90 Pages

mbs_lecture3

Course: FIN 464, Fall 2008
School: Kentucky
Rating:
 
 
 
 
 

Word Count: 4464

Document Preview

Mortgage-Backed Commercial Securities National University of Singapore July 27, 2001 Notes from lecture given by Brent Ambrose at National University of Singapore July 2001 July 2001 Brent W. Ambrose, University of Kentucky 1 COMMERICAL MORTGAGEBACKED SECURITIES What is a CMBS? A commercial mortgage-backed security (CMBS) is a financial asset. created when an issuer places a commercial mortgage (or collection...

Register Now

Unformatted Document Excerpt

Coursehero >> Kentucky >> Kentucky >> FIN 464

Course Hero has millions of student submitted documents similar to the one
below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.

Course Hero has millions of student submitted documents similar to the one below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.
Mortgage-Backed Commercial Securities National University of Singapore July 27, 2001 Notes from lecture given by Brent Ambrose at National University of Singapore July 2001 July 2001 Brent W. Ambrose, University of Kentucky 1 COMMERICAL MORTGAGEBACKED SECURITIES What is a CMBS? A commercial mortgage-backed security (CMBS) is a financial asset. created when an issuer places a commercial mortgage (or collection of mortgages) into a trust the trust issues classes of bonds backed by the underlying principal and interest payments. July 2001 Brent W. Ambrose, University of Kentucky 2 Objectives for this Session This session will cover the following topics: July 2001 Differences between CMBS and MBS. The anatomy of a CMBS deal. Prepayment penalties on commercial mortgages. CMBS risks (prepayment & default) CMBS underwriting and the role of rating agencies. The role of CMBS servicers. Empirical studies of CMBS default and loss severity. How CMBS deals are rated. Brent W. Ambrose, University of Kentucky 3 CMBS vs. MBS Basic difference between residential MBS and commercial MBS: PREPAYMENT Commercial mortgages have prepayment lockouts and penalties. This changes the termination options. Default is now paramount. July 2001 Brent W. Ambrose, University of Kentucky 4 U.S. CMBS Legalities Real Estate Mortgage Investment Conduits (REMIC) U.S. tax code provision that allows the pooling and securitization of mortgages. Once pool is formed, not allowed to substitute mortgages. July 2001 Brent W. Ambrose, University of Kentucky 5 U.S. CMBS Legalities Financial Asset Securitization Investment Trust (FASIT) created in 1997 allows issuers to substitute and add collateral after securitization now issuer can add mortgages to pool as they are originated July 2001 Brent W. Ambrose, University of Kentucky 6 U.S. CMBS Legalities FASIT Allow securitization of a broader class of assets including: 1. 2. 3. 4. 5. July 2001 construction loans commercial property bridge loans automobile loans credit card receivables home equity loans Brent W. Ambrose, University of Kentucky 7 CMBS Anatomy Cash Flows Subordination Prepayment Penalties CMBS Risks Call Protection Property Diversification Credit Enhancements Underwriting Role of Servicers Brent W. Ambrose, University of Kentucky 8 July 2001 Typical CMBS Structure Whole Loan Pool Rating Sub DSCR LTV Spread Rating Sub CMBS DSCR LTV Spread Interest Principal BBB 0% 1.40 75% 250 BP AAA 30% 2.00 52.50% 136 BP AA A BBB BB B NR 24% 18% 11% 6% 3% 0% 1.84 1.71 1.57 1.49 1.44 1.40 57.00% 61.50% 66.75% 70.50% 72.75% 75.00% 156 BP 176 BP 240 BP 535 BP 825 BP 2200 BP Borrower Equity Borrower Equity Losses Source: Anthony Sanders July 2001 Brent W. Ambrose, University of Kentucky 9 Actual CMBS Examples GMAC Mortgage Pass-through Certificates NationsLink Commercial Mortgage Passthrough Certificates Series 1998-1 July 2001 Brent W. Ambrose, University of Kentucky 10 GMACMortgagePassthroughCertificates,Series1997C1 DMGwastheLeadManagerofa$1.7Billionconduitsecuritizationthelargestofitskindatthe timeofissuance CoLeadManagers: Servicer: Sellers: DeutscheMorganGrenfellInc. LehmanBrothersInc. GMACCommercialMortgageCorporation ContiTradeL.L.C. GermanAmericanMortgageCorporation GMACCommercialMortgageCorporation September25,1997 Go ra h d trib tio e g p ic is u n $ ,6 6 8 ,2 8 1 9 ,9 4 7 35 5 30 8 $ ,7 0 3 4 8 ,2 7 8 2 .6 % 7 .0 % 1 4 1 3 .3 x Clifo ia a rn N wY rk e o P n s lv n e n y a ia Cn e tic t onc u N wJ rs y e e e Tx s ea 1 .2 % 7 9 1 .2 % 2 5 7 4 .5 % 6 7 .0 % 6 1 .0 % 5 1 .4 % IssuanceDate: Collateral: M rtg g p o c a c ris s o a e o l h ra te tic In l p o b la c itia o l a n e N . o mrtg g lo n o f o ae as N . o mrtg g dp p rtie o f o a e ro e s A g Ct- ff d teb la c v . u o a a ne W . A g M rtg g ra td v . o a e te W . A g Ct- ff d teL V td v . u o a T W .A g D C td v . S R M r rpt t p a p eyy s j o or e Ri el t a Mi aiy uf m l t l Oe fi fc I d ti l ns a ur H i ay o t lt s i p Sl dui g kens i l rn Mds i eu x e S-t rg es a l o e f Ml h ea o eo pk b m r i C rgea/ ss dvg o e t c e st l i n a r ai e i n g Or te h 25 51 . % 17 89 . % 10 86 . % 12 01 . % 92 . % 4 73 . % 1 49 . % 2 26 . % 8 25 . % 6 08 . % 9 00 . % 1 The other remaining mortgaged properties are located throughout 37 other states, Puerto Rico and the District of Columbia Source: Anthony Sanders July 2001 Brent W. Ambrose, University of Kentucky 11 GMACMortgagePassThroughCertificates,Series1997 C1 Bonds: Class A-1 A-2 A-3 B C D E F G H J K X Total S ecurities Initial Cert. Balance or Notional Am t. $261, 582,000 $227, 661,000 $724,100,000 $67,879,000 $50,909,000 $50,909,000 $93,334,000 $25,454,000 $84,849,000 $59,394,000 $16,969,000 $33,944,278 $1,696,984,278 $1,696,984,278 Spread 48 62 65 70 75 85 100 118 Rating (Moody's/ Fitch) Ass/AAA Aaa/AAA Aaa/AAA Aa2/AA+ A1/AA A2/A+ Baa2/BBB Baa3/BBBBB/BB B BUnrated Aaa/AAA Percent of Initial Pool B alance 15.4% 13.4% 42.7% 4.0% 3.0% 3.0% 5.5% 1.5% 5.0% 3.5% 1.0% 2.0% N/A Initial Pass- Weighted Through Rate Average L ife Sub-ordination (approx.) (yrs) 28.5% 28.5% 28.5% 24.5% 21.5% 18.5% 13.0% 11.5% 6.5% 3.0% 2.0% 0.0% N/A 6.830% 6.853% 6.869% 6.918% 6.898% 6.997% 7.085% 7.222% 7.414% 6.600% 6.600% 6.600% 1.629% 4.00 7.50 9.71 9.94 9.96 10.01 11.45 13.53 14.93 17.99 19.78 22.0 N/A Paym ent Window 1 - 75 75-108 108-119 119-120 120-120 120-125 125-158 158-170 170-195 195-235 235-242 242-358 1-358 Notional Amt Source: Anthony Sanders July 2001 Brent W. Ambrose, University of Kentucky 12 NationsLink Pass-through Certificates Series 1998-1 $ 1 Billion (approximately) of CMBS Class A-1 A-2 A-3 A B C D E F G IO July 2001 Rating AAA AAA AAA AAA AA A BBB BB B unrated AAA Principal $ 200mm $ 82mm $ 434mm $ 716mm $ 54mm $ 56mm $ 80mm $ 60mm $ 20mm $ 40mm notional Coupon Subordination 6.49% 6.43% 6.40% 6.44% 6.60% 6.90% 7.00% 7.00% 7.00% 0.90% 30% 25% 20% 12% 6% 2% 0% na 101.00% 101.00% 101.00% 101.00% 100.75% 100.50% 85.00% 75.00% 45.00% 4.75% 13 Brent W. Ambrose, University of Kentucky CMBS Anatomy Cash flow prioritization: 1) Principal repayments (both scheduled amortization and unscheduled prepayments) go to retire senior class debt first. CF go to senior classes AAA through BBB Intermediate class Junior class Unrated Equity holder 2) Coupon interest paid to all classes July 2001 Brent W. Ambrose, University of Kentucky 14 CMBS Anatomy Loss prioritization: Principal and interest due the most junior class bondholder must be completely exhausted before any loss is assigned to the class above it. July 2001 Brent W. Ambrose, University of Kentucky 15 The Anatomy of a CMBS Required Subordination The required level of subordination is computed as the expected loss in the event of a recession in the real property market. More specifically, required subordination = probability of loss (given a recession) x severity of loss (given a default) The probability of loss varies from small (say 10%) to large (say 50%), depending on the magnitude of the real property recession. The severity of loss is the amount of the loss conditional on a default. For example, a Class B real property recession will result in loan losses with a 10% probability. The severity of the loss is typically 20% of the loan balance. Therefore, the required subordination for a Class B real property recession is 10% x 20% = 2%. July 2001 Brent W. Ambrose, University of Kentucky 16 The Anatomy of a CMBS Example of Required Subordination Level Calculation (NationsLink Example) Type of Recession AAA AA A BBB BB B Probability of Loss 50% 45% 40% 30% 20% 10% x x x x x x x Severity of Loss 60% 55% 50% 40% 30% 20% = Required Subordination 30% 25% 20% 12% 6% 2% July 2001 Brent W. Ambrose, University of Kentucky 17 The Anatomy of a CMBS The Unrated Piece The unrated piece is used to provide subordination for the lowest rated junior piece. The size of the unrated bond reflects rating agency requirements for loans that are not cross-collateralized and cross-defaulted. The unrated piece is sold privately and typically purchased by the special servicer. July 2001 Brent W. Ambrose, University of Kentucky 18 The Anatomy of a CMBS The Interest Only (IO) Piece The notional balance of the IO piece is initially the aggregate issue amount ($ 1 billion in the example) The notional balance of the IO piece equals the sum of the certificate balances for the sequential pay certificates. The IO piece typically pays a small coupon (e.g. 90bp) and sells at a steep discount. July 2001 Brent W. Ambrose, University of Kentucky 19 CMBS Anatomy Expected Cash Flows Review Principal repayment Scheduled amortization Unscheduled prepayment Interest Penalties July 2001 Hyperamortization Prepayment Penalty Balloon Default Brent W. Ambrose, University of Kentucky 20 CMBS Anatomy (Penalties) A. Hyperamortization (cash trap): all cash flows in excess of operating expenses go to retire debt. Triggered by i. ii. iii. iv. Delinquency failure to maintain required DSCR failure to maintain debt rating failure to maintain adequate reserves July 2001 Brent W. Ambrose, University of Kentucky 21 CMBS Anatomy (Penalties) B. Prepayment penalty: penalty assessed the borrower for early repayment of debt. Penalty may be computed in various ways. July 2001 Brent W. Ambrose, University of Kentucky 22 CMBS Anatomy (Penalties) C. Balloon default: penalty assessed the borrower for failing to refinance at the end of the loan term. July 2001 Brent W. Ambrose, University of Kentucky 23 CMBS Anatomy (Penalties) Prepayment Penalties 1. A (declining) percent of the outstanding balance (e.g. 5-4-3-2-1) paid when the loan is prepaid 2. Yield maintenance: the prepayment penalty is computed as the difference between the book value of the loan and the PV of the remaining contractual payments discounted at some required interest rate. The required interest rate is expressed as some spread over the rate prevailing on comparable maturity Treasuries. A. B. July 2001 300bp over Treasuries zero spread: Treasuries flat Brent W. Ambrose, University of Kentucky 24 CMBS Anatomy (Penalties) Prepayment Penalties: 3. Lockout complete prohibition of prepayment of principal. Usually only in effect during the first few years of the mortgage. July 2001 Brent W. Ambrose, University of Kentucky 25 Simple Prepayment Example Mortgage