio1995                             (0.004)
     ∆LTV ratio1995-2000           -             -            −0.03**          -
     *unsecured debt1995                                       (0.01)
     ∆LTV ratio1995-2000           -             -              -               −0.2**
     *unsecured debt>0.51995                                                    (0.05)
     H income1995 (£)            0.004**       0.004**       0.004**           0.004**
                                 (0.002)       (0.002)       (0.002)         (0.002)
     Employed=1                    0.08**        0.09**        0.09**           0.09**
                                   (0.03)        (0.03)        (0.03)          (0.03)
     Fin. Assets1995 (£)         −0.002*       −0.002*       −0.002*         −0.002*
                                  (0.001)       (0.001)       (0.001)         (0.001)

Notes: Coefficient (standard errors in parentheses) are dF/dX. N=2006. **=1% level of
significance; *=5% level of significance.
Specifications also include constant term and dummies for highest educational qualifications,
marital status in 1995 and change in marital status 1995-2000, number of children in 1995,
whether retired in 1995, and whether female head of household in 1995, number of benefits,
household income & financial assets sqd, whether smoker, saver, gender of HofH, other
labour market statuses
∆LTV ratio is the predicted change in the LTV ratio given the original mortgage value, and
excludes consequences of remortgaging.
House Prices and Total Debt: Changes in Total Debt and House Prices

            Change in unsecured debt limited to +/- £60,000 (excludes 2.4%)
            Change in house price limited to +/- £275,000 (excludes 1.37%)

                                        Non- Movers            Heckman            All (incl. movers)

 OLS Estimates                     (1)             (2)             (3)             (4)            (5)
 ∆total debt 1995-2000
 ∆self-reported house            -0.01           -0.01           -0.01         0.02**            0.01
 value1995-2000                 (0.008)         (0.008)         (0.008)        (0.008)         (0.008)
 ∆self-reported house              -             -0.05           0.003            -            0.12**
   value1995-2000                                (0.04)          (0.03)                         (0.25)
   *constrained1995
    350 observations

 ∆self-reported house               -           0.13**          0.09*              -             0.29
 value1995-2000                                 (0.05)          (0.04)                          (0.04)
   *constrained1995
    *unsecured debt
 1995>£500
   300 observations

 Age1995                        -0.54**         -0.48**        -0.52**         -1.65**          -1.4**
                                 (0.21)          (0.21)         (0.21)          (0.21)          (0.21)
 Age21995                       0.005**         0.005**        0.005**          0.01**           0.01
                                (0.002)         (0.002)        (0.002)         (0.002)         (0.002)

Notes: Coefficient (standard errors in parentheses). **=1% level of significance; *=5% level of
significance. Specifications also include controls as before plus constant term and dummies for highest
educational qualifications, marital status in 1995 and change in marital status 1995-2000, number of
children in 1995, whether retired in 1995, and whether female head of household in 1995. Note on
Heckman: 2566 observations (861 censored, 1705 uncensored).
Total effect of ∆debt is (% constrained) X (coeff on ∆debt)
                                        Impact of change in LTV on total debt by LTV
                                               (where unsecured debt >£500)

                                  0.5




           Coeff or % of sample   0.4




                                  0.3




                                  0.2




                                  0.1




                                   0
                                        0.95    0.9      0.85        0.8         0.75    0.7     0.65
                                                                LTV ratio 1995

                                                 Coeff                           +1 SE
                                                 -1 SE                           Coeff*% of sample
                                                 % of sample
              Interpretation and conclusion
•   We do find a ‘financial accelerator’ effect
•   But it only works through households substituting cheaper
    secured debt for unsecured debt if their house value rises
•   The more loosely we define a ‘collateral constrained household’,
    the smaller the effect
•   The macroeconomic effect on debt-financed consumption is
    ∆C/∆VH..
•   In our study this is (probability of being constrained) X (the effect
    of being constrained). This is roughly constant at 0.03.
•   This is much lower than Campbell & Cocco (0.08 to 0.12) and
    effect implied by other simulation-based studies
•   However, it is much closer to the effect implied by standard LCH
    models of house price ‘surprises’.
•   Conclusion: Don’t overstate the impact of volatile asset prices!

				
DOCUMENT INFO
Shared By:
Tags: what, loan, value, ratio
Stats:
views:79
posted:12/21/2008
language:English
pages:30