QuarcHome
Model catalogueBack to the model index

Intensity Gamma and Stochastic Business Time

Calendar time passes at the same speed for all borrowers. In a stochastic business-time model, the clock driving defaults can move faster in some scenarios than others. A shared clock can cluster defaults even when marginal default probabilities remain similar. Comparing tranche losses with a one-factor Gaussian baseline makes the effect of the dependence assumption easier to see.

How it works

Intensity Gamma introduces a stochastic business-time clock. When a shared clock advances quickly, the effective credit environment can worsen for several names together. Marginal credit curves and the portfolio dependence parameters provide separate calibration targets. The random clock creates a mechanism for clustering loss without fixing a single Gaussian dependence structure.

Worked example

Consider two loans with equal marginal probability of default by year end. Compare a model with independent clocks against a common-clock version re-calibrated to the same single-name probabilities. Conditional on the common clock, calculate each name's default probability and then average the product. Clock volatility can increase the joint probability while leaving the expected default count unchanged.

Validation and limitations

Verify calibration to individual survival curves before comparing portfolio tails. Test the fitted common-clock parameters against different maturities and tranche attachments. If models fit the same base tranche quotes, compare out-of-sample events or portfolio payoffs and report the dependence sensitivity.

Further reading