Assumptions: Two year, $10 million interest only mortgage 10% interest rate on the loan at date of issuance Loan is repaid after 1 year when interest rates fall to 8%. July 2001 Brent W. Ambrose, University of Kentucky 26 Simple Prepayment Example Prepayment Penalties Yield Maintenance penalty provision Penalty = ($1,000,000 - $800,000) / (1+0.08) = $185,185 Percent of Prepaid Amount penalty provision Assume 1% penalty in this example Penalty = 1% * $10 million = $100,000 July 2001 Brent W. Ambrose, University of Kentucky 27 Allocation of Prepayment Penalties Allocation is based on the language in the CMBS prospectus Ultimately determined by investment bankers and lawyers during the creation of CMBS Underwriters have a great deal of latitude No standard approach exists July 2001 Brent W. Ambrose, University of Kentucky 28 Yield Maintenance Calculations One bullet loan tranched into two classes: Assumptions: 1. Underlying loan is a ten year bullet loan priced at par and pays a 9% coupon 2. Multi-class structure: Senior class is $90 million and pays an 8% coupon (priced at par). Subordinate class is a $10 million classs that pay an 18% coupon (priced at par). 3. Borrower prepays in full at at year 3. Current interest rates are 100bps lower than in year 0. July 2001 Brent W. Ambrose, University of Kentucky 29 CMBS Prepayment Example Whole Mortgage Prepayment Calculation -- Bullet Loan Principal Loan Type Term of Loan Discount Rate $ 100 Year Prepaid Non-amortizing Original Interest Rate 10 New Interest Rate 8% 3 9% 8% Year 1 2 3 4 5 6 7 8 9 10 $ $ $ $ $ $ $ $ $ $ Original Coupon Payments 9.00 9.00 9.00 9.00 9.00 9.00 9.00 Discounted Cash Flows $ $ $ $ $ $ $ $ $ $ $ $ 8.33 7.72 7.14 6.62 6.13 5.67 5.25 46.86 5.21 Reinvested Interest Payments $ $ $ $ 8.00 $ 8.00 $ 8.00 $ 8.00 $ 8.00 $ 8.00 $ 8.00 Discounted Cash Flows $ $ $ $ 7.41 $ 6.86 $ 6.35 $ 5.88 $ 5.44 $ 5.04 $ 4.67 $ 41.65 Penalty (Difference in Cash Flows) July 2001 Brent W. Ambrose, University of Kentucky 30 Allocation of Penalty: 1. Percentage Prepayed Senior Class (90%) Junior Class (10%) $ $ $ 4.69 0.52 5.21 July 2001 Brent W. Ambrose, University of Kentucky 31 Principal Loan Type Term of Loan Discount Rate Using a Make-whole Calculation Senior Tranche $ 90 Year Prepaid Non-amortizing Original Interest Rate 10 New Interest Rate 7% 3 8% 7% Year 1 2 3 4 5 6 7 8 9 10 $ $ $ $ $ $ $ $ $ $ Original Coupon Payments 7.20 7.20 7.20 7.20 7.20 7.20 7.20 Discounted Cash Flows $ $ $ $ $ $ $ $ $ $ $ $ 6.73 6.29 5.88 5.49 5.13 4.80 4.48 38.80 4.85 Reinvested Interest Payments $ $ $ $ 6.30 $ 6.30 $ 6.30 $ 6.30 $ 6.30 $ 6.30 $ 6.30 Discounted Cash Flows $ $ $ $ 5.89 $ 5.50 $ 5.14 $ 4.81 $ 4.49 $ 4.20 $ 3.92 $ 33.95 Penalty (Difference in Cash Flows) July 2001 Brent W. Ambrose, University of Kentucky 32 Principal Loan Type Term of Loan Discount Rate Using a Make-whole Calculation Junior Tranche $ 10 Year Prepaid Non-amortizing Original Interest Rate 10 New Interest Rate 17% 3 18% 17% Year 1 2 3 4 5 6 7 8 9 10 $ $ $ $ $ $ $ $ $ $ Original Coupon Payments 1.80 1.80 1.80 1.80 1.80 1.80 1.80 Discounted Cash Flows $ $ $ $ $ $ $ $ $ $ $ $ $ $ 1.54 1.31 1.12 0.96 0.82 0.70 0.60 7.06 0.39 5.24 (0.04) Reinvested Interest Payments $ $ $ $ 1.70 $ 1.70 $ 1.70 $ 1.70 $ 1.70 $ 1.70 $ 1.70 Discounted Cash Flows $ $ $ $ 1.45 $ 1.24 $ 1.06 $ 0.91 $ 0.78 $ 0.66 $ 0.57 $ 6.67 Penalty (Difference in Cash Flows) Total Penalty Payment (Junior + Senior) Difference in Penalty Payments July 2001 Brent W. Ambrose, University of Kentucky 33 CMBS Anatomy Risk Rating Sub DSCR LTV Price Prepayment Risk Credit/De fault Risk Premium AAA 30% 2.00 52.50% 102 Discount AA A BBB BB B NR 24% 18% 11% 6% 3% 0% 1.84 1.71 1.57 1.49 1.44 1.40 57.00% 61.50% 66.75% 70.50% 72.75% 75.00% 101 100 98 75 65 35 Extension Risk Credit/De fault Risk Source: Anthony Sanders July 2001 Brent W. Ambrose, University of Kentucky 34 CMBS Anatomy CMBS Risk impacted by: property quality geographic location tenant creditworthiness July 2001 Brent W. Ambrose, University of Kentucky 35 CMBS Risks Default risk: Income property loans are typically nonrecourse. Borrower has the financial incentive to default when the market value of the property falls below the outstanding balance of the loan (negative equity). Also referred to as optimal, strategic, or financial default. July 2001 Brent W. Ambrose, University of Kentucky 36 CMBS Risks Balloon risk: income property mortgages typically have terms that are less than the loan amortization period, thus the borrower must refinance to continue making mortgage payments. Circumstances in the property and capital markets may have changed in ways that make refinancing difficult or even impossible. Also referred to as refinancing risk July 2001 Brent W. Ambrose, University of Kentucky 37 CMBS Risks Prepayment risk: Many income property mortgages provide some call protection. lock-out provisions prepayment penalties Treasury defeasance Some income property mortgages do not have any of these features. July 2001 Brent W. Ambrose, University of Kentucky 38 CMBS Call Protection Lock-out provisions: prohibit loan prepayment over given period Prepayment penalties: paid in a lump sum at the time of prepayment; Cash flows are proportionally allocated to remaining certificate classes. See previous examples of a (declining) percent of the outstanding loan balance or yield maintenance agreements July 2001 Brent W. Ambrose, University of Kentucky 39 CMBS Call Protection Treasury defeasance: the borrower must purchase a series of Treasuries that provide the same future cash flows assuming the loan not prepaid. Property release provision: prohibits asset substitution; prevents the issuer/lender from removing the stronger properties from the pool. July 2001 Brent W. Ambrose, University of Kentucky 40 Property Diversification Diversification across: Loan size: usually no single loan exceeds 5% of the aggregate issue amount. An exception to this is a fusion deal, where a single large loan is packaged with several smaller loans. Property type Property location State metropolitan area July 2001 Brent W. Ambrose, University of Kentucky 41 Loan Diversification Table 2. The twenty largest loans underlying the GMAC 1999-C3 deal. