Bank Loan Renewal by owu16919

VIEWS: 44 PAGES: 21

More Info
									Banking Relationships and
   Conflicts of Interest

              Wook Sohn

  KDI School of Public Policy and Management
                 MBA Program
                 Seoul, Korea




        FDIC/JFSR Conference
            Research Questions

Highlight conflicts of interest in banking relationships.

Key feature is pre-existing relationships (or financial
claims) between banks and borrowers before the banks’
lending decisions.

Approach
  Part 1: Loan announcement effect in the stock
           market. (empirical investigation)
  Part 2: Loan announcement effect in the stock
           market. (theory)
  Part 3: Banks’ lending decisions (empirical).
               Event Description

Closed Banks       Borrowing firms   Acquiring Banks

     C1                                     A1
                         Prior
                     Relationship
     C2               (Type P)              A2


     C3                                     A3
                        No Prior
                      Relationship
     C4                (Type N)
                                            A4


     C5
                                            A5



                    16 Other Banks
         Features of the Event
Exogenous selection and matching and the transfer
of loans in their entirety.

No personnel were transferred from the closed
banks to the acquiring banks (purchase and
assumption).

Distinctions between the closed banks and the
acquiring banks.

 - BIS ratios: 5.32% vs. 9.64%
 - Non-performing loans: 9.08% vs. 3.01%
 - Market shares: 6.95% vs. 30.96%

Firms’ pre-existing relationships with the acquiring
banks are identified.
               Event Description

Closed Banks       Borrowing firms   Acquiring Banks

     C1                                     A1
                         Prior
                     Relationship
     C2               (Type P)              A2


     C3                                     A3
                        No Prior
                      Relationship
     C4                (Type N)
                                            A4


     C5
                                            A5



                    16 Other Banks
        Estimation of CARs


CARs                Mean         Median [pos./neg.]
  (-7,-2)   -5.79**    (-2.48)     -4.39 [37/81]***
 (-1,+1)    -4.85**    (-2.94)     -7.85 [32/86]***
(+2,+5)     -8.81***   (-4.62)     -8.31 [24/94]***
 (-7,+5) -19.45***     (-5.65)   -20.56 [22/96]***
(+6,+51)    22.39***   (3.46)     27.09 [74/44]***
(-1,+51)     8.73       (1.26)    10.92 [65/53]*
               Event Description

Closed Banks       Borrowing firms   Acquiring Banks

     C1                                     A1
                         Prior
                     Relationship
     C2               (Type P)              A2


     C3                                     A3
                        No Prior
                      Relationship
     C4                (Type N)
                                            A4


     C5
                                            A5



                    16 Other Banks
           CARs for subsample



           Prior Relationship
  CARs            0            (0,1)           1
           (Type N)    (mixed type)     (Type P)   t test
      N          45               28         45
 (-1,+1)       -5.62            -7.10     -2.68     1.61
(-1,+51)       13.26            11.29      2.62    -1.28
                Explanatory Variables

Bank-firm      Exposure to        Firm-specific     Firm           Bank-
relationship   event              characteristics   ownership      pair
                                                                   dummy
 Prior rela-      #Closed           Age              Chaebol        Bank1
 tionship        banks /            Size (asset)     firm           Bank2
 Locational      #Lending           Sales            Proportion     Bank3
 advantage       banks              growth           of largest     Bank4
 Main            Loan_closed        Profit /         shareholder    Bank5
 creditor       / Loan_all          interest         Small
 bank            banks              Bond             share-
                 Collateralized     finance          holders
                 loan_closed        Equity           Foreign
                 banks              finance          share-
                                    Loan finance     holders
            OLS regressions of CAR (-1,+51)

                                          CAR(-1,+51)             CAR(-1,+51)

