Structured Credit: CDO/CLO Tranche Pricing

Structured credit products were at the centre of the 2008 financial crisis, yet the models that price them remain indispensable on credit desks, at rating agencies, and in CLO management today. This course starts where the textbooks stop — with a fully implemented tranche pricer that you build yourself, from the Gaussian copula through to a working CLO waterfall.

You will implement two complementary approaches: the Vasicek large-pool approximation, which gives you an analytical loss CDF for a homogeneous infinite portfolio in a single closed-form expression, and the Andersen-Sidenius-Basu recursive algorithm, which handles real heterogeneous pools exactly via generating functions. Both are used in practice; understanding both gives you the ability to choose the right tool for each problem.

The course culminates in base correlation — the market's way of quoting CDO tranche prices — and a simplified CLO waterfall that shows how interest and principal diversion tests protect senior noteholders. Every calculation is grounded in the original academic literature, from Li's 2000 copula paper to the ASB 2003 extension, so you leave with both working code and the theoretical foundation to defend every line of it.

You build a CDO tranche pricer end to end. Starting from a pool of reference names with individual default probabilities, notionals and recovery rates, you calibrate the Gaussian copula, compute the conditional default probability given the systematic factor, and integrate numerically over the factor distribution to recover the full unconditional loss distribution of the portfolio.

With the loss distribution in hand, you price equity, mezzanine and senior tranches by computing expected tranche losses and converting them into par spreads. You then invert this process: given market-quoted tranche prices, you extract the base correlation for each tranche and build the base correlation curve — the industry-standard representation of the correlation smile.

Finally you implement a simplified three-tranche CLO waterfall with interest and principal diversion tests, run correlation and PD stress scenarios, and compare large-pool approximation prices against ASB exact prices to understand where each model is sufficient and where it breaks down.