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Name Biltmore Fashion Prime Outlets Equity Inns One Colorado Comerica Bank 120 Monument 125 Maiden Texas Development Sherman Plaza Alliance TP Bush Tower County Line Sherwood Lakes Laurel Portfolio Sweet Paper Sheraton Portsmouth Trinity C ommons Village Square Golden Books Air Touch Location, MSA Phoenix, Arizona Niagara Falls, New York Various Pasadena, San California Jose, California Indianapolis, Indiana New York, New York Houston, Texas Van Nuys, California Various New York, New York Jackson, Mississippi Schererville, Indiana Various Various Portsmouth, New Hampshire Fort Worth, Texas Indianapolis, Indiana Fayetteville, North Carolina Dublin, Ohio Category Retail Retail Hotel Retail Office Office Office Apartment Office Apartment Office Retail Apartment Apartment Warehouse Hotel Retail Apartment Warehouse Office Loan Amount $80,000,000 $62,835,426 $46,511,317 $42,628,093 $33,640,510 $28,955,362 $28,500,000 $26,926,701 $25,984,904 $24,888,157 $23,000,000 $20,990,264 $20,162,442 $17,950,331 $17,420,000 $15,949,087 $15,242,981 $14,993,950 $14,493,350 $13,992,523 Source: Charter Research. July 2001 Brent W. Ambrose, University of Kentucky 42 Geographic Diversification Table 4. Aggregate loan amounts by state for GMAC 1999-C3 deal. State California Texas New York Arizona Indiana Ohio Mississippi New Jersey Other Loan Amount $257,522,410 $162,355,125 $130,070,471 $99,942,794 $68,623,516 $44,982,528 $23,067,864 $22,983,973 $342,473,371 No. of Loans 33 26 7 5 5 5 2 5 50 % of Pool 22.35% 14.09% 11.29% 8.68% 5.96% 3.90% 2.00% 2.00% 29.73% Total $1,152,022,052 138 100.00% Source: Charter Research. July 2001 Brent W. Ambrose, University of Kentucky 43 Property Type Diversification Table 5. Aggregate loan amounts by property type for GMAC 1999-C3 deal. Property Type Apartment Office Retail Warehouse Hotel Other Loan Amount $259,779,802 $322,053,844 $350,683,062 $99,126,075 $105,832,139 $14,547,130 No. of Loans 39 36 34 15 8 6 % of Pool 22.55% 27.96% 30.44% 8.60% 9.19% 1.26% Total $1,152,022,052 138 100.00% Source: Charter Research. July 2001 Brent W. Ambrose, University of Kentucky 44 Credit Enhancements Subordination Cross collateralization: properties that collateralize individual loans are pledged against all loans in the pool Cross default: allows the lender to call ALL LOANS in the event a single loan is in default. July 2001 Brent W. Ambrose, University of Kentucky 45 Credit Enhancements Lock box: Gives the trustee control of the property gross revenues. The trustee assigns priority in the following order: (1) taxes and insurance; (2) operating expenses; (3) debt service; (4) management fees; (5) reserves for replacements; (6) equity investor July 2001 Brent W. Ambrose, University of Kentucky 46 Credit Enhancements Overcollateralization: When the book value of the loans exceed the par value of the bonds issued. Most common in residential MBS Especially common in CMO structure July 2001 Brent W. Ambrose, University of Kentucky 47 Reserve Funds Established at loan closing to: Provide liquidity: to pay interest for investment grade bonds Service the asset: to pay July 2001 property taxes property insurance legal fees Maintenance Brent W. Ambrose, University of Kentucky 48 Standardized CMBS Underwriting Key Underwriting Characteristics Debt Service Coverage Ratio (DSCR) Loan to Value Ratio (LTV) Average loan size Max loan not to exceed certain percentage (e.g. 5%) Diversification across Property types Geographic locations Prepayment terms Loan maturities July 2001 Brent W. Ambrose, University of Kentucky 49 Role of rating agencies Establish different rating criteria for various property types. Negotiate subordination levels with issuers. Track property performance/delinquencies Servicer and trustee report ongoing loan level performance Monthly/quarterly DSCRs occupancy levels updated bond information July 2001 Brent W. Ambrose, University of Kentucky 50 Role of Servicers Master Servicer: Oversees the deal and servicing agreements Facilitates timely payment of principal and interest May provide (servicer) advances for delinquent/defaulted loans Sub-Servicer: loan originator in a conduit deal who retains servicing July 2001 Brent W. Ambrose, University of Kentucky 51 Role of Servicers Special Servicer: Becomes engaged when loan more than 60 days delinquent. Has the authority to Extend the loan Modify/restructure the loan (based on an appraisal) Foreclose July 2001 Brent W. Ambrose, University of Kentucky 52 Pricing CMBS Unlike residential MBS, the underlying mortgages have little prepayment risk. However, default risk is now relevant due to these risks: Lease termination risk Lease rollover risk Imperative to monitor developments in the overall real estate market. For example, low vacancy rates may lead to additional construction. This additional supply can result in reduced real lease rates in future years. Has