U.S. Private Label and European Residential Mortgage-Backed Securities

Transcription

1 U.S. Private Label and European Residential Mortgage-Backed Securities Evaluation Methodology Interactive Data offers daily evaluations and related data for U.S. Private Label and European residential mortgage-backed securities (RMBS), including Prime, Jumbo, ALT-A (Fixed and ARM) and Subprime Coverage Interactive Data provides evaluations and related data for fixed- and floating-rate U.S. non-agency/government sponsored enterprise (referred to as Non-Agency or private label ) and European residential mortgagebacked securities (RMBS) issued by private institutions. Covered tranche types include, but are not limited to: Senior Tranches Junior Tranches NAS (Non-Accelerating Senior) Tranches Floaters and Inverse Floaters* (LIBOR, CMT, COFI, and Prime) IOs (Interest Only)* and IOettes* POs (Principal Only)* Z-bonds (accrual bonds) including Z-PACs and Jump-Z bonds * Does not include Private Label RMBS backed by ARMs and Subprime. Evaluation Methodology Our teams of experienced fixed income evaluators and methodologists closely monitor the structured product markets, interest rate movements, new issue information, and other pertinent data. Interactive Data obtains and applies available direct market color (including trades, covers, bids, offers and price talk) along with market color for similar bonds. Evaluations of tranches (non-volatile, volatile, or credit sensitive) are based on Interactive Data s interpretation of accepted global market modeling, trading, and pricing conventions. As input to our model, Interactive Data uses spreads and other information solicited from market buy- and sell-side sources, including primary and secondary dealers, portfolio managers, and research analysts. To determine a tranche evaluation, Interactive Data: 1. Generates cash flows for each tranche 2. Determines the applicable benchmark yield 3. Applies market based tranche specific spreads to determine yield Incorporates deal collateral performance and tranche level attributes, as available, and market color, to determine tranche-specific spreads to adjust the benchmark yield. All U.S. Private Label and European RMBS are currently evaluated on a clean basis; i.e., without accrued interest. Generating a Tranche Cash Flow To generate a tranche cash flow, Interactive Data s evaluators use the following as model inputs: Deal files: Intex Solutions, an industry leader for securitized fixed- income cashflow models and related analytical software analytics modeling of U.S. Private Label and European RMBS, and provides Interactive Data with new deal files on a daily basis and updated trustee reports monthly. These files enable us to use actual pool and loan level collateral information. Performance Assumptions: When applicable, we apply one of the following methods to project prepayments: a calculated speed using Interactive Data s prepayment model for RMBS, or a speed that is updated based upon the behavior of the collateral. For U.S. Private Label and European RMBS, defaults, severity and Evaluations are provided in the U.S. through Interactive Data Pricing and Reference Data LLC

2 trigger events are also applied to the cash flows where applicable. These inputs are then used to generate daily cash flows that incorporate deal level data to determine principal and interest payments along with an average life. Depending upon the characteristics of a given tranche, a volatility-driven, multi-dimensional single cash flow stream model or option-adjusted spread (OAS) model is used. For U.S. Private Label RMBS and European RMBS with calls or puts, we utilize global market consensus for running cashflows to call or maturity, when deriving an evaluated price. Verifying Tranche Cash Flow When an initial cash flow deal file is received from our source, our evaluators perform checks to confirm that the cash flows appear appropriate for the deal structure. Interactive Data s evaluators analyze features such as the pricing speed, spread, and yield, in order to confirm the deal structure. If any significant deviation between the source data and the evaluators review of the trustee report is noted, Interactive Data performs additional research on that structure. After issuance, information becomes available in the form of the actual payment history (known as factors ). Determining Applicable Benchmark Yield Curves Benchmark treasury and swap curves are created daily using observed values from a market snapshot: o U.S. snapshot taken at 15:00 or 16:00 ET o European snapshot taken at 16:15 UK Time / 17:15 CET Applying Spread/Yield and Generating an Evaluation Tranche specific spreads/yields are determined based on relevant market information and assessment of tranche characteristics including asset type, underlying collateral, issuer, tranche type, average life, collateral performance, ratings and prevailing market conditions. Interactive Data s models determine tranche spread/yield and evaluations for fixed- and floating-rate issues based on techniques consistent with Interactive Data s interpretation of accepted local market trading conventions, including but not limited to: market spread/yield concordance derivation from benchmark curve. Specifically, For fixed rate RMBS, the model takes the average life or expected maturity for each tranche and matches it to the interpolated value on a relevant treasury, swap or Euro Dollar Synthetic Forward rates (EDSF) curve. It then uses that benchmark spread/yield as the base spread/yield. For floating rate RMBS, the current value of the index over which the security resets is used as a discount basis. A tranche specific Discount Margin (DM) is applied to the relevant index value to derive the yield.. The appropriate tranche spread/yield or DM is applied to the corresponding benchmark. This value is then used to discount the cash flows to generate an evaluated price. Derivative U.S. Private Label and European RMBS Tranche Evaluations For some volatile derivative tranches, Interactive Data uses prepayment and evaluated pricing models. These tranche types include, but are not limited to, IOs (Interest Only), POs (Principal Only), IOettes, Floaters, and Inverse floaters. The prepayment model produces varying monthly prepayment projections based on levels of interest rates, seasoning, seasonality, exposure to refinance opportunities, and other variables that reflect loan turnover and homeowner refinancing. The model is derived using historical prepayment data and current market price data. An option adjusted spread (OAS) model may be used when evaluating volatile derivative U.S. Private Label and European tranches. The OAS model uses a combination of Monte Carlo simulation and benchmark matching to determine evaluations, as follows: 1. A Monte Carlo method is used to simulate future interest rate paths. Each path projects the yield curve of swap rates, Treasuries, LIBOR, COFI, MTA and mortgage-backed rates for each month in the future (up to 40 years): The volatility of rates along/among the paths is constructed to match the implied volatilities of available financial instruments such as caps and swaptions. The paths are designed to model a multifactor lognormal process. Collateral specific prepayment speeds are projected at each point in time on each path. Cash flows are projected using the

