JUNE 2006

JPMORGAN MBS PRIMER NTIAL E CONFID ND A PRIVATE Y L T STRIC MBS Analyst Certification The strategist(s) denoted by an asterisk (“*”) certify that: (1) all of the views expressed herein accurately reflect his or her personal views about any and all of the subject instruments or issuers; and (2) no part of his or her compensation was, is, or will be directly or indirectly related to the specific recommendations or views expressed by him or her in this material, except that his or her compensation may be based on the performance of the views expressed.

This research contains the views, opinions and recommendations of research strategists with JPMorgan US Strategy. Research strategists routinely consult with JPMSI trading desk personnel in formulating views, opinions and recommendations in preparing this research. Trading desks may trade or may have traded as principal on the basis of the research strategist(s) views and report(s). Therefore, this research may not be independent from the proprietary of JPMSI trading desks which may conflict with your interests. In addition, research strategists receive compensation based, in part, on the quality of their analysis, firm revenues, trading revenues, and competitive factors.

Copyright 2006 J.P. Morgan Chase & Co. All rights reserved. JPMorgan is the marketing name for J.P. Morgan Chase & Co. and its subsidiaries and affiliates worldwide. J.P. Morgan Securities Inc. is a member of NYSE and SIPC. JPMorgan Chase Bank is a member of FDIC. J.P. Morgan Futures Inc. is a member of the NFA. J.P. Morgan Securities Ltd. and J.P. Morgan plc are authorised by the FSA and members of the LSE. J.P. Morgan Europe Limited is authorised by the FSA. J.P. Morgan Equities Limited is a member of the Johannesburg Securities Exchange and is regulated by the FSB. J.P. Morgan Securities (Asia Pacific) Limited is registered as an investment advisers with the Securities & Futures Commission in Hong Kong and itsCE numbers is AAJ321 Jardine Fleming Singapore Securities Pte Ltd is a member of Singapore Exchange Securities Trading Limited and is regulated by the Monetary Authority of Singapore (“MAS”). J.P. Morgan Securities Asia Private Limited is regulated by the MAS and the Financial Supervisory Agency in Japan. J.P.Morgan Australia Limited (ABN 52 002 888 011) is a licensed securities dealer.

R This material is provided for information only and is not intended as a recommendation or an offer or solicitation for the purchase or sale of any security or financial instrument. E JPMorgan and its affiliates may have positions (long or short), effect transactions or make markets in securities or financial instruments mentioned herein (or options with respect thereto), or provide advice or loans to, or participate in the underwriting or restructuring of the obligations of, issuers mentioned herein. The information contained herein RIM is as of the date and time referenced above and JPMorgan does not undertake any obligation to update such information. All market prices, data and other information are not P warranted as to completeness or accuracy and are subject to change without notice. Transactions involving securities and financial instruments mentioned herein (including S futures and options) may not be appropriate for all investors. Clients should contact their salespersons at, and execute transactions through, a JPMorgan entity qualified in their B

M home jurisdiction unless governing law permits otherwise J.P. Morgan Securities Inc. is a member of NASD, NYSE and SIPC In the UK and other EEA countries, this material is not available for distribution to persons regarded as private customers (or equivalent) in their home jurisdiction. AN G R O M

JP MBS Market Overview and Origination 1

Demand 11

Mortgage Cashflows and Intro to Prepayments 23

Valuation and OAS 32

Prepayments Analysis and Reports 46

TBA Market and Specified Pools 62

Relative Value Trading 74 R

E Case Studies 99 RIM P

S ARMs 110 B M AN

G CMOs 116 R O M

JP MBS Index 163 MBS 1 Agency MBS market composition and issuance

Securitized agency market composition Annual fixed-rate net issuance ($ billions)

Hybrid ARM 1/1 ARM Hybrid ARM 292 IO ($107.1 ($29.4 276 ($229.5 217 211 231 billion) billion) billion) 156 3% 1% 109 93 7% Other Fixed ($157.2 -17 billion) 1998 1999 2000 2001 2002 2003 2004 2005 2006 5% Total = $__mm Source: JPMorgan, FNMA, FHLMC, GNMA

15-year Annual hybrid ARM net issuance ($ billions) ($650 billion) Annual hybrid ARM net issuance ($ billions)

TION 19% 30-year 79 ($2.2 74

GINA 49 52

I trillion)

R 65%

O 10 13 D N

A -13 -11 -31 W E

I 1998 1999 2000 2001 2002 2003 2004 2005 2006

V Total = $3.4 trillion Source: JPMorgan, FNMA, FHLMC, GNMA Source: JPMorgan, FNMA, FHLMC, GNMA OVER RKET A

M MBS 2 MBS in the U.S. fixed income market

Fixed income market composition Overview Fixed income market composition „ Largest US fixed income asset class

„ Many products to choose from within the MBS Asset- market backed Municipal 8% 9% „ Agency fixed-rates and ARMs Money „ Non-agency fixed-rates and ARMs (Jumbos, Alt- Market U.S. As) 14% Treasury 17% „ Whole loans „ CMOs and other structured MBS

Fed Agencies „ Superior liquidity 10%

TION „ The TBA market adds unique liquidity to MBS

„ MBS market often used to express duration and GINA I Mortgage R curve views (due to its liquidity and size)

O Corporate Related

D „ Agency fixed-rate pass-throughs is 34% of the 19% 23% N

A Lehman U.S. Aggregate Index (a benchmark of the

W U.S. investment grade debt). E I

V Total = $25.9 trillion Source: The Market Association, as of March 2006 OVER RKET A

M MBS 3 The mortgage market has surged, thanks to a strong housing market and cash-out refis

1-4 Family Mortgage Debt Outstanding ($ billions) 1-4 Family Mortgage Debt Outstanding ($ billions) 8, 978 8, 683 8, 079

7, 120

6, 317 5, 614 5, 133 TION GINA I R O D N A W E I

V 2000 2001 2002 200 3 2004 2005 2006

Source: Association, Federal Reserve Board OVER RKET A

M MBS 4 The MBS market links borrowers and investors

Agencies Mortgage Loans MBS Pass-through Securitization

Mortgage MBS Dealers lenders

TION Borrowers MBS Investors GINA I R „ O The issuer of the pass-through obtains the mortgages either by purchasing or originating the loans D

N „ Loans with similar characteristics are pooled together; loans are securitized A

W „ The investor has undivided ownership (the investor is entitled to the pro-rata share of interest and E

I principal payments of the underlying loans) V „ A “pass-through” is the basic MBS structure

OVER „ It passes the monthly principal and interest payments, minus a servicing spread, from a pool of mortgages to investors RKET A

M MBS 5 Origination: The Menu of Mortgages Has Expanded

ƒ Origination: production of new loans in primary market

ƒ Products

– Fixed-rate mortgages (30-year / 20-year / 15-year) – Adjustable rate mortgages (Hybrid ARMs: 3/1, 5/1, 7/1, 10/1) – Interest-Only – MTAs; Option ARMs – Other ƒ TION Balloon mortgages (5-year / 7-year) ƒ Prepayment penalty mortgages GINA I