implications on the ability of the property to service the debt in future years. July 2001 Brent W. Ambrose, University of Kentucky 53 Pricing CMBS Rating agencies play a critical role in the CMBS pricing process. S&P, Moodys, and Duff & Phelps maintain internal models of collateral risk in order to rate default risk associated with CMBS deals. In addition to having to price deals in accordance with rating agency opinions, it is wise to understand the underlying real estate markets in order to anticipate payments delays or defaults. Example. Lease rates and vacancy rates may look great at the moment, but will these indicators prompt developers/banks into another frenzy of construction? July 2001 Brent W. Ambrose, University of Kentucky 54 Pricing CMBS Remember CMBS product is part of the fixed-income universe. Capital markets are linked such that shocks in one market impact others. For example Russian debt crises in 1998 See following charts. July 2001 Brent W. Ambrose, University of Kentucky 55 July 2001 Brent W. Ambrose, University of Kentucky 56 NCREIF Total Returns 0.04 0.03 0.02 0.01 0 1990 -0.01 -0.02 -0.03 -0.04 -0.05 Date 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 July 2001 Return Brent W. Ambrose, University of Kentucky 57 July 2001 Brent W. Ambrose, University of Kentucky 58 July 2001 Brent W. Ambrose, University of Kentucky 59 July 2001 Brent W. Ambrose, University of Kentucky 60 Bond Market Yields 12 10 8 AAA 6 BBB Treas 4 % 2 0 Jan-90 Jan-91 Jan-92 Jan-93 Jan-94 Jan-95 Jan-96 Date Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 July 2001 Brent W. Ambrose, University of Kentucky 61 Bond Spreads Bond Credit Spreads 3 2.5 2 Spread 1.5 AAA BBB 1 0.5 0 Jan-90 Jan-91 Jan-92 Jan-93 Jan-94 Jan-95 Jan-96 Date Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 July 2001 Brent W. Ambrose, University of Kentucky 62 Income Property Debt: Default and Loss Severity * What do we know about income property default and loss severity? 1. 2. 3. 4. Fitch ICBA, Inc. (1998) Corcoran and Kao (1998) Vandell, Barnes, Hartzell, Kraft, and Wendt (1993) Snyderman (1994) *-these notes are based on lecture material provided by Thomas Thibodeau at Southern Methodist University. July 2001 Brent W. Ambrose, University of Kentucky 63 Income Property Debt: Default and Loss Severity Fitch IBCA, Inc. (1998) Study examines Fitch rated transactions between 1991 and 1996 18,839 loans (in 33 CMBS transactions) total principal $16.1 billion 84% thrift loans (mostly RTC) 16% conduit loans July 2001 Brent W. Ambrose, University of Kentucky 64 Income Property Debt: Default and Loss Severity Fitch IBCA, Inc. (1998) Examines relationship between default/loss severity and: Debt service coverage ratio (DSCR) Property type State Loan Size Fixed/floating rate loan Loan type (e.g. amortizing, balloon) Servicer flexibility Foreclosure type July 2001 Brent W. Ambrose, University of Kentucky 65 Income Property Debt: Default and Loss Severity Fitch IBCA, Inc. (1998) Default: > 60 days past due on debt service or > 90 days past due on balloon payment 3,134 lifetime defaulted loans (16.64%) 3,002 RTC loans (96% of defaults) annual default rate 4.3% July 2001 Brent W. Ambrose, University of Kentucky 66 Income Property Debt: Default and Loss Severity Fitch IBCA, Inc. (1998) Loss = loan balance at securitization + interest advanced + property protection expenses - loan amortization - property income - net sales proceeds Losses reported as a percent of loan balance at securitization for loans COMPLETELY resolved (e.g. properties sold). July 2001 Brent W. Ambrose, University of Kentucky 67 Income Property Debt: Default and Loss Severity Fitch IBCA, Inc. (1998) Source of losses: Decrease in property value + advanced interest + advanced property protection expenses - amortization - property income (combined) Average Loss Rate: 35.8% 10.5% 7.7% 14.9% 39.1% July 2001 Brent W. Ambrose, University of Kentucky 68 Income Property Debt: Default and Loss Severity Fitch IBCA, Inc. (1998) DSCR 0.01-0.49 0.50-0.79 0.80-0.89 0.90-0.99 1.00-1.14 1.15-1.24 1.25-1.34 1.35-1.49 1.50-1.74 1.75+ July 2001 Default Rate 8.0% 7.0% 6.6% 6.5% 4.4% 2.6% 2.4% 2.8% 3.1% 2.9% Brent W. Ambrose, University of Kentucky Loss 47% 52% 42% 41% 29% 48% 27% 36% 41% 22% 69 Income Property Debt: Default and Loss Severity Fitch IBCA, Inc. (1998) Property Type Lodging Multifamily Nursing Office Industrial Other Retirement Warehouse Default Rate 4.2% 3.9% 4.0% 4.8% 4.7% 4.2% 4.7% 2.5% Loss 27% 46% 11% 38% 27% 46% 34% 29% July 2001 Brent W. Ambrose, University of Kentucky 70 Income Property Debt: Default and Loss Severity Fitch IBCA, Inc. (1998) State Highest 5: New York Louisianna New Mexico Arizona Massachusetts Iowa Florida Texas Washington Oregon Default Rate 6.8% 5.8% 5.5% 5.2% 5.2% 3.5% 3.1% 3.1% 2.4% 1.9% Loss 32% 69% 25% 22% 40% 65% 44% 45% 25% 34% 71 Lowest 5: July 2001 Brent W. Ambrose, University of Kentucky Income Property Debt: Default and Loss Severity Fitch IBCA, Inc. (1998) Loan Size $0.0M - $0.5M $0.5M - $1.0M $1.0M - $5.0M $5.0M - $10M > $10M July 2001 Default Rate 4.0% 5.5% 4.9% 3.9% 2.0% Brent W. Ambrose, University of Kentucky Loss 38% 42% 39% 37% 34% 72 Income Property Debt: Default and Loss Severity Fitch IBCA, Inc. (1998) Other Results Interest Rate: Floating rate Fixed rate Loan Type: Amortizing Balloon 2.6% 6.0% 5.4% 3.5% 42% 35% Default Rate Loss Judicial foreclosures take longer and cost (between 8% and 26%) more than nonjudicial (e.g. power of sale states) foreclosures. July 2001 Brent W. Ambrose, University of Kentucky 73 Income Prop...

