CITIGROUP INC

Citi estimates expected credit losses for its quantitative component using (i) its comprehensive internal data on loss and default history, (ii) internal credit risk ratings, (iii) external credit bureau and rating agencies information and (iv) reasonable and supportable (R&S) forecasts of macroeconomic conditions.

For its consumer and corporate portfolios, Citi’s expected credit losses are determined primarily by utilizing models that consider the borrowers’ probability of default (PD), loss given default (LGD) and exposure at default (EAD). The loss likelihood and severity models used for estimating expected credit losses are sensitive to changes in macroeconomic variables, including unemployment rate, real GDP and housing prices, and cover a wide range of geographic, industry, product and business segments.

In addition, Citi’s models determine expected credit losses based on portfolio characteristics, including loan delinquencies, changes in portfolio size, default frequency, risk ratings and loss recovery rates, as well as other credit trends.