R ƒ “Conforming” balance loans O D N A ƒ “Non-conforming” loans (Private label, Non-agencies) W E I

V ƒ Jumbos and Alt-As OVER RKET A

M MBS 6 Understanding Mortgage Collateral : Borrower Credit & Housing Leverage

Borrower Credit and Information „ FICO Score – Historical Credit Use and Management — Avg FICO Score for Jumbo Mortgages : ~730 — Avg FICO Score for Alt-A Mortgages : ~700 — Avg FICO Score for Subprime Mortgages : ~600 — Non-Linear Relationship Between FICO and Propensity to Default

„ Documentation — Full vs. Limited/Reduced/No Doc

TION „ Leverage (Debt to Income Ratios) „ Reserves : Staying Power in the event of financial trouble GINA I R O D N A W E I V OVER RKET A

M MBS 7 Understanding Collateral cont…

Housing Leverage „ Loan-to-Value Ratio — House Value / Mortgage Amount

— Higher LTV Æ Less Equity Protection for the Mortgage Investor Æ Higher Risk

„ Occupancy — Owner Occupied – Borrower Lives in the Property (Most Secure) — Second Home – Borrower has personal ties to the property — Investor – Business Decision on Economic Situation (Least Secure) TION „ Property Type

GINA — Single Family Property (Most Secure) I R

O — Condos

D — Multi-Family N A W E I V OVER RKET A

M MBS 8 Conforming loan limits rose by 16%, reaching $417,000 in 2006

Conforming Limits ($’000)

450

Conforming loan 400 limit for 2006

350

300

TION 250 GINA I

R 200 O D N A 150 W E

I 1990 1992 1994 1996 1998 2000 2002 2004 2006 V Source: FHFB, JPMorgan OVER RKET A

M MBS 9 Origination channels

ƒ $3.0 trillion origination volume in

2005 ƒ Top 5 mortgage banking companies account for over 48% of all new ƒ Retail (42%) - loan officer employed by origination volume mortgage banking company; mortgage loan is closed in the name of the lender 1. Countrywide ƒ Broker (34%) – mortgage loan broker 2. Wells Fargo represents borrower to lending institution; 3. Washington Mutual mortgage loan is closed in the name of the 4. Chase Home Finance lender

TION 5. CitiMortgage

ƒ Correspondent (24%) – independent Source: National Mortgage News, as of Q1 2006 GINA

I mortgage banking company; mortgage loan R

O is closed in “ABC” Mortgages name and D

N sold to mortgage banker A

W ƒ Internet – mortgage loan originated and E I

V funded by mortgage banking company through website OVER RKET A

M MBS 10 Market Overview and Origination 1

Demand 11

Mortgage Cashflows and Intro to Prepayments 23

Valuation and OAS 32

Prepayments Analysis and Reports 46

TBA Market and Specified Pools 62

Relative Value Trading 74

Case Studies 99 R E ARMs 110 RIM P S B

M CMOs 116 AN G R MBS Index 163 O M

JP MBS 11 Major MBS investors

MBS Investor Breakdown MBS Investor Breakdown MBS Investors ($ billion) YE 2004 Mid-2005 % Chg Dealer Personal Investor Type All MSRs Non-Agency All MSRs Non-Agency % of MSRs since 2004 REITs Inventory Sector Fannie Mae/Freddie Mac $ 1, 261 $ 2 67 $ 1,192 $ 3 63 26% -5% 2% 1% 6% FDIC Commercial Banks $ 8 76 $ 124 $ 913 $ 158 20% 4% Life Insurance Cos. $ 46 5 N.A. $ 480 $ 150 10% 3% Finance Foreign Investors $ 2 80 $ 3 0 $ 400 $ 5 0 9% 43% Companies Agencies Mutual Funds $ 31 8 N.A. $ 325 N.A. 7% 2% 2% 29% Personal Sector $ 27 0 N.A. $ 290 N.A. 6% 7% Pension Public Pension Funds $ 27 0 N.A. $ 275 N.A. 6% 2% Funds All Thrifts $ 23 4 $ 7 $ 228 $ 6 5% -3% 9% Priv. Pension Funds $ 12 5 $ 1 3 $ 128 $ 1 5 3% 2% FHLBanks $ 11 3 $ 7 1 $ 117 $ 7 1 3% 3% REITs $ 95 $ 50 $ 105 $ 6 0 2% 11% Mutual Funds Finance Companies $ 85 N.A. $ 88 N.A. 2% 4% 7% MBS Dealer Inventory $ 4 1 $ 15 $ 55 $ 20 1% 34% Federal Credit Unions $ 2 8 N.A. $ 29 N.A. 1% 2% Subtotal: $ 4, 462 $ 5 77 $ 4,625 $ 8 93 Foreign All Other Investors* $ 31 7 $ 474 Investors Total Outstanding $ 4, 779 $ 1,076 $ 5,098 $ 1,289 9% Banks Insurance 25% Source: Inside MBS & ABS Companies 10%

Total = $4.6 trillion Source: Inside MBS & ABS D

DEMAN MBS 12 Foreign demand has dominated the mortgage market over the past several years

Net Purchases ($ billions, annual)

250 Foreign GSE Bank 200

150

100

50

0

-50

-100

-150 2001 2002 2003 2004 2005 2006

Source: US Treasury, Federal Reserve, FNMA, FHLMC, JPMorgan * Foreign purchase data: March ’06, GSE: April ’06, Bank: May ‘06 D

DEMAN MBS 13 Trends in bank demand

„ Bank holdings are still growing, but at a slower pace.

„ Some signs that bank demand could slow:

„ MBS holdings as a percentage of deposits have increased

„ C&I loan growth has picked up

„ But, there are mitigating factors:

„ Deposit rates have remained sticky

„ Unrealized losses are significant

„ Mortgages remain one of the few sectors that offers the size and liquidity that large banks need D

DEMAN MBS 14 Large banks have been drawn to the mortgage market

TopTop 3 3Bank Bank Holdi Holdingsngs as as % %of of Total Total Banks’ Banks’

40% 38% 36% 34% 32% 30% 28% 26% 24% 22% 20% Mar -01 Sep-01 Mar -02 Sep-02 Mar -03 Sep-03 Mar -04 Sep-04 Mar -05 Sep-05 Mar -06

Source: Federal Reserve, JPMorgan „ Large banks have enjoyed the liquidity of the mortgage market for large trades, and average trade size has increased significantly.

D „ Diversification and movement away from credit risk have been themes, but could shift if mortgages remain tight.

