|
Moody’s KMV develops
models that facilitate active management of credit portfolios. Our models accurately
calculate stand-alone risk and return measures of instruments in a portfolio,
as well as portfolio-referent risk of those instruments. These risk and return measures
are combined to yield an extensive set of economic performance metrics, including portfolio return,
expected loss, unexpected loss, economic capital, and expected shortfall. Our portfolio framework supports
a wide range of credit investments and contingencies, including bonds, equities, term loans, revolving credit lines,
credit derivatives, and structured instruments. In addition, we maintain open and transparent design principles,
allowing clients to utilize their own data and third-party research in the portfolio model.
To compute risk and return measures of single-name instruments, we offer several valuation methods, including valuation based on book, user-input price or spreads, credit curves, and lattice modeling. The proprietary lattice valuation framework accounts for the dynamic nature of the credit quality of a reference entity over time, which enables accurate modeling of instruments with complex fee structures and credit contingencies. The lattice framework models migration between credit states using the Moody’s KMV Empirical Credit Migration model or using a user defined credit migration model (e.g., ratings-based or custom migration). Expected spread, expected loss, and unexpected loss of individual instruments are computed from the value distribution at the horizon date.
To measure portfolio concentration and diversification effects, we created a global factor model that provides pair-wise asset correlations for all counterparties within a portfolio. Our approach imposes structure on the correlation of asset returns by decomposing a counterparty’s risk into a systematic component (for example, an industry-specific component and a country-specific component for a large corporate borrower), and an idiosyncratic component. This structure enables robust estimates of pair-wise correlations, which we update annually to reflect changing market dynamics for publicly listed firms. In addition, we develop models that provide mid-market correlations for private firms, as well as correlation parameterizations for a wide range of other asset classes, such as retail instruments and commercial real estate. Structured instruments are then combined to form the distribution of returns for a credit portfolio using Monte Carlo techniques. The portfolio value and loss distributions are analyzed to provide different views of portfolio risk, including economic capital, stress testing, and risk attribution.
|
