Studies in debt valuation adjustments

Lin, Wen (2021) Studies in debt valuation adjustments. PhD thesis, UNSPECIFIED.

[img]
Text (2021WenLinphd)
2021WenLinphdv2.pdf - Published Version
Available under License Creative Commons Attribution-NonCommercial-NoDerivs.

Download (2MB)

Abstract

This thesis consists of two self-contained studies in Debt Valuation Adjustments (DVAs). The first study is motivated by the debate about the introduction of the Fair Value Option for financial Liabilities (FVOL) and the requirement to recognize and separately disclose DVAs in financial statements. This study investigates what we can learn regarding own credit risk from DVAs. Using a sample of U.S. bank holding companies that adopt the FVOL, we show that DVAs generally cannot be explained by the same factors that explain contemporaneous changes in the credit quality of these institutions. These results may be driven by the opportunistic use of the FVOL or the superior ability of managers to estimate own credit risk. Further tests indicate that DVAs for fair value Level 3 reporters can explain future changes in credit risk, providing support for the latter explanation. The second study compares the reported Debt Valuation Adjustments provided by managers with the estimated DVAs based on market information. To obtain the estimated DVAs we use two structural credit risk models: the Merton (1974) model and the Leland (1994) model. We find that the private information contained in the reported DVAs causes a significant deviation of the estimated DVAs from the reported DVAs. This deviation is more pronounced for the banks with high volatile creditworthiness and gets better for the banks with stable credit standing. Findings suggest that the reported DVAs reflect more private information on credit risk when the economy is volatile rather than stable. In addition, the comparison of estimation errors shows that the Merton model outperforms the Leland model with regard to the estimation of DVAs over the sample period, suggesting that the incorporation of additional information in structural models does not improve the performance of pricing DVAs.

Item Type:
Thesis (PhD)
ID Code:
155703
Deposited By:
Deposited On:
08 Jun 2021 11:05
Refereed?:
No
Published?:
Published
Last Modified:
27 Nov 2021 08:02