Textbooks related to the document above:
Find millions of documents on Course Hero - Study Guides, Lecture Notes, Reference Materials, Practice Exams and more. Course Hero has millions of course specific materials providing students with the best way to expand their education.

Below is a small sample set of documents:

Kentucky - MA - 551
[Munkres, Problem 6, page 181] Problem. Let (X, d) be a metric space. If f : X - X satisfies the condition d(f (x), f (y) = d(x, y) for all x, y X, then f is called an isometry of X. Show that if f is an isometry and X is compact, then f is bijectiv
Kentucky - MA - 321
MA/CS 321:001 MWF 11:0011:50 FB 213 Fall 2004Instructor: Russell Brown Oce: POT741 Phone: 257-3951 russell.brown@uky.eduAnnouncements. Homework 7 will be due on Monday, 1 November 2004. The exam will be delayed until Friday, 5 November 2004. Plea
UMass (Amherst) - BIEP - 540
PubHlth 540Hypothesis TestingPage 1 of 55Unit 7. Hypothesis TestingTopic1. The Logic of Hypothesis Testing . 2. Beware the Statistical Hypothesis Test . 3. Introduction to Type I, II Error and Statistical Power . 4. Normal: Test for , 2 Kno
Taylor IN - COS - 104
Telecommunications (Chapter 6)Thursday, September 26Agenda TWOtestimonies? Video 7:00 PM Thursday & Friday Questions? LectureAnalog vs. DigitalAnalog: signal of continuously varying strength and/or quality Digital: signal represente
Taylor IN - CSE - 121
While WaitingRichard Pattis quotes Programming languages, like pizzas, come in too sizes; too big and too small. The code for a computer system provides the ecology in which [more] code is born, matures, and dies. A well-designed habitat allows fo
Taylor IN - CSE - 280
Lisp-ish quotes while waiting "Lisp is a programmable programming language." - John Foderaro, CACM, September 1991 "One can even conjecture that Lisp owes its survival specifically to the fact that its programs are lists, which everyone, including
Duke - STA - 216
Bayesian Analysis of Structural Equation Models Sperm Motility Example Summary of sperm motility data Outcome Dose Mean SD Y1 0 88.4 9.21 8 76.1 7.54 24 82.1 15.6 72 77.2 13.3 Y2 0 0.219 0.013 8 0.216 0.013 24 0.207 0.012 72 0.206 0.020 Y3 0 25.5 2.7
Duke - STA - 216
STA 216, Generalized Linear Models, Lecture 8September 19, 2008High-dimensional PredictorsData Augmentation for Binary DataAlternatives to SSVSA variety of fast alternatives to SSVS have been proposed Many approaches rely on sparse maximum
Duke - STA - 216
STA 216 Generalized Linear ModelsMeets: 2:50-4:05 T/TH (Old Chem 025)Instructor: David Dunson 219A Old Chemistry, 684-8025 dunson@stat.duke.edu Teaching Assistant: Jenhwa Chu 114 Old Chemistry jenhwa@stat.duke.eduSTA 216 SyllabusTopics to be c
Duke - STA - 103
STA103 Spring 2001Name Circle section: F 8:00, F 9:10, F 10:30, F 11:50Diagnostic QuizSTA103 is more math-intensive than STA101 or STA102; you need to have completed at least MTH31 or its equivalent to do well in the class. The simple problems t
Duke - STA - 294
Pairwise comparison table Calculate all pairwise alignment scores and arrange them in a table S1 S2 S3 S4 S5 2 0 9 1S1 10 5 4 S2 10 25 8 S3 5 25 11 S4 4 8 11 S5 2 0 9 1Convert all score into distances . 1. FengDoolitle : D=log(SSrand)/(SmaxSrand)
Duke - STA - 216
STA 216 Fall 2000 Assignment 4 Refer to the binary regression O-ring example from class. 1. Write down the expression for the working response Z and the weights W for complementary log-log link. 2. Carry out k steps of the Fisher scoring algorithm us
Duke - STA - 216
alpha0alpha1pi[i]Y[i]for(i IN 1 : 24)alpha1 5.55112E-17 -0.2 -0.4 -0.6 10850 10900 10950 iteration 40.0 30.0 20.0 10.0 0.0alpha1 1.0 0.5 0.0 -0.5 -1.0 0 20 lag 40 1.0 0.5 0.0 -0.5 -1.0 2XWSXIURGHIDXRJPRGHO*UDSK
Duke - STA - 103
Multivariate probability distributions Often we are interested in more than 1 aspect of an experiment/trial Will have more than 1 random variable Interest the probability of a combination of events (results of the different aspects of the experim
Duke - STA - 104
STA 104 MTH 135Name: Probability First Test 2:10-3:30 pm Thursday, 3 October 1996This is a closed-book examination, so please do not refer to your notes, the text, or to any other books. If you dont understand something in one of the questions fe
Duke - STA - 216
STA 216 Fall 2000 Assignment 3 Refer to the O-ring example from class and the last assignment. Assume that you have M possible models (M1 , . . . , MM ) for O-ring failure and that you can calculate the posterior probability of each model (Mj |Y ). F
Taylor IN - COS - 381
1IntroductionFor this lab, you are going to begin the construction of your simulated computer. The resulting component of this assignment is a 32 32 register file, that is a set of 32 registers each of which is 32 bits in size. See Figure 5.7 in
Taylor IN - COS - 382
Lexical AnalysisJonathan GeislerFebrurary 8, 2006Jonathan GeislerLexical AnalysisLanguage RecognitionLets use the same grammar as Monday and validate a sentence for that grammar: 1/2.5=Jonathan GeislerLexical AnalysisParse treesThi