3 prepayment speeds and then discounted using the path specific short term interest rates plus an option adjusted spread. Call option features embedded in the structures are taken into account at the time of discounting. Evaluations of Distressed Sectors Interactive Data also provides evaluations for issues that may require a more detailed level of research in addition to the process described above. These issue types include, but are not limited to, non-investment grade issues, distressed issues, Prime, Alt-A, Subprime and issues with special terms and conditions. Our evaluation process incorporates techniques and procedures designed to capture the volatility in these sectors. Interactive Data s evaluators start by grouping or sectoring the securities. Deals are grouped by how similar deals are reacting to prevailing market conditions, and are based on attributes such as the collateral type, country of collateral issuer, vintage and tranche. Evaluators then review pool collateral, specifically loan types (e.g., % ARMs, GEO or location, LTV, etc.), structure and statistics to gain insight into the type of collateral backing the deal, and examine the delinquencies, foreclosures, REOs, bankruptcies, and losses among other pipeline indicators to review the health of the deal. After initial review of the above items for a given security, evaluators run stress scenarios to obtain an indication of how the structure performs taking into consideration prepayment, severity, delinquencies, over collateralization, and step-down features. The scenarios and the combination of variables used in the analysis vary on a deal-by-deal basis. Taking this information into consideration, evaluators then review market observations to reflect a market-based evaluation. Evaluation spread adjustments in the model can be made on a tranche and/or deal level, and evaluators typically review all of the tranches in the deal to assess whether the resulting adjustments are consistent and appropriate. Additional adjustments on a tranche level might also be made after this review based on market sentiment, reported trades, bid lists, offerings, etc. received through our network of contacts. Interactive Data seeks to initiate deal level reviews (as described above if it learns of market events, and/or changes in collateral performance from monthly trustee reports, when that information would be expected to impact evaluations. During volatile market conditions, Interactive Data may require a period of time to complete these reviews. In cases where market information is needed to evaluate a specific sector, evaluators work directly with broker/dealer contacts, which may be provided by the client, for market color. Settlement Interactive Data uses T+0 settlement for U.S. Private Label RMBS and T+2 for European RMBS. Evaluation Types Interactive Data offers bid, mean, and ask evaluations. We apply a price or yield adjustment to the bid-side evaluation to calculate an ask evaluation, in some instances a dollar price adjustment is used for the bid/ask relationship. The mean evaluation is the mid-point between the bid and ask evaluated price. If the bid/ask relationship is based on yield, the evaluated mid-price is derived from the mid-yield value Bid to ask spread relationships are set up in Interactive Data s proprietary models by our evaluators and can incorporate security level information including price volatility, credit exposure and liquidity, which are factors typically taken into account in a two-sided market. Creating the bid to ask spread involves the evaluator s judgment, based on his/her experience and on the available market and credit information. Bid to ask spread relationships typically remain constant once established, but can be adjusted as market conditions warrant. For U.S. broker- quoted issues, we apply a zero spread relationship to the bidside valuation, resulting in the same values for the mean and ask. Quality Processes and Controls Following are some of the quality controls performed by Interactive Data when we generate our evaluations: System requirements to enter an explanation for all adjustments made to the evaluation Daily review of market information and data changes that may have an impact on evaluations Review of unchanged evaluations and other applicable data Daily reviews by managers of tolerance reports to confirm processes are being followed Monthly management reviews of evaluator work samples (tolerance reports, client challenges, and other evaluation-related matters)

