Two Essays in Financial Intermediaries Open Access

He, Ai (Spring 2019)

Permanent URL:


This dissertation investigates the activities of financial intermediaries and financial institutions in the credit markets. In the first essay, I studies how exogenous shocks on a subset of borrowers constrain bank lending and affect real economic activities of non-shocked firms. I separate a loan supply effect from a loan demand effect by identifying borrower-level shocks with the occurrence of major U.S. natural disasters. Financially constrained banks reallocate post-disaster lending by restricting credit supply as well as increasing loan pricing to non-shocked firms but prioritizing the disaster firms with which they have strong pre-disaster relationships. I find one dollar of additional lending to disaster firms is associated with 11.5 cents of decline of the same bank's lending to non-shocked firms. Non-shocked firms' pre-disaster dependence on such banks for financing accounts for economically significant reductions of their total loan borrowing, investment, profitability, and sales-growth in the year following a natural disaster. Consistent with frictions deriving from asymmetric information, the real outcome losses are larger for financially constrained firms. In the second essay, I explore the nature and impacts of relationship lending in shadow banks by analyzing a cross-holding relation (CHR) between financial firms who mutually hold each other's debt through their own affiliated money market funds (MMFs). Using novel security-level holdings data, I show that non-European financial firms increased their MMFs' portfolio weights on bilateral-connected European financial firms after Moody's downgrade review of European banks in mid-2011, a special period when MMFs generally reduced their exposure to European issuers to avoid further redemption. I provide evidence that this bias represents reciprocity between the bilateral-connected financial firms. In return, during the same period, the European financial firms accepted more unsecured than secure debt from their bilaterally-connected non-European partners through their affiliated MMFs. Issuer- or fund-characteristics do not explain the results. CHR is also found to affect MMFs after the 2013 Dodd-Frank stress test in the same pattern. A further test shows a spillover effect on issuers unconnected with MMFs due to MMFs' tilt to connected issuers.

Table of Contents

1 Introduction 2

2 Sample Construction 8

2.1 Data 8

2.1.1 Corporate Loans 8

2.1.2 Bank Characteristics and Firm-Level Information    9

2.1.3 Major Natural Disasters 10

2.1.4 Other Datasets 10

A. Bank Branches and Deposits 11

B. Supplier-Customer Links  11

C. Geographic Dispersion of Borrowers’ Business Operations 11

2.2 Measures of Relationships 12

2.3 Sample Characteristics 12

3 Identification Strategy 14

3.1 Classify Borrowers  14

3.2 Natural disasters as Negative Demand Shocks  14

3.3 Exposure to Natural Disasters through Disaster Firms  15

3.3.1 Banks’ Pre-Disaster Exposure to Disaster Firms 16

3.3.2 Connected Firms’ Pre-Disaster Exposure to Disaster Firms 16

3.4 Other Identification Concerns 17

4 Methods and Results 18

4.1 Lending Spillovers on Connected Firms 19

4.1.1 Trace out capital flows: the firm-bank level lending change 19

4.1.2 The loan-level evidence 23

4.1.3 Financially constrained banks 25

4.2 Real Outcomes of Connected Firms 27

4.2.1 Firm-level total loan borrowing 28

4.2.2 Post-disaster economic activities 29

4.2.3 Financially constrained firms 31

5 Conclusion 31

1 Introduction 34

2 Background and Hypotheses Development 40

2.1 CHR in the U.S. MMFs Market 40

2.2 The European Bank Crisis in 2011 41

2.3 Hypothese Development 42

3 Data and Summary Statistics 44

4 Methodology and Empirical Results 48

4.1 Construction of CHR Measures 48

4.2 Tests of Hypothesis One 49

4.2.1 Univariate Analysis 49

4.2.2 Multivariate Analysis50

4.2.3 MMF Flows 54

4.2.4 Securities Comparison 56

4.3 Tests of Hypothesis Two 56

4.3.1 Univariate Analysis 57

4.3.2 Multivariate Analysis 58

4.4 Spillover Effects 59

5 Conclusion 61

Appendices 62

A Essay One: Variable Definitions 62

B Lead Lenders in Syndicated Loans 66

C Loan-level Tests of Disaster Firms 67

D Essay Two: Variable Definitions 68

E Investment and Issuer Categories 70

F Financial Firms Serving Dual-Roles in the MMFs Market 71

G Bilateral Connection and Past Relationship 72

H Multivariate Analysis of Securities Comparison 74

I The Dodd-Frank Banking Stress Test 75

References 77

Tables 83

Figures 118

About this Dissertation

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Last modified

Primary PDF

Supplemental Files