Prior Relationship                     -0.174**(-2.012)          0.436*(1.719)
Prior Relationship* Loan Finance                             -1.147***(-2.697)
Locational Advantage                        0.196(1.513)           0.176(1.498)
Main Creditor Bank                        -0.012(-0.084)           0.026(0.176)
Log_Age                                     0.249(1.629)         0.262*(1.828)
Log_Size                                  -0.014(-0.142)           0.007(0.080)
Sales Growth                              -0.154(-1.081)         -0.145(-1.141)
Profit/Interest                             0.007(0.333)           0.014(0.753)
Bond Finance                           -0.566**(-2.155)      -0.940***(-3.300)
Equity Finance                        -2.261***(-4.862)      -2.639***(-5.441)
Chaebol                                   -0.161(-1.593)         -0.147(-1.537)
#Closed/#Lending Banks                    -0.136(-0.213)         -0.140(-0.240)
Loan_Closed/Loan_All                      -0.193(-0.604)         -0.160(-0.540)
Collateralized Loan                         0.034(0.429)           0.020(0.286)
Bank 1                                      0.067(0.851)           0.074(1.042)
Bank 2                                   -0.176*(-1.684)      -0.208**(-2.050)
Bank 3                                      0.009(0.088)         -0.023(-0.221)
Bank 4                                    -0.140(-1.431)         -0.136(-1.416)
Intercept                                   0.527(0.559)           0.552(0.715)
F test                                        3.580***               3.820***
R-square                                          0.333                  0.368

Three firm ownership variables are included in the regressions.
             Probit regressions of the selection equation
                                                     [1]                 [2]                 [3]
Prior Relationship                       2.312***(4.350)   2.307***(3.692)     2.719***(4.007)
Market/Book Value                         1.502**(1.966)      1.925*(1.937)       2.056*(1.919)
Prior Relationship*Market/Book Value   -4.929***(-3.325)   -4.599**(-2.365)    -5.257**(-2.456)
Locational Advantage                                          0.851*(1.811)       0.882*(1.747)
Main Creditor Bank                                           -0.221 (-0.437)     -0.174 (-0.297)
Log_Age                                                        0.091 (0.122)     -0.124 (-0.160)
Log_Size                                                       0.354 (0.948)       0.525 (1.174)
Sales Growth                                                   0.300 (0.549)     -0.040 (-0.071)
Bond Finance                                                 -0.177 (-0.200)       0.277 (0.295)
Equity Finance                                              -2.994*(-1.674)      -2.650 (-1.397)
Chaebol                                                      -0.037 (-0.095)     -0.140 (-0.310)
Loan_Closed/Loan_All                                          4.101*(1.909)       3.452*(1.676)
Collateralized Loan                                          -0.071 (-0.273)     -0.243 (-0.839)
Bank 1                                                                             0.508 (1.366)
Bank 2                                                                             0.348 (0.893)
Bank 3                                                                           -0.370 (-1.145)
Bank 4                                                                             0.082 (0.209)
Intercept                                -0.372 (-1.276)     -3.562(-1.010)      -5.688 (-1.439)
Chi-square                                    20.850***           25.030**           47.810***
Pseudo R-square                                   0.167               0.277               0.343

 Three firm ownership variables are included in the regressions.
         Heckman estimation of CAR(-1,+51) for the subsample

                                                  [1]                   [2]          Full sample
Prior Relationship                      0.545*(1.729)         0.513*(1.673)       0.486**(2.104)
Prior Relationship* Loan Finance    -1.073**(-1.966)        -0.933*(-1.674)    -1.237***(-3.095)
Locational Advantage                     0.200 (1.241)        0.283*(1.816)         0.156 (1.430)
Main Creditor Bank                     -0.023 (-0.118)       -0.025 (-0.125)        0.025 (0.172)
Log_Age                                  0.330 (1.276)         0.365 (1.543)       0.255*(1.853)
Log_Size                               -0.130 (-1.325)       -0.093 (-0.929)        0.015 (0.188)
Sales Growth                             0.037 (0.201)         0.010 (0.058)      -0.129 (-1.099)
Profit/Interest                        -0.047 (-0.894)       -0.044 (-0.809)        0.014 (0.781)
Bond Finance                          -0.774*(-1.874)       -0.788*(-1.891)    -0.973***(-3.516)
Equity Finance                     -3.023***(-4.255)     -2.941***(-4.268)     -2.653***(-5.870)
Chaebol                                -0.158 (-1.362)       -0.131 (-1.229)      -0.144 (-1.551)
Bank 1                                                         0.111 (1.217)        0.069 (1.038)
Bank 2                                                       -0.148 (-1.308)    -0.217**(-2.283)
Bank 3                                                         0.059 (0.497)      -0.043 (-0.485)
Bank 4                                                       -0.141 (-1.121)      -0.143 (-1.573)
Inverse Mills ratio                    0.308 (0.687)           0.554 (1.128)
Intercept                              1.336 (1.465)           0.820 (0.948)       0.483 (0.695)
F test                                      2.880***              3.060***             4.670***
                                                                     0.385
R-square                                        0.326                                      0.371
No. of firms                                       82                    82                  118
         Main Results of Part 1
Overall effect on firm value in the stock market:
positive

   Loss of the relationship-specific advantage with the
    closed banks is outstripped by the gain from good
    quality of the acquiring banks.