DEMAN MBS 15 Top 10 banks ranked by MBS portfolios as of 1Q 2006

Bank Holding Company Total MBS Change Pass-through Change CMO Change BANK OF AMERICA CORPORATION 212,273,791 15,121,205 206,670,221 15,808,550 5,603,570 -687,345 WACHOVIA CORPORATION 94,293,000 3,173,000 78,527,000 2,338,000 15,766,000 835,000 JPMORGAN CHASE & CO 41,644,000 18,775,000 40,354,000 18,633,000 1,290,000 142,000 WELLS FARGO & COMPANY 40,042,000 7,676,000 33,969,000 8,729,000 6,073,000 -1,053,000 US BANCORP 35,975,000 -1,547,000 22,928,000 -576,000 13,047,000 -971,000 CITIZENS FINANCIAL GROUP INC 31,828,058 1,087,257 11,429,339 176,681 20,398,719 910,576 BANK OF NEW YORK COMPANY INC 22,671,000 189,000 2,908,000 -129,000 19,763,000 318,000 COMMERCE BANCORP INC 20,908,478 1,076,917 3,651,529 104,519 17,256,949 972,398 STATE STREET CORPORATION 20,252,615 581,874 6,324,590 -626,064 13,928,025 1,207,938 SUNTRUST BANKS INC 17,287,171 264,477 11,580,470 -159,294 5,706,701 423,771

·The top 10 banks account for over 58% of all bank MBS holdings.

Source: Federal Reserve, JPMorgan D

DEMAN MBS 16 Bank MBS holdings continue to grow

MBS Holdings of Large Banks ($ billions)

650

600

550

500

450

400

350

300

250

200 May-99 May-00 May-01 May-02 May-03 May-04 May-05 May-06

Source: Federal Reserve D

DEMAN MBS 17 Security purchases and C&I loan growth has typically been inversely correlated

MBS and C&I Holdings (Annual changes since 2000)

15%

10%

g 5% Ch % l a 0% u n n

A -5% C&I

-10% R2 = 0.6155 -15% -5% 0% 5% 10% 15% 20% 25% 30% 35%

MBS Annual % Chg

Source: Federal Reserve, JPMorgan D

DEMAN MBS 18 With many bank positions underwater, it is unlikely that there will be large selling

150 20 Quarterly Changes in Security Holdings Unrealized Gains/Losses 15 100 10 ngs ($bn) di 5

50 osses ($bn) es hol 0 ns/l

securiti 0 n (5)

(10) Unrealized gai (50) Changes i Banks did not sell amid losses in 2000… (15) … While most bank sales have occurred amid gains (100) (20) 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 D

DEMAN MBS 19 The GSEs

•Fannie Mae

•Freddie Mac

•Ginnie Mae

•FHLBs – the MPF and MPP programs

•GNMA pools are backed by FHA/VA loans which are government insured. These pools have an explicit U.S. government guarantee and a zero risk- capital weighting.

•FNMA and FHLMC pools are backed by conventional conforming loans, have an implicit U.S. government guarantee, and a 20% risk-capital weighting. Single-

D family loan limit is $359,650 in 2005.

DEMAN MBS 20 Unique role of GSEs: issuer / investor

ƒ Mission is to facilitate secondary mortgage market in U.S. which provides steady flow of low cost mortgage funds

ƒ Issue

ƒ Hold MBS, CMOs, and loans as well as ABS, CMBS, and mortgage-related spread products

ƒ Large portfolios (FN + FH hold over $1.4 trillion loans and MBS) demand active hedging via swap and swaption markets D

DEMAN MBS 21 Agency portfolio growth has slowed

Retained Portfolio ($ billions)

1000 FNMA Portfolio FHLMC Portfolio

900

800

700

600

500

400

300 Apr-01 Oct-01 Apr-02 Oct-02 Apr-03 Oct-03 Apr-04 Oct-04 Apr-05 Oct-05 Apr-06

Source: FNMA, FHLMC, JPMorgan D

DEMAN MBS 22 Market Overview and Origination 1

Demand 11

Mortgage Cashflows and Intro to Prepayments 23

Valuation and OAS 32

Prepayments Analysis and Reports 46

TBA Market and Specified Pools 62

Relative Value Trading 74

R Case Studies 99 E RIM P ARMs 110 S B M

AN CMOs 116 G R O M MBS Index 163 JP MBS 23 MBS Terminology

ƒ Pools are comprised of mortgage loans with similar rates and terms

ƒ WAC – weighted average of all loans in pool (vs Coupon)

ƒ WAM – weighted average of loans in pool MENTS Y ƒ WALA – weighted average loan age EPA R P

O ƒ Original face – original principal amount of pool T O ƒ Current face – remaining principal balance of pool NTR I

ƒ Origination year – average origination year of loans in pool; age (WALA) is important in prepayment AND assessment (“seasoning”)

ƒ CPR – Constant Prepayment Rate – annualized percentage of remaining principal prepaid CASHFLOWS E G

MORTGA MBS 24 Mortgage cash-flow characteristics

Homeowner's August After processing, security payment due holders receive shares of (in arrears) August payment MENTS

Y FHLMC FNMA 14th 24th EPA R P

O August 1st September 1st October 1st T 30 days in arrears 14 to 24 days processing O NTR I 44 to 54 day delay AND CASHFLOWS E G

MORTGA MBS 25 Mortgage cash-flow

„ Example: $500,000 purchase price; $400,000 loan amount; 6% mortgage rate; 30-year fixed-rate loan

„ Using “MP” function on BBG… MENTS Y EPA R P O T O NTR I AND CASHFLOWS E G

Source: Bloomberg

MORTGA MBS 26 Mortgage cash-flows: without prepayments MENTS Y EPA R P O T Interest O NTR I AND Principal CASHFLOWS E G

Source: Bloomberg

MORTGA MBS 27 Mortgage cash-flows: with prepayments MENTS Y EPA R

P Interest O T O NTR I Pre-paid AND Principal Principal CASHFLOWS E G

Source: Bloomberg

MORTGA MBS 28 Prepayments: source of MBS optionality

ƒ Borrowers have the right to prepay at any time without penalty – in effect “calling” their loans away from investors; prepayments may be partial or complete

ƒ Valuing this call option and the cash flow uncertainty it creates is the key to MENTS Y understanding MBS EPA R P ƒ Timing and rate of prepayments vary and produce non-level, less-predictable O T cash flows O

NTR ƒ Prepayment (Call) Risk vs extension risk I AND CASHFLOWS E G

MORTGA MBS 29 Determinants of prepayments

ƒ Prepayments can be for economic / non-economic reasons

ƒ Interest rate incentive

ƒ shape: refi down the curve MENTS Y

EPA ƒ WALA: mobility increases over time R P O T

O ƒ Seasonality: Summer months have the fastest speeds owing to school vacation NTR I

AND ƒ “Burnout”: Remaining borrowers in a pool may be credit constrained

ƒ Cash-out refinancing: Take built-up equity out of the home CASHFLOWS E G ƒ Default: Shows as a prepayment in agency pools

MORTGA MBS 30 Prepayment standards ƒ CPR – Constant Prepayment Rate – annualized percentage of remaining principal prepaid