Duke - STA - 242
% D0 @8 3 @ 6 ( 6 d (' 6 ( D F ' 62 B (' 3 E ( 3A F @ 2A YB u 1 (A D B c 8 ( V H 9 E u x Rc G(D F Q( ( 0 ` W (A B W t CD B 3 B' F ( 3 @D F G 6C 3 C E5 d kl 35 (B D1 B Rc 1 ( Q(@ d 3A 6B G k 0 B2 3 D' C5 dj D i Q3 G@ A (C 3 9 9(@
Duke - STA - 242
' (' c 9rQI S a C S Y A sA A )`VAR%3yeav)wYBSVC%ifVWeC4d4VIi9zVQw9`T4BI9 VI9x3eC%BIH4ea4)iYwss vqUWVSBc4eaBSDBAb`4VIxEABYXDVADeWeS4`BYHVAR3R%BIp%eYBSR%Bfw9ESheapvCb`eAiShi}iYeRY S o S 9 A Sd W A G ( 9 c 9aAW A 9 Cf 9 9Q sr YW Y 9 cQ 9Q 9 Ca 9Q
Duke - STA - 244
STA2444/23/2003Take Home Final ExamDue 5/1/2003 by 5pm This is an open note/open book test. All work must be your own.Study of the growth of plants can be a crucial element in understanding how they compete for resources. For example, soybean
Duke - STA - 244
STA2444/7/2003Homework 7Due 4/14/2001 1. The matrix X(i) X(i) can be written as X(i) X(i) = X X - xi xi where xi is the ith row of X and X(i) is the matrix X with the ith row removed. Use this to show (X(i) X(i) )-1 = (X X)-1 + (X X)-1 xi xi (X
Duke - STA - 244
STA2441/15/2003Homework 2Due 1/22/2003 1. Write the following two way analysis of variance (AOV) model with interactions Yijk = + i + j + ij + with i = 1, 2, 3, j = 1, 2, k = 1, 2 in matrix notation. 2. Suppose we have a k k matrix S partition
Duke - STA - 244
STA2442/28/2005Homework 5Due 3/7/2001 1. For a random vector n , is called exchangeable if has the same distribution as any permutation of the vector . If is exchangeable, prove that E( ) = 1 ( ), and that the Cov( ) = has the forma a b .
Duke - STA - 244
STA2441/15/2001Homework 1Due 1/22/20011. Assume that we have a sample of size n where Y i = 0 + 1 Xi + e i and the errors ei are iid N (0, 2 ). (a) Find the maximum likelihood estimator of 2 , 2 . Hint: let = 2 and maximize. ^ (b) Under
Duke - STA - 103
j3{a" r"D jmR"jE9" "h9r! Em {"6{!9h4!" 6 Em!ED j RR4 Ew{9r4j 7"{h9 m!9 ' r9D jmREp"j7tj r7E{m!9 49 6r9BX EBr{p !4hD 3h b rq T1I b ` X '
Duke - STA - 244
STA2442/5/2005Homework 3Due 2/12/2001 1. Recall from class that a non-central 2 (m, ) can be represented as a Poisson mixture of central 2 random variables, where Y P (/2) and X|Y 2 (m + 2y, 0). Find the mean and variance of a non-central Chi-
Duke - STA - 103
% # " &$!
Duke - STA - 103
p2hGuBgB@G8ey$uHx3G4(f'Be2Cy$02#dBr}X`v2!XhX1gB'0(fX`'dufdGF'0ffxEVz w Gxp( 7 7 C ) W!t h I# % %t 1` 1 # 3 h C h )t ) # % h c! !t c! ! E # W p p t T x` 1 h` q u@G f h g f d 8Ve (H f q u@G q u@G f h qp 8@ruf u Yr@ruiH f s f qp q u@G
Duke - STA - 103
v @ ' @ 9 u@4Cb'0B")Q~u5q d5s9tgts 84iq8qB")qYU'rRQuIQ8D ' @ R1p YV1 6 qY'AIqVIPUCg9dqd&FB6&BrV8AapQ6U@rTrR2EFQ9p$ @9 'Y T d ' '7 @ 1 x T1 f 9 TD1D d vgIG"dw6 rRQuh'C u@ qVIswqYb't@4ph'h'UCw|UCIqG4p0$ ' @ R1 d T ') ` d T T d p 4g9bdqdF
Duke - STA - 244
STA2444/9/2002Homework 7Due 4/16/2001 1. Problem 15.7 in CW. To obtain case diagnostics in S-Plus, fit a model using the QR option, i.e. mylm.obj <- lm(Y X1 + X2, data=mydataframe, qr=T) To obtain the case diagnostics, use the function ls.diag(
Duke - STA - 103
g c gycp x c j e d p p x uy e d t r ey c d pc p e ~isq~n"cfreqqg "q%tsmis|q~g c d uc u xe x c d c p | pe re c jl s'}ers weup1Wy~y sfesqk qknk yc d ey c d pc p e g p upp rcyc u t yc t r sc q%qe srs|q~nuc1ux%q Tssxv3s"9 a p p x uy e d
Duke - STA - 103
h7W9AAP Br2rW 014sGC3 d bc2G90'DbcGAVb{b652IIr2B@I65{Dbfb8fD28fS65rS6db`U01`@DrA 3 q l P 37A) w v 3 H7 9 H7 1 3)7 5 PW ')CA97 3 7) H7 v 5 ) iuxhfi xh F 3A w v d w 3 ) d 3 97) H42FtSt2PzBBbH G9BSG942@Dp2tbI@b822@01b01b6501SxSH s 7 PA 9 H7 9A Q9A
Duke - STA - 103
()c 5 q1 )h oefen'a5sf q1 r AeqAdGeC(oC9edPAq0adPA0P#( y f HR cp HR H FE f ( f @ V c H HRp p c 9 H HR D H h ( d FACE@ y CS@0G%dHeQnmeqiC0G0CH0giACBse6T 3E R h f 3 R cp H S 3 ` H FE f S 3E E H FE E 3 D @ c c x } ( Q 5 sf q1 r a5 h
Duke - STA - 244
STA2443/19/2005Homework 6Due 3/28/2002 1. For the usual linear model Y N (X, -1 In ) with prior distributions N (bo , Vo ) independent of and p() 1/: (a) Find the posterior distribution of |. (b) Can you find a closed form expression for th
Duke - STA - 102
ill sandwich "Yes" "Yes" "Yes" "Yes" "Yes" "Yes" "Yes" "Yes" "Yes" "Yes" "Yes" "Yes" "Yes" "Yes" "Yes" "Yes" "Yes" "Yes"
Duke - STA - 102
exposure nma "high" 28 "high" 35 "high" 37 "high" 37 "high" 43.5 "high" 44 "high" 45.5 "high" 46 "high" 48 "high" 48.
Duke - STA - 102
subject fev1 gender 1 2.30 0 2 2.15 1 3 3.50 1 4 2.60 0 5 2.75 0 6 2.82 1 7 4.05 1 8 2.25 1 9 2.68 0
Duke - STA - 244
x1 x2 y1 y2 y3 y4 10 8 8.04 9.14 7.46 6.58 8 8 6.95 8.14 6.77 5.76 13 8 7.58 8.74 12.74 7.71 9 8 8.81 8.77 7.11 8.84 11 8 8.33 9.26 7.81 8.47 14 8 9.96 8.1 8.84 7.04 6 8 7.24 6.13 6.08 5.25 4 19 4.26 3.1 5.39 12.5 12 8 10.84 9.13 8.15 5.56
Duke - STA - 244
D m S WS y 0 10 1 3408 623 0.04 5 1 206.8 680.2 0.1 5 1 1841.2 721.4 0.16 5 1 1223.2 750.4 0.28 5 1 861.2 789.4 0.04 5 2 2810.8 672.2 0.1 5 2 860.8 709.2 0.16 5 2 592.8 731.2 0.28 5 2 2642.8 778.2 0.04 5 3 2399.2 668.4 0.1 5 3 327.2 715.6
Duke - STA - 244