Effect of the pre-existing relationships on the
positive valuation: negative

   Informational advantage from the pre-existing
    relationships is more than offset by banks’ incentives
    to misuse the information.

   The larger the size of pre-existing loan, the more
    negative the effect of pre-existing relationships.
           Overview of Theory in Part 2
     Market’s valuation of bank’s lending decisions


Firms with prior relationships         Firms with no prior relationships
          (Type P)                                 (Type N)

  whether to renew loans to good         whether to invest in costly screen-
 firm or bad firm (potentially more     ing to identify good firm (potentially
 informed decisions).                   less informed decisions).

  investor’s concerns arise from         investor concerns arise from
 bank’s incentive to renew bad firm.    bank’s incentive not to screen.

  bank’s loan renewal does not           bank’s loan renewal does signal
 signal good borrower quality           good borrower quality
 when bank holds large bad loans        when the screening costs are low.
       Loan Announcement Effects: Summary

Ex ante bank    Equil. strategy for     Market valuation Market valuation
reputation h1   Type P                   to loan renewal to loan renewal
                (Type N)                to Type P at t=1 to Type N at t=1
High            Renew good firm
                                              V (h1 )  kL.                     V (h1 )  kL.
                (Screen)
Intermediate    Mixed strategy:
                Renew good firm                         h1 ( kL) 2                     eh1 ( kL) 2
                                        V ( h1 )                      .   V ( h1 )                   .
                                                     (1  e)(1   ) D                      4c
                with a prob.
                (Screen with a prob.)
Low             Renew bad firm
                                                                                          1  h1
                                                                             V ( h1 )           kL .
                (Not screen)                  V (h1 )  0                                   2
                     Loan Announcement Effects

2.5




 2


                   Value of Type P firms
                                           Value of Type N firms


1.5




 1




0.5




      Direction of increase in D                    Direction of increase in c
 0
          Motivation of Part 3

Firm’s abnormal stock returns to estimate the net
gain for the firm may not tell the whole story .

 For example, the market’s reactions to the event
  may be inconsistent with the actual behavior of loan
   officers of the acquiring banks.

Examine directly how the pre-existing relationships
between banks and borrowers affect the banks’
lending decisions.

 whether the lending relationship is maintained.
 how the size of loans changes.
    Random effect panel regressions of
   changes in loan size for the subsample
                                         [1]             [2]          [3]
 Prior Relationship                -0.022***      -0.015***    -0.015***
                                    (-4.291)        (-2.818)     (-2.889)
  Log_Loan_All                                                -0.122***
                                                                 (-9.124)
 Prior Relationship*                                            0.126***
         Log_Loan_All                                           (6.370)
 Collateralized Loan_Acquiring                       -0.001       -0.001
                                                    (-0.542)     (-0.546)
 Collateralized Loan_Closed                          0.005*      0.005**
                                                    (1.697)      (2.013)
 Constant                             -0.017        -0.099*       -0.094
                                    (-0.290)        (-1.760)     (-1.547)
 Wald (chi2)                      114.97***        43.09***    365.82***
 Log-likelihood                      661.74          602.91       606.27

All other control variables are included in the regressions.
                Main Results
Banks tend to continue relationships with firms that
have prior relationships even when their market
values are lower.
 Evidence for conflicts of interest

Bank is aggressive in expanding loans to firms that
have no prior relationships once the new
relationships are continued.


 Value of bank relationship to its client firms
  decline over time.
 Bank quality does not necessarily conveys risk
  classes of its client firms.
                Contributions


Investigates in great detail the effect of pre-existing
relationship between banks and borrowers
on banks’ lending decisions and market’s reactions
to the lending decisions.

Sheds light on the fundamentals of bank-borrower
relationships,

 especially the dark sides of banking relationships:
  conflicts of interest.
             Policy Implications

Underscores the importance of the specific mechanisms
employed to replace failed banks – the liquidation of
banks followed by transfers of their loans to better
banks can make client firms better off.

Mitigates the policy-maker’s concerns about potential
negative effects of banking sector restructuring on the
values of sound client firms of failed banks.

Suggests that the intensity of firms’ pre-existing
relationships with acquiring banks is important in
understanding how successful a bank consolidation is
in speeding up the resolution of financially distressed
firms.

								
To top