SITUATION

A German Landesbank engaged Devon to assess their $500M investment in a CDO2 issued by a leading investment bank.

Devon modelled the CDO based on market data and the post-closing substitution option on the reference pool and realized that the investment bank had made approximately $350M of P&L selling the $500m CDO to investors including the Landesbank.

The Landesbank hired a U.S. law firm to work alongside Devon and open a litigation

The law firm’s i-Connect document system was inadequate for the volume and complexity of the document productions

Devon requested and the client agreed that Devon should build a platform and host a copy of the discovery documents

DEVON’S EXPANDED APPROACH

The investment bank sought to “bury” the case in paper, producing a huge volume of unstructured, unsorted and duplicative documents and spreadsheets

The deluge proved an opportunity for Devon to build a bespoke platform, Occam, tailored to the needs of the case, in particular the capacity to read spreadsheets and trace emails and other communication through the organization

Devon built matching algorithms and other bespoke tools to sort and tag the productions

Using the Occam platform, Devon was able to review incoming documents in a matter of days rather than weeks or months

Devon brought the financial modelling in-house, reducing time and cost

The client’s law firm insisted on labor intensive reviews of all documents by associates; Devon used our data platform to review and sort document templates to filter for those that were actually relevant

Devon not only assessed the tranches the Landesbank bought, but also reverse engineered the CDO and each of the seven other CDOs in the series that the investment bank issued.

PROVEN RESULTS

Devon drafted the deposition questions that led to proof of the investment bank’s bad intent selling the CDO to the Landesbank. In depositions, the investment bank conceded that it had taken roughly $350M in P&L on the $500M investment

Devon requested and the client agreed that Devon should build a platform and host a copy of the discovery documents. This project resulted in the first version of OCCAM Devon’s E- discovery platform.  

THIS IS THE TWO CASH FLOWS ANALYSIS TO VALUE A CDO

If the arranging bank reports positive value for the CDS on its balance sheet, it is expecting that the investors will lose money.

*IFRS rules often require the investment bank to retain the net day1 P&L in reserves rather than immediately showing in the total NPV after costs in the P&L. The Discounted Sum of CDS Cash Flows = Swap Value.

This graphic does not include the initial principal legs of the cash flow on the CDO purchase or the anticipated interest earned on the initial principal leg which is often retained within an SPV to make it bankruptcy remote.

 

BANKS SOLD THE TRANCHES TO INVESTORS ON THE RATINGS-BASED APPROACH, WHICH IMPLIED THAT THE SUBORDINATION ADEQUATELY PROTECTED THE INVESTORS’ TRANCHES FROM LOSS

Devon’s stochastic model evaluation was the same methodology that the bank itself would have used. With that information, Devon traced the substitutions and valuations back to the bank’s internal communications and financials.

MODELLING THE CDOS TRUE VALUE PRODUCED EVIDENCE OF SCIENTER

  • Devon calculated the NPV of both CDS payments (Protection and Coupon) to show that from day one the investment bank expected the investor to lose money on the investment including increased risk to the investor through substitutions.
  • Focusing on the CDS cash flows strips away complexity, making plain the investment banks acted wrongly and knew that the CDO tranches posed a much higher risk of loss to investors than implied by the rating. Additionally, poor substitutions significantly increased both the risks and losses to the investors and the profit to the investment bank.

IN ADDITION DEVON REVERSE ENGINEERED THE PERFORMANCE OF OTHER CDOS IN THE SERIES

In discovery, Devon requested performance information on the bank’s both earlier and subsequent CDOs.

All of the CDOs exhibited similar poor asset selection for the reference pool as well as eroding credit quality.