Dec 15, 2010

Advanced Valuation Methodologies for CLOs

The Exploration of Traditional Methods, Contrasted with Advanced Methodologies Based on Monte-Carlo Methods

With the demand for Collateralized Loan Obligations (CLOs) reviving, Numerix recently hosted a webinar presented by industry expert Benoit Fleury, Head of Financial Engineering and Products at R2 Financial Technologies to discuss the different valuation methodologies available today for CLOs.

There is a time and a place for everything; and, indeed traditional CLO valuation methods have been effective to a certain degree, but the financial crisis has clearly shown that traditional valuation methods alone have many shortcomings?ranging from the utilization of ineffective structured finance ratings, to inconsistent IDs from multiple data sources, to the scarcity of available data.

So here’s the deal… Pricing methods for structured finance instruments can be essentially divided into two classes:1 ) bond models (single- scenario) and 2) stochastic methods. In addition, the Net Asset Value is a third approach commonly used to monitor these instruments. Read on to discover important insights into today’s best practices for CLO modeling and valuations.


What Are the Key Shortcomings of the Bond Model?


•       Bond models generally rely on credit ratings as determinants of yields (spreads) and risk

–      Basing valuations largely on structured finance ratings, as has been common practice, is ineffective, at best, and potentially dangerous (as the crisis has shown!)

–      Even when computed correctly, they only reflect one “statistic” (PD or EL) of a complex loss distribution

•       Fails to account for all relevant market data

–      Underlying assets may have liquid prices, which are often overlooked with this approach

–      Fails to employ important information derived from synthetic markets (CDX tranches)

•       Fails to capture the option-like payoff profile of some of the bonds/tranches ( Specifically, the lower portion of the capital structure)

•       Can easily lead to pricing inconsistency across deals

–      Not comparable across asset classes (Bond, ABS, CDO) or even structure


The Need for Second Generation Models


In light of the above, Mr. Fleury highlighted the need for ‘second generation’ models, including the development of “practical” models for structured credit, which are outlined below:


1) Net Asset Value (NAV) Model: Produces a liquidation value for each tranche, using market prices of the underlying securities 


2) Monte Carlo Model: Simulates the price of the bonds across a large number of Monte-Carlo scenarios to capture the option-like payoff profile of the structure


Collateral Market Value Approach & NAV

•       NAV reports are used extensively by CLO investors to screen potential investment vehicles

•       NAV statistics measure relative “protection/cushion” implied in a tranche

•       Collateral market prices are taken (or calculated) to generate market values for underlying


“Easier Said, Than Done:” The NAV Approach & Key Challenges ‘in Practice’

•       Key to producing NAV statistics: Reliable, up-to-date information on the structure and collateral pool market values

•       “Easier Said, Than Done”

–      Four types of assets: Loans, Bonds, CDO/CLO and Equity

–      Prices (or ‘dealer quotes’) from multiple sources: (e.g. Bloomberg and LoanX: They can be inconsistent, with holes in the coverage)

–      Inconsistent IDs from multiple sources (Intex, LoanX, etc.)

–      Much of the work is typically performed manually ( inefficient, time consuming and potential operational risk)

•       Requires robust infrastructure (in addition to the analytics):

–      For example, mapping tools (e.g. Intex – LoanX) and other tools (such as rule-based algorithms) needed to find and store the most relevant price across a source


What Are the Shortcomings of the NAV Approach?

•      Need a very good infrastructure and detailed workflow procedures to make them work ‘in practice’


•       NAV does not perform a direct modeling of cashflows; and as a result, it is more of a relative value indicator than real valuation approach


•       Fails to capture the option-like payoff profile of some of the bonds/tranches


Stochastic Valuation Methods As Best Practice

      Key Strengths of Weighted Monte Carlo Approach 

•       Structured, multi-scenario approach – effectively uses advances in credit models, Monte Carlo methods and CDO analytics

–      Ability to model waterfall details and incorporate a full bottom-up approach

–      Factor models for credit risk – characterize “correlations”

•       Basic idea is simple: set of scenarios under which instruments are consistently valued

–      Imply “risk-neutral” distribution (process) for underlying systematic risk factors

–      Effective use of market (pricing), as well as fundamental credit information

–      Produces consistent, arbitrage-free prices

–      Integrated view of cash and synthetic products

•       Explicit modeling of key risks: Credit (default, LGD, spread), prepayment and market risk

•       Explicitly captures systematic risk, correlations and concentrations


Summary: Implied Factor Models and WMC Approach

•       Structured, multi-scenario approach ( as indicated above)

•       Consistent valuation and risk through the use of common scenarios

–      Deals (tranches) based on the same collateral pool, bespoke portfolios, and across asset classes (synthetics, cash; CLO, CDO, CDO^2, MBS)

–      Transparent loss distributions, parameters, sensitivities and risk metrics

•       Integrated view of cash and synthetic products

•       Explicit modeling of key risks

–      Credit (default, LGD, spread), prepayment and market risk

–      Explicitly captures systematic risk,  correlations and concentrations

•       “Arbitrage-free” – uses market (pricing), as well as fundamental credit information


CLO Modeling and Valuation: Conclusions & Best Industry Practices

When it comes to best practices for CLO Modeling and Valuations (and learning from our past mistakes!), it is wise to adhere to the best practice strategy outlined below:

•       Collateral and Structure modelling – full bottom-up approach

–      Reasonable proxies, waterfall and comparables (where information is missing)

–      Handle cash and synthetic deals consistently

•       Valuation Models

–      NAV as a monitoring and trading tool

–      Stochastic (Monte Carlo) – primary valuation and risk model

–      NPV model (bond + spreads) – secondary model (quoting, fast computation)

•       Valuation model inputs

–      Prior collateral model (estimated): credit/prepayment model for pool

•       Distributions of PDs, LGDs, prepayments and correlations

•       Ideally loan level/buckets, but possible at pool level – adjustments for credit quality and concentrations

•       Based on historical data and loan or LCDS/CDS prices( where available)

–      Benchmark prices ideally from actual trades or reasonable quotes/indices


If you have further questions or would like more information about Numerix/R2 Financial Technologies’ CLO or other valuation solutions, please contact

Blog Post - Apr 13, 2010

LIBOR Market Models for Smile-Dependent Exotics

Need Assistance?

Want More From Numerix? Subscribe to our mailing list to stay current on what we're doing and thinking

Want More from Numerix?

Subscribe to our mailing list to stay current on what we're doing and thinking at Numerix

Subscribe Today!