ƒ PSA – prepayment vector expressed as a series of CPRs; begins at .2% in the first month, increases .2% per month, leveling out at 6.0% in month 30;

MENTS prepayment assumptions for pricing stated as linear multiples of PSA Y schedule EPA R P

O 14 T 200 PSA

O 12

NTR 10 I )

% 8

( 100 PSA R AND 6 CP 4

2

0 CASHFLOWS 0 1020304050

E Age G

MORTGA MBS 31 Market Overview and Origination 1

Demand 11

Mortgage Cashflows and Intro to Prepayments 23

Valuation and OAS 32

Prepayments Analysis and Reports 46

TBA Market and Specified Pools 62

Relative Value Trading 74 R

E Case Studies 99 RIM P

S ARMs 110 B M

AN CMOs 116 G R O M

JP MBS Index 163 MBS 32 Many Different Types of Spreads ƒ Basic: static over a single point on the curve ƒ “I” : spread to Treasury ƒ “N” : spread to swaps

ƒ Intermediate: zero volatility yield curve spread ƒ “Z” : spread to Treasury curve ƒ “E” : spread to Libor/swap curve Libor ZSpread on JPMorgan’s analytic reports. ƒ Advanced ƒ OAS : option-adjusted spread S A

O ƒ LIBOR OAS ND A ƒ Treasury OAS N O

VALUATI MBS 33 Yield analysis in the MBS market

„ Static Spread (Yield Spread): standard measure of incremental return over a single benchmark Treasury Î Compares MBS to single point on the yield curve, usually to the interpolated point closest to the Weighted Average Life of the MBS Î But MBS does not return principal in one lump sum but over many periods. A better assumption would include multiple data points on the yield curve. Z Spread takes this another step further.

„ Z Spread (Yield Curve Spread) : discounts each monthly MBS cashflow by the monthly forward rates derived from the curve Î More accurate for securities that return principal over many periods as opposed to bullets Î Still a static measure since it assumes that interest rates and MBS cashflows remain constant S A O ND A N O

VALUATI MBS 34 Evaluating pass-throughs: yield / average life S A O ND A N O

Source: Bloomberg

VALUATI MBS 35 Prepayments and OAS

„ Prepayment issues: Î Reinvestment risk: n When rates decline and speeds increase the investor has to reinvest an increased amount of principal at lower rates o When rates increase and speeds decline, the investor has less cashflow to reinvest at higher rates Î Discount bonds: when rates decline, the benefit of earlier return of principal at par may mitigate reinvestment risk Î Premium bonds: when rates increase, the benefit of a larger outstanding principal balance and longer average life means higher and more interest payments which may mitigate the reinvestment risk

„ OAS has been derived to account for the dispersion and uncertainty associated with this return of principal from MBS S A O ND A N O

VALUATI MBS 36 OAS Calculation

„ To incorporate prepayment volatility in the valuation of MBS, we can calculate a theoretical price for a given OAS 1. Hundreds of hypothetical interest rate paths are simulated 2. On each interest rate path the prepayment model is used to predict prepayment speeds and thus, MBS cashflows 3. For each path, the present value of the projected cashflows are calculated using a specified spread, s, which is added to the forward rates 4. Value of MBS = Average value of PV(s) over all simulated interest rate paths = AVGPV(s) where s is OAS

„ To find OAS given market price: 1. Start with an initial estimate for OAS 2. Calculate AGVPV(s) and keep adjusting until AVGPV(s) = market price

„ Drawback of OAS: 1. The spread earned by the investor depends on the actual path realized and can be drastically different from the OAS 2. Wide differences in OASs are produced by different firms models due to different term structures, S volatility assumptions and prepayment projections A

O 3. Doesn’t account for dollar roll financing

ND 4. Is a “black box” – difficult for investors to decompose OAS into its component parts. A N O

VALUATI MBS 37 Pass-through risk measurement (duration)

ƒ Various measures of duration: % change in price for a 1% change in rates.

ƒ Modified duration is inappropriate for pass-throughs as it cannot accommodate varying cash flows.

ƒ OAD is found by calculating constant OAS prices for parallel curve shifts.

ƒ Empirical duration uses actual observations regressed against a Treasury benchmark. Directional/empiricals could be different against different parts of the yield curve.

ƒ None of these measures is perfect. We tend to use a combination of them all. S A O ND A N O

VALUATI MBS 38 Empirical durations

FN 5.5 Empirical Durations (using 10-yr Tsy; Dec.05 – Jun. 06) FN 6 Empirical Durations (using 10-yr Tsy; Dec.05 – Jun. 06) FN 5.5 Empirical Durations (using 10-yr Tsy; Dec.05 – Jun. 06)

FN 6 FN 5.5

y = -4.8009x + 0.0056 1.0 1.0 y = -3.3453x + 0.0015 R2 = 0.8927 2 g 0.5 g R = 0.8266 h h 0.5 C C

x 0.0 x 0.0

% P -0.5 % P -0.5 -1.0

-0.15 -0.10 -0.05 0.00 0.05 0.10 -0.15 -0.10 -0.05 0.00 0.05 0.10

Yield Chg (%) Yield Chg (%)

Source: JPMorgan S Source: JPMorgan A O ND A N O

VALUATI MBS 39 Rates have little effect on OAS

FN 5.5 OAS vs rates (using 10-yr Tsy; Dec.05 – Jun. 06) FN 6 OAS vs rates (using 10-yr Tsy; Dec.05 – Jun. 06) FN 5.5 OAS vs rates (using 10-yr Tsy; Dec.05 – Jun. 06)

FN 6 FN 5.5 y = 6.455x - 0.0508 R2 = 0.0713 4.0 4.0

y = -2.2723x - 0.0758 ) ) 2.0 R2 = 0.0079 2.0 hg (bps

hg (bps 0.0 0.0 C -2.0 -2.0 OAS OAS C -4.0 -4.0 -0.15 -0.10 -0.05 0.00 0.05 0.10 -0.15 -0.10 -0.05 0.00 0.05 0.10

Yield Chg (%) Yield Chg (%)

Source: JPMorgan Source: JPMorgan S A O ND A N O

VALUATI MBS 40 Pass-through risk measurement (convexity)

ƒ Convexity: the rate at which the duration of a security changes as interest rates change.

– Positive convexity implies that for small, equal and opposite changes in interest rates, the increase in price if rates go down will be more than the decrease in price if rates rise.

– Negative convexity implies that the increase in price if rates go down will be smaller than the decrease in price if rates rise.