4 Our Evaluations Interactive Data s bid-side evaluations are marketbased measurements that represent our good faith opinion as to what the holder would receive in an orderly transaction (for an institutional round lot position typically 1MM or greater current value USD or local currency equivalent) under current market conditions. Trades and bids are reviewed to determine that the lot size is representative of an institutional round lot, though smaller or retail sized lots may be considered especially if this is the only or primary trading information available. Interactive Data uses valuation techniques that reflect market participants assumptions and maximize the use of relevant observable inputs including quoted prices for similar assets, benchmark yield curves and market corroborated inputs. Interactive Data s evaluators regularly review the evaluation inputs for securities covered, including executed trades, broker quotes, credit information and collateral attributes and/or cashflow waterfall as applicable. If we determine we do not have sufficient objectively verifiable information about a security's valuation, we will discontinue evaluating the security until we can obtain such information. Interactive Data seeks to obtain market color, including bid information received by our clients, as part of our evaluation methodologies. To that end, we request that clients forward market information to the evaluation team (market information should be sent to: mdr.structured@interactivedata.com). Verified information is reflected in our evaluations to the extent that we deem it formative of our good faith opinion. Interactive Data does not advise clients as to what securities they should buy or sell. Note: For certain U.S. Private Label RMBS, Interactive Data provides a broker quote when sufficient information, such as cash flows or other security structure or market information, is not available to produce an evaluation. Broker- quoted securities are adjusted based solely on our receipt of updated quotes from market makers or broker-dealers recognized as market participants. A list of such issues is compiled daily as of market close and is available via 360View SM. (Please contact your Interactive Data representative for more information.)

5 About Interactive Data Interactive Data Corporation is a trusted leader in financial information. Thousands of financial institutions and active traders, as well as hundreds of software and service providers, subscribe to our fixed income evaluations, reference data, real-time market data, trading infrastructure services, fixed income analytics, desktop solutions and web-based solutions. Interactive Data s offerings support clients around the world with mission-critical functions, including portfolio valuation, regulatory compliance, risk management, electronic trading and wealth management. Interactive Data is headquartered in Bedford, Massachusetts and has over 2,400 employees in offices worldwide. For more information, please visit Interactive Data s Pricing and Reference Data business provides global securities pricing, evaluations, and reference data designed to support financial institutions and investment funds' pricing activities, securities operations, research, and portfolio management. Interactive Data collects, edits, maintains, and delivers data on more than 10 million securities, including daily evaluations for approximately 2.7 million fixed income and international equity issues. Interactive Data specializes in hard-to-get information and evaluates many hard-to value instruments. Pricing, evaluations and reference data are provided in the U.S. through Interactive Data Pricing and Reference Data LLC and internationally through Interactive Data (Europe) Ltd. and Interactive Data (Australia) Pty Ltd. Bedford Chicago Cologne Dubai Dublin Frankfurt Geneva Hayward Hong Kong Jersey, CI London Lombard Luxembourg Madrid Melbourne Milan Minneapolis New York Paris Rome Santa Monica Singapore Sydney Tokyo Zurich Interactive Data Pricing and Reference Data LLC 32 Crosby Drive Bedford, MA Tel: Fax: info@interactivedata.com 100 William Street, 17 th Floor New York, NY Tel: Fax: West Fulton, 7th Floor Chicago, IL Tel: Fax: Limitations Use of this documentation is limited to authorized clients of Interactive Data Pricing and Reference Data services. This material contains information that is confidential and proprietary property and/or trade secrets of Interactive Data Pricing and Reference Data LLC and/or its affiliates, and is not to be published, reproduced, copied, disclosed, or used without the express written consent of Interactive Data Pricing and Reference Data LLC. This document is provided for informational purposes only. The information contained in this document is subject to change without notice and does not constitute any form of warranty, representation, or undertaking. Nothing herein should in any way be deemed to alter the legal rights and obligations contained in agreements between Interactive Data Pricing and Reference Data LLC and/or affiliates and their clients relating to any of the products or services described herein. Interactive Data Pricing and Reference Data LLC does not provide legal, tax, accounting, or other professional advice. Clients should consult with an attorney, tax, or accounting professional regarding any specific legal, tax, or accounting situation. Interactive Data Pricing and Reference Data LLC makes no warranties whatsoever, either express or implied, as to merchantability, fitness for a particular purpose, or any other matter. Without limiting the foregoing, Interactive Data Pricing and Reference Data LLC makes no representation or warranty that any data or information (including but not limited to evaluations) supplied to or by it are complete or free from errors, omissions, or defects. Interactive Data SM and the Interactive Data logo are registered service marks or service marks of Interactive Data Corporation in the United States or other countries. Other products, services, or company names mentioned herein are the property of, and may be the service mark or trademark of, their respective owners Interactive Data Pricing and Reference Data LLC 5575 (0116)