FUEL/POP INC LIC/POP POP TAX VEH/POP VM/VEH 644.147 14.826 0.70923 4041 13 0.911408 11.0684 474.545 21.761 0.549091 550 8 0.669091 10.5625 552.524 16.297 0.660573 3665 18 0.777899 12.2119 683.539 14.218 0.735857 2351 18.7 0.615908 14.0981 501.34
Duke - STA - 244
Pressure Temp 20.79 194.5 20.79 194.3 22.4 197.9 22.67 198.4 23.15 199.4 23.35 199.9 23.89 200.9 23.99 201.1 24.02 201.4 24.01 201.3 25.14 203.6 26.57 204.6 28.49 209.5 27.76 208.6 29.04 210.7 29.88 211.9 30.06 212.2
Duke - STA - 244
Duke - STA - 244
STA2444/18/2002Homework 8Due 4/26/2001 Refer to Exercise 11.5 in CW (page 285). Use any appropriate methods covered in class to answer the problem (Bayesian, Frequentist, or compare both). Provide a typed solution describing the problem and how
Duke - STA - 244
STA2442/14/2005Homework 4Due 2/21/2001 1. Consider the linear model Y = X 1 1 + X 2 2 + where X1 is n q and X2 is n (p q), with both matrices of full column rank. Consider the problem of testing N H : 1 = 0. Assume that N (0, 2 In ). (a) Gi
Duke - STA - 244
U X1 X2 Y 0.493151 1 1 0.872302 1.40245 2 1 1.59988 2.31175 3 1 2.4019 3.22104 4 1 3.25942 4.13034 5 1 4.14616 5.03964 6 1 5.04607 5.94894 7 1 5.95154 6.85823 8 1 6.85928 7.76753 9 1 7.76795 8.67683 10 1 8.677 0.0770038 1 2 0.557762 0.986
Duke - STA - 113
Name:Section:STAT 113 Midterm 31 Otis 1979, Journal of Psychology interviewed people waiting to see the space aliens lm Close Encounters of the Third Kind." Each person was asked to state his or her degree of agreement with the statement Life on
Duke - STA - 113
Name:Section:STAT 113 Midterm 21a. 1pt Suppose y is a normally distributed random variable with mean 0 and variance 1.0, i.e. y is standard normal. Find P ,1:0 y 0:5.1b. 2pt Suppose y is normally distributed random variable with mean 10 and va
Duke - STA - 113
Homework 9a SolutionsAs yi iid Bernoullip, then E yi = = p. Setting this expression to its respective sample P y =1 ^ y moment, we obtain: = n , or p = n ^ n! y n,y where K = 8.8 c. For the Binomial experiment, the likelihood function is L = K p
Duke - STA - 113
Name:1a. 2pt Suppose y is normally distributed random variable with mean = 5:0 and variance 2 = 4:0, i.e. y N 5:0; 4:0. Find P 3:0 y 12:0. 1b. 2pt Suppose y is a 2 distributed random variable with = 12 degrees of freedom. Find cuto s c and d, su
Duke - STA - 113
Name:Section:STAT 113 Midterm 31 Otis1 1979 interviewed people waiting to see the space aliens lm Close Encounters of the Third Kind." Each person was asked to state his or her degree of agreement with the statement Life on Earth is being observ
Duke - STA - 113
Name:Section:STAT 113, Spring 99 Midterm 3On all problems, please show your work. Just the correct answer without justi cation and intermediate results is not acceptable.Note:1.In a survey of college students, it was found that X = 69 of th
Duke - STA - 113
SOLUTIONNote: Version B had slightly di erent numbers. But the basic problems were the same. You can recognize Version B by Name" instead of Name:" on the top line, i.e., a missing :" after Name". 1. On questions 1a-f: 2pts for the correct choice;
Duke - STA - 110
Min. 1st Qu. Median Mean 3rd Qu. Max. 41 76 89 85.28 96 100 Decimal point is 1 place to the right of the colon 4 : 14 4 : 5 : 5 : 55 6 : 1 6 : 55569 7 : 0011123333444
Duke - BME - 227
Articulating External Fixation Device for the ElbowBrace Yourself: Blake Iceton Lisa Richards Advisor: Aura Gimm, PhD. Gimm, BME 227L 4/22/08Clients: Dr. Marc Richard Dr. Michael RichardTeam IntroductionsBlake Iceton Lisa Richards Dr. Marc Ric
Duke - BME - 227
Easy-to-Use Bacterial Growth ChamberXin Zheng Derek Hsu Bruno Gugelmin1AgendaIntroduction II. Background Information III. Objective IV. Design Ideas V. Comparison of Design Ideas VI. Conclusion VII. Whats Next?I.21IntroductionClient: D
Duke - BME - 227
Creation and Testing of a Novel Ultrasonic Imaging System for Pre-Operative Staging of Rectal CancerAndrew Sobel and Brian LemisterClients: Dr. Kathy Nightingale, Department of Biomedical Engineering, Duke University Liang Zhai, Department of Biome
Duke - BME - 227
ECPrepAlexandru Avram Aaron Globerman Blake Sowerby September 27, 2005Presentation OverviewTechnical Background Client Background Problem Statement Client Requirements Proposed Designs Slush Model Sandwich Model Spot ModelDesign Evaluations
Duke - BME - 227
Endoscopic Surgical Platform with Magnets!Team ESPM The Worldwide LeaderBy Dave Abdollahian, Nicole Bell, Jarred Callura, Matt Furlow and Dean ParasFor Dr. George Mutafyan ClientBME 277L, Duke University, Spring 20071Overview Background
Duke - BME - 227
Articulating External Fixation Device for the ElbowBlake Iceton and Lisa Richards Clients: Dr. Marc Richard, M.D., Department of Orthopedic Surgery; Dr. Michael Richard, M.D., Department of Ophthalmology Advisor: J. Aura Gimm, Ph.D., Pratt School of
Duke - BME - 227
Articulating External Fixation Device for the ElbowTeam Brace Yourself: Blake Iceton Lisa Richards Advisor: Aura Gimm, PhD. BME 227L 2/12/08Team IntroductionsBlake Iceton Lisa Richards Dr. Marc Richard Orthopedic Surgery Dr. Michael Richard - Op