Centre for Finance, Credit and Macroeconomics (CFCM)

CFCM 18/02: Preliminary credit ratings and contact disclosure

Preliminary credit ratings


A recent amendment to the European Regulation on credit rating agencies (CRA) requires CRAs to disclose any issuers' request of initial reviews (ie preliminary ratings). This theoretical paper focuses on the widespread practice of obtaining an initial indicative evaluation prior to purchasing the published rating. The confidentiality of these preliminary ratings (or indicative ratings) is widely regarded as a problem because it can bias ratings and causes poor information provision.

The author's main question is whether having to disclose preliminary ratings affects the CRAs' reputation and how their incentives might change in response to this different environment. In order to understand the theoretical consequences of policy proposals, the author develops a game theoretic model of the interaction between the issuers of financial products and the credit rating agency. The latter wants to create and maintain a reputation for accurate ratings. Two different regimes are considered; one characterised by no disclosure of the contact between the players while the other features compulsory contact disclosure.

The author finds that the CRA benefits from the opaqueness of the market and identifies a trade-off between the fee and the CRA's reputation. This trade-off is such that when the project out for evaluation is likely to be good, the CRA issues a good preliminary rating because the risk of losing reputation is extremely low whereas when the project is likely to be bad the CRA prefers to issue a bad preliminary rating avoiding to risk reputation. When there is disclosure of the contact between the CRA and the entrepreneur, the CRA issues more good preliminary ratings then when there is no evidence of preliminary contact. Disclosure results in a lower probability of stopping high quality projects but also in a higher probability of financing low quality ones.

There are three main implications. First, consistently with a strand of the literature, disclosure is not necessarily welfare improving. Second, disclosure policies and regulation cannot be a one off decision, they require tailoring and timely revisions. Thirdly, for disclosure to be effective, the rating process needs to be as accurate as possible.

Download the PDF of this paper


Marta Allegra Ronchetti

View all CFCM discussion papers | View all School of Economics featured discussion papers


Posted on Wednesday 31st January 2018

Centre for Finance, Credit and Macroeconomics

Sir Clive Granger Building
University of Nottingham
University Park
Nottingham, NG7 2RD

+44 (0)115 951 5620