– Bullet Treasuries have positive convexity. Pass-throughs typically have negative convexity. S A O ND A N O

VALUATI MBS 41 Negative convexity of mortgages

FN 6 prices ($) vs shift in rates (bps) 105

100

) 95 ($ x

FN 6 P 90

85 S A

O 80 -300 -225 -150 -75 0 75 150 225 300 ND

A Bps Source: JPMorgan N O

VALUATI MBS 42 Mortgages have embedded options – valuation needs to incorporate vol

ƒ Homeowners have the right to prepay at any time during the life of the mortgage

ƒ Consequently, an MBS investor is short many options to the homeowner:

Underlying Short Long Short 1m x 1y 1m x 10y Option Long 5m x 1y 5m x 10y

ƒ Term structure models are calibrated to the entire vol surface in swaptions

ƒ Higher vol should cause mortgages to cheapen

FN 30 Vega FN15 Vega

S 5.0 -0.21 4.5 -0.091 A

O 5.5 -0.26 5.0 -0.123 6.0 -0.27 5.5 -0.145 ND

A 6.5 -0.23 6.0 -0.111 N O

Source: JPMorgan

VALUATI MBS 43 Mortgage efficiency of pricing in changes in implied vol has been increasing

Correlation of 1-week change in current coupon ZV spread and 3x7 swaption premium, rolling six-month window

1.0

0.8 0.6

0.4 0.2

0.0 (0.2) S A

O (0.4) ND A (0.6) N O 97 98 99 00 01 02 03 04 05 06

Source: JPMorgan

VALUATI MBS 44 Changes in mortgage market duration can impact the rates markets

The rate of extension of the mortgage A sell-off could cause the curve to market will slow in a sell-off steepen

Change in 10-year equivalents of the agency fixed rate market for various parallel shifts in rates Change in 10-year equivalents for the mortgage market across the curve for a parallel +50 rate shock

400 200 bn)

$ 200 150 vs ( 0 ($bn) equi 100 yr -200 quiv 10- n -400 50

-600 10-year E

hange i 0 C -800 S

A -100 -75 -50 -25 0 25 50 75 100 -50 O Rate change (bp) 2 5 Tenor 10 30 ND A N O

VALUATI MBS 45 Market Overview and Origination 1

Demand 11

Mortgage Cashflows and Intro to Prepayments 23

Valuation and OAS 32

Prepayments Analysis and Reports 46

TBA Market and Specified Pools 62

Relative Value Trading 74

Case Studies 99 R E RIM

P ARMs 110 S B M CMOs 116 AN G R O

M MBS Index 163

JP MBS 46 A closer look at turnover

„ Existing Home Sales (EHS)

„ Strong Lock-in

„ Seasonality and Calendar Effect

„ Cumulative Home Appreciation (CHPA)

„ Overall Shorter Baseline Ramp:

„ Interaction between EHS, CHPA, Lock-in and Aging Ramp successfully captures periods of apparent shorter and longer ramps „ In effect lower CHPA lengthens the ramp REPORTS

D „ Lock-in also lengthens the apparent ramp N A S I S Y L NA A NTS ME Y A P E R

P MBS 47 Home price appreciation and discount speeds have been highly correlated…

„ 87% correlation between discount speeds and the housing strength now „ … in the weaker housing environment of 2000, there was no correlation.

Discount Speeds by State (Last 12 mo.) versus HPI in 2005 14 Discount Speeds by State vs. HPI in 2000

14 13 13 AZ 12 CA 12 11 11 % 10 VA MD FL % 10 NJ AZ MA 9 9 8 TX FL

REPORTS CA 8 1-year CPR, WA MI D 7 1-year CPR, 7 MI N PA MA TX PA VA A NY 6 6 OH NJ S

I 5 5 OH MD

S NY

Y 4 4 L 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40

NA Home Price Appreciation,% Home Price Appreciation, % A

NTS Notes: Deep discount: 75bps or more out of the money; balance weighted average 12-month CPR observed in the past year ME Y A P E R

P MBS 48 Seasoning ramps under different HPA assumptions

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3

Turnover Multiplier 0.2 0.1

REPORTS 0.0

D 1 112131415161718191101111 N A

S WALA I S

Y 0pct hpa 1pct hpa 4pct hpa 8pct hpa L NA A NTS ME Y A Source: JPMorgan P E R

P MBS 49 Turnover (cont’d) Lock-in (disincentive to move)

„ Captures the relationship between turnover and refinancing “disincentive” Lock-in (long WAM) 1 „ Strong Lock-in. However, home price appreciation can strongly mitigate 0.9 lock-in 0.8

0.7 iplier lt

u 0.6 M

REPORTS 0.5 D N

A 0.4 S I

S 0.3 Y

L 1 1.1 1.2 1.3 1.4 1.5 1.6 NA A MtgRate/WAC NTS ME Y A P E R

P MBS 50 Turnover (cont’d)

Seasonality: Patterns tend to be impacted by weather and school schedules

„ School schedules and weather conditions are the main reasons for seasonal behavior 1.6

„ There is also a separate “day count” 1.4 adjustment to account for different collection days in each month 1.2

1.0

0.8 REPORTS D

N 0.6 A S I

S 0.4 Y L NA

A 0.2

NTS 0.0

ME Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Y A P E R

P MBS 51 Economics of refinancing

• CATO (curve at origination) – borrowers who take out a 30-year mortgage in a steep curve environment are likely to exhibit slow turnover

• SATO (spread at origination) – high mortgage rates relative to prevailing rates at origination indicate credit impairment. These borrowers are less likely to refinance

• Loan size – with similar fixed costs for refinancing, borrowers with larger loan sizes are more likely to refi

REPORTS • Home price appreciation – higher HPA allows borrowers to “cash-out refi”, or may allow some D

N borrowers to “cure” and obtain a better mortgage rate with A lower LTV A S I S

Y • Mortgage banking capacity – in large refi waves mortgage bankers may become inundated L

NA with supply, causing mortgage spreads to widen A NTS ME Y A P E R

P MBS 52 The Refinancing Index

Seasonally-adjusted refinancing index

12,000

10,000

8,000

6,000

4,000 REPORTS D N

A 2,000 S I S Y L 0 NA

A Jun-00 Dec-00 Jun-01 Dec-01 Jun-02 Dec-02 Jun-03 Dec-03 Jun-04 Dec-04 Jun-05 Dec-05 Jun-06

Source: MBAA NTS ME Y A P E R

P MBS 53 Burnout

1.2

1

0.8

0.6

0.4 Burnout Multiplier 0.2

REPORTS 0 D

N 0 0 0 0 0 0

A 0 10 3 50 70 90 70 90 30 11 130 15 1 190 210 230 25 27 2 310 330 350 370 39 41 4 450 470 490 S I S

Y Cumulative Incentive (bps) L NA A NTS ME Y A Source: JPMorgan P E R

P MBS 54 Primary differences between GNMA and conventional pools

Characteristic Conventional GNMA I GNMA II

Assumable No Yes Pass-through rate 25 to 250bps below loan rate 50bps below 50 to 150bps below loan rate Guarantee fee Negotiated (10 to 25bps) 6bps 6bps Servicing fee 25bps minimum 44bps 44bps minimum Mortgage insurance LTVs worse than 80% Mandatory Mandatory Excess servicing Allowed Not allowed Allowed Buy-ups/-downs Allowed Not allowed 10% buy-down allowed Delay days 24 for FNMA, 14 for Golds 14 days 19 days Prepayment reports Fifth business day Fifth business day Seventh business day

Revised GNMA II pooling guidelines came into effect July 1, 2003. REPORTS D N A S I S Y L NA A NTS ME Y A Source: JPMorgan, FNMA, FHLMC, GNMA P E R

P MBS 55 GNMA prepayments

ƒRefinancing into conventionals

ƒServicer buy-out

ƒHigher delinquencies REPORTS D N A

S Rolling 90-days delinquent with only one missing payment; I ƒ S Y L changed in 2003 to 90-days delinquent NA A NTS ME Y A P E R

P MBS 56 GNMA delinquencies

14 Conventional (Prime) VA FHA 12

10 e (%) u

8

6 Total Past D

4 REPORTS D N

A 2

S Dec-95 Dec-97 Dec-99 Dec-01 Dec-03 Dec-05 I S Y L NA A NTS ME Y A Source: JPMorgan, MBA P E R

P MBS 57 Prepayment reports: speeds by origination year REPORTS D N A S I S Y L NA A NTS ME Y A

P Source: JPMorgan E R

P MBS 58 Prepayment reports: speeds by WALA REPORTS D N A S I S Y L NA A NTS ME Y A P Source: JPMorgan E R

P MBS 59 Prepayment reports: speeds by servicer REPORTS D N A S I S Y L NA A NTS ME Y

A Source: JPMorgan P E R

P MBS 60 Prepayment expectations REPORTS D N A S I S Y L NA A NTS ME Y

A Source: JPMorgan P E R

P MBS 61 Market Overview and Origination 1

Demand 11

Mortgage Cashflows and Intro to Prepayments 23

Valuation and OAS 32

Prepayments Analysis and Reports 46

TBA Market and Specified Pools 62

Relative Value Trading 74

Case Studies 99 R E

RIM ARMs 110 P S B

M CMOs 116 AN G R

O MBS Index 163 M

JP MBS 62 TBA pass-through market

ƒ TBA = “To Be Announced”. Essentially, a cheapest-to-deliver market (like a futures contract). Most liquid market. ƒ As with other delayed delivery transactions, a seller agrees to issue a TBA security at a future date. However, in a TBA trade, the seller and the buyer do not identify the specific underlying mortgage pools, simply certain pre-specified terms

ƒ TBAs are identified by agency, term, coupon, settle month, and traded on a dollar-price basis ƒ “Cheapest-to-deliver” gives the seller a delivery option that the buyer is short

ƒ Standardized delivery dates (see Bloomberg TDAT).

ƒ Settles once a month like a futures contract S OOL

P ƒ Delivering pools: seller must provide pool information by 3 p.m. 2 business days prior to settlement (48-hour day). IFIED C

SPE ƒ Variance: the amount by which the face value at delivery can vary from the amount

ND specified at the time of the trade, expressed as a percentage of the initial face value A

T requested. The Bond Market Association suggests 0% variance on all TBA trades E K R A M A B

T MBS 63 Bloomberg Generic TBA Tickers S OOL P IFIED C SPE ND A T E K R A M A B

T MBS 64 Dollar rolls

ƒ Dollar rolls are transactions where an institution sells mortgage backed securities with a commitment to buy similar, but not identical, mortgage backed securities on a future date at a lower price.

ƒ In the case of mortgage pass-throughs, “similar” securities refers to securities with the same coupon, security type, and mortgage

S collateral. OOL P ƒ Dollar rolls offer an attractive means of borrowing at a low cost IFIED C primarily because they allow dealers to cover their short positions. SPE ND A ƒ Dollar rolls offer dealers a convenient way to obtain promised T E K

R mortgage securities, avoiding the higher costs of failing to deliver. A M A B

T MBS 65 TBA transactions: evaluating dollar rolls

ƒDrop

ƒCoupon

ƒPrepayments S OOL P ƒDelivery optionality IFIED C

SPE ƒRe-investment rate ND A T E K R A M A B

T MBS 66 TBA transactions: evaluating dollar rolls S OOL P IFIED C SPE ND A T E K R A

M Source: Bloomberg A B

T MBS 67 Roll Specialness

12

10 30-yr CC 15-yr CC 8

6

ll specialness (tks) 4 ro n S 2 upo OOL P 0 IFIED

C -2 Current Co SPE -4 ND

A Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Jan-04 Jan-05 Jan-06 T E K

R Source: JPMorgan A M A B

T MBS 68 Price spreads influence coupon production

TBA 6.0% TBA 6.5% Mortgage note rate (a) 6.88 6.88 Less agency guaranty fee (b) 0.15 0.15 Retained servicing spread (c) 0.25 0.25 Remaining note rate (a-b-c) 6.48 6.48 Excess servicing (a-b-c-e) 0.48 0.00 Buydown of Guaranty Fee (%) 0.00 0.03 Cost to buydown guaranty fee (d) 0.00 0.12

S Desired MBS coupon (e) 6.00 6.50

OOL Market price of MBS (f) 98.05 100.13 P Adjustment to TBA price for low WAC (g) 0.00 0.03 Price adjusted for cost of buydown (f+g-d) 98.05 100.04

IFIED Value of excess servicing (h) 2.09 0.00 C Price adjusted for servicing value (f+g+h) 100.14 100.04 SPE ND A T E K

R Note: Prices are for illustration purposes only A

M Source: JPMorgan A B

T MBS 69 The many dimensions of specified pools

„ WALA

„ FICO

„ Low WAC

„ Low loan balance (LLB)

„ Geographic

„ Prepayment penalty

„ Relocation

S „ Originator OOL

P „ High LTV IFIED C SPE ND A T E K R A M A B

T MBS 70 Specified pools make up roughly 80% of the mortgage market

Spec pool market composition

New Specified 4% New TBA 20%

S Seasoned (>30 WALA)

OOL 49% Moderate P (13-30 WALA) 27% IFIED C SPE

ND Total = $2.814 trillion A Source: JPMorgan T

E Data is as of May 2006 K R A M A B

T MBS 71 New issue specified pools report S OOL P IFIED C SPE ND A T E K R A M Source: JPMorgan A B

T MBS 72 Seasoned specified pools report S OOL P IFIED C SPE ND A T E K R A M Source: JPMorgan A B

T MBS 73 Market Overview and Origination 1

Demand 11

Mortgage Cashflows and Intro to Prepayments 23

Valuation and OAS 32

Prepayments Analysis and Reports 46

TBA Market and Specified Pools 62

Relative Value Trading 74

Case Studies 99 R E RIM

P ARMs 110 S B M CMOs 116 AN G R O

M MBS Index 163

JP MBS 74 Relative value strategies and analysis

Trading Strategies Trading Strategies Evaluation Approaches Mortgage - Swap basis OAS Mortgage - Tsy basis Spread Coupon swap Hedge-Adj Carry 15s / 30s Regressions Ginnie / Fannie Deliverable TBA / Seasoned Sponsorship Agency / Non-agency Pass-through / ARM

G CMO / Collateral RADIN T E VALU E

RELATIV MBS 75 Where to find JPMorgan MBS data

Factor Where to find it Supply Net supply MRV charts, Dataquery Gross supply MRV charts, Daily Packet, Dataquery Demand Banks Federal Reserve W ebsite JPM quarterly bank report

Agencies Fannie, Freddie website

Dealers Bloomberg (PDPPMORT )

Non-US investors TIC data Prepayments by WALA, servicer, etc. JPM Monthly Prepay Appendix MorganMarkets.com OAS TBAs Current Specifieds, strips, hybrids, etc Daily packet Coupon swap history Daily packet Hedge adjusted carry Daily packet (soon)

G Volatility swaption premiums, bp vol Daily packet, Dataquery Regression Historical OAS by relative coupon JPM trading bloomberg Coupon swaps, butterflies MBS Analyzer

RADIN MBS Index/performance Performance by coupon vs swaps and Treasuries Index Monitor T Dataquery E Daily packet TBA Performance Report Dataquery

VALU Other Analytics/historical data Daily packet

E MBS Analyzer

RELATIV MBS 76 Mortgages have widened back to the widest levels since the beginning of the year – but are they fundamentally cheap?

30-year current coupon OAS (bps)

60

50

40

30

20

10

0 G (10) RADIN

T (20) E (30)

VALU 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 E Source: JPMorgan

RELATIV MBS 77 Declining long-dated vol has caused nominal spreads and OAS to diverge

OAS vs. CC ZV spread (bps)

„ Current coupon OAS has been 0 80 OAS ZV spread (5) relatively stable, while nominal 70 spreads have continued to tighten (10) 60 recently … (15)

(20) 50 Dec-05 Jan-06 Feb-06 Mar-06 Apr-06 May-06 Jun-06

3YX10Y swaption premium (bps) „ … Declining long-dated vol has 475 465 G

been the driver (bps) 455 m u

i 445 435 RADIN

T 425 on Prem i t E p 415 405 Swa Y 0 VALU

1 395 x Y E

3 385 375 Dec-05 Jan-06 Feb-06 Mar-06 Apr-06 May-06 Jun-06 Source: JPMorgan RELATIV MBS 78 Mortgage / Swap basis G RADIN T E VALU E

RELATIV MBS 79 30- and 15-year current coupon OAS G RADIN T E VALU E

RELATIV MBS 80 MBS Fixed-Rate Daily Analytics

COB June 9, 2006 G RADIN T E VALU E

RELATIV MBS 81 Hedge-adjusted carry

„ We introduce a hedge-adjusted carry methodology which provides a short- term (1- to 2-month) measure of performance using rather straightforward calculations

„ Option-adjusted spreads serve as a long-term spread measure and assume vega hedging, among other risk measures

„ It serves as a good complement to OAS valuations, capturing the return for taking duration and convexity risks

G „ Essentially, hedge-adjusted carry gives us the net carry of holding a TBA after hedging for duration and adjusting for convexity hedging costs RADIN T E „ Higher carry could suggest overweighting certain coupons. However, this

VALU valuation ignores other risks like vega and mortgage Libor spread E

RELATIV MBS 82 Hedge-adjusted carry components

„ The valuation incorporates factors such as the roll, hedge ratios, durations,

convexities. Specifically, the main components of the analysis are:

„ Swap Hedge Ratios (using partial durations).

„ Duration Hedged Carry

„ Total Negative Convexity

„ Convexity Hedging Cost G RADIN T E VALU E

RELATIV MBS 83 Hedge-adjusted carry components

„ We hedge three partial durations on the swap curve (2-yr, 5-yr and 10-yr) with the cost being the carry and rolldown on each of the swaps

„ To figure out the duration adjusted carry we simply calculate: „ Duration Hedged Carry= TBA roll – ∑(cost of swap hedge) „ Cost of swap hedge = hedge ratio * swap carry

„ In calculating total convexity cost, we incorporate swap convexity since we are long MBS and short swaps (which are positively convex) „ Total Convexity = MBS convexity – swap convexity

„ We can estimate the convexity cost using short-dated swaption implied vols (1-month x 10-year swaptions) 1 ⎡⎛ bp ⎞ 1⎤ ∗C⎢⎜ ⎟ ∗ ()22days ∧ ⎥ ∧ 2 ∗32 2 ⎣⎝ day ⎠ 2⎦

TBA Passthroughs - 30 Year Conventionals TBA Passthroughs - 30 Year Conventionals Roll Dur He dge d Roll Dur He dge d Cnv. Hedging Hedge Adj. (32nds) Carry (32nds) Carry Cost Carry G Sec Price .. (32nds) Sec Price .. (32nds) (32nds) (32nds) FN 30 4.5 91-00 1.25 .. 0.1 FN 30 4.5 91-00 1.25 .. 0.1 -0.7 -0.6 FN 30 5.0 93-29 1.87 .. 0.8 FN 30 5.0 93-29 1.87 .. 0.8 -1.1 -0.4 RADIN FN 30 5.5 96-12 2.25 .. 1.2 FN 30 5.5 96-12 2.25 .. 1.2 -1.6 -0.4 T FN 30 6.0 98-25+ 3.00 .. 2.1 FN 30 6.0 98-25+ 3.00 .. 2.1 -1.9 0.3 E FN 30 6.5 100-29 3.25 .. 2.6 FN 30 6.5 100-29 3.25 .. 2.6 -2.1 0.4 FN 30 7.0 102-14+ 2.00 .. 1.5 FN 30 7.0 102-14+ 2.00 .. 1.5 -2.1 -0.6 VALU E Source: JPMorgan (as of June 1, 2006)

RELATIV MBS 84 Hedge Adjusted Carry Report

New York J.P. Morgan Securities Inc. Jun 05, 2006 MBS Research (1-212) 834-3121 MBS Research morganmarkets.jpmorgan.com MBS Hedge Adjusted Carry

TBA Passthroughs - 30 Year Conventionals Settle Dates Spot = [Jun-13-2006], 1M = [Jul-13-2006], 2M = [Aug-14-2006] Roll Dur Hedged Carry Cnv. Hedging Cost Hedge Adj. Carry (32nds) Hedge Ratio / Swap (32nds) Total (32nds) (32nds) Sec Price 1m 2m OAD OAS 2Y 5Y 10Y 1m 2m Cnvx 1m 2m 1m 2m FN 30 4.5 91-21 1.25 2.25 6.0 -5 0.21 0.32 0.49 0.3 0.6 -1.2 -0.8 -1.6 -0.4 -1.0 FN 30 5.0 94-16+ 1.87 3.37 5.5 -6 0.28 0.33 0.42 1.0 1.8 -1.9 -1.2 -2.5 -0.2 -0.6 FN 30 5.5 96-29+ 2.12 4.12 4.8 -7 0.37 0.32 0.34 1.3 2.7 -2.4 -1.6 -3.2 -0.3 -0.5 FN 30 6.0 99-07+ 3.00 6.00 3.8 -9 0.49 0.25 0.23 2.4 4.9 -2.8 -1.8 -3.7 0.6 1.2 FN 30 6.5 101-05+ 3.25 6.50 2.9 -1 0.51 0.20 0.15 2.7 5.7 -3.2 -2.0 -4.2 0.7 1.5 FN 30 7.0 102-19 2.00 4.00 1.9 0 0.51 0.11 0.07 1.7 3.5 -2.9 -1.9 -3.9 -0.2 -0.4

TBA Passthroughs - GNMA I 30 Years Settle Dates Spot = [Jun-19-2006], 1M = [Jul-18-2006], 2M = [Aug-17-2006] Roll Dur Hedged Carry Cnv. Hedging Cost Hedge Adj. Carry (32nds) Hedge Ratio / Swap (32nds) Total (32nds) (32nds) Sec Price 1m 2m OAD OAS 2Y 5Y 10Y 1m 2m Cnvx 1m 2m 1m 2m GN 30 4.5 92-29 2.00 4.00 5.4 -6 0.24 0.30 0.42 1.2 2.6 -1.2 -0.7 -1.5 0.5 1.1 GN 30 5.0 95-26+ 2.00 4.00 5.2 -23 0.29 0.32 0.40 1.2 2.6 -1.8 -1.1 -2.3 0.1 0.3 GN 30 5.5 97-31 2.50 5.00 4.8 -20 0.37 0.32 0.34 1.8 3.8 -2.4 -1.5 -3.1 0.3 0.7 GN 30 6.0 100-06+ 2.50 5.00 3.7 -20 0.47 0.26 0.22 1.9 4.0 -2.9 -1.8 -3.7 0.2 0.3 GN 30 6.5 102-03 3.00 6.00 2.3 -21 0.50 0.17 0.09 2.7 5.5 -2.7 -1.7 -3.5 1.0 2.0 GN 30 7.0 103-14 0.00 0.00 1.4 -31 0.52 0.09 0.01 -0.2 -0.3 -2.3 -1.4 -3.0 -1.6 -3.2

TBA Passthroughs - 15 Year Conventionals Settle Dates Spot = [Jun-21-2006], 1M = [Jul-20-2006], 2M = [Aug-21-2006] G Roll Dur Hedged Carry Cnv. Hedging Cost Hedge Adj. Carry (32nds) Hedge Ratio / Swap (32nds) Total (32nds) (32nds) Sec Price 1m 2m OAD OAS 2Y 5Y 10Y 1m 2m Cnvx 1m 2m 1m 2m FN 15 4.0 93-00 0.50 1.00 4.3 -1 0.30 0.35 0.25 -0.1 -0.1 -1.0 -0.6 -1.3 -0.7 -1.3 RADIN

T FN 15 4.5 95-06+ 0.75 1.25 4.3 -15 0.34 0.35 0.25 0.1 0.2 -1.3 -0.8 -1.7 -0.7 -1.5 FN 15 5.0 97-01 1.75 3.75 4.0 -15 0.40 0.33 0.22 1.1 2.7 -1.7 -1.1 -2.2 0.1 0.5 E FN 15 5.5 98-28+ 2.00 4.00 3.4 -14 0.48 0.28 0.17 1.5 3.1 -2.1 -1.3 -2.8 0.2 0.4 FN 15 6.0 100-26 3.50 6.50 3.0 -8 0.48 0.23 0.14 3.0 5.8 -2.1 -1.3 -2.7 1.7 3.0 FN 15 6.5 101-28 2.00 4.00 2.2 5 0.51 0.18 0.06 1.7 3.5 -1.8 -1.1 -2.3 0.6 1.2 VALU

E Rolldown and Carry of Swaps between settles Sw ap 1m 2m Dur 2Y 0.29 0.36 1.89 5Y 0.82 1.46 4.38 10Y 1.25 2.32 7.68

RELATIV MBS 85 Performance vs. swaps and Treasuries G RADIN T E VALU E

RELATIV COB June 9, 2006 MBS 86 MBS Issuance Report G RADIN T E VALU E

RELATIV MBS 87 MBS coupon price spread G RADIN T E VALU E

Source: JPMorgan

RELATIV MBS 88 MBS butterfly price spread G RADIN T E VALU E

Source: JPMorgan

RELATIV MBS 89 Mortgage/swap basis vs Index G RADIN T E VALU E

Source: JPMorgan

RELATIV MBS 90 Mortgage/Agency basis vs Index G RADIN T E VALU E

Source: JPMorgan

RELATIV MBS 91 Longer-dated Vol vs Index G RADIN T E VALU E

Source: JPMorgan

RELATIV MBS 92 Shorter-dated Vol vs Index G RADIN T E VALU E

Source: JPMorgan

RELATIV MBS 93 30-year relative coupon OAS G RADIN T E VALU E

Source: JPMorgan

RELATIV MBS 94 30-year relative coupon swaps G RADIN T E VALU E

Source: JPMorgan

RELATIV MBS 95 15-year relative coupon OAS G RADIN T E VALU E

Source: JPMorgan

RELATIV MBS 96 15-year relative coupon swaps G RADIN T E VALU E

Source: JPMorgan

RELATIV MBS 97 Trust IO analytics G RADIN T E VALU E

RELATIV MBS 98 Market Overview and Origination 1

Demand 11

Mortgage Cashflows and Intro to Prepayments 23

Valuation and OAS 32

Prepayments Analysis and Reports 46

TBA Market and Specified Pools 62

Relative Value Trading 74

Case Studies 99 R E RIM

P ARMs 110 S B M CMOs 116 AN G R O

M MBS Index 163

JP MBS 99 Case Study #1: FNMA 6.5s – The Deliverable The relationship between delivery size and Over $400mm TBA 6.5 pools prepaid above paydown cost is non-linear; smaller trades 60CPR are far more exposed to adverse selection

Distribution of 1-mo speeds on 3-17 WALA FNMA 6.5s, Aggregate 1-month speeds on TBA 6.5s (3-17 WALA) for different (April 2006) delivery size (based on Apr prepayments)

2500 100 90 2000 80 l i

% 70 ,

$m 1500 R ce, P 60 n

a 1000 l o C a 50 B 500 1m 40 30 0 20 60+ 50-60 40-50 30-40 20-30 10-20 0-10 CPR Range 0 1000 2000 3000 4000 5000 6000 Balance, $mil S E I

D •There are still a lot of fast pools available in TBA 6.5s STU SE

CA MBS 100 Case Study #1: FNMA 6.5s – Issuance and supply

•FN 6.5 supply should be robust

FNMA 30-year relative coupon issuance Issuance by coupon as a % of total FNMA 30 year issuance

17,000 50% ) R2 = 0.8045 2 M 14,000 R = 0.8793 M 40% ($

e 11,000 FN 6.5 c 30% n Issuance Projection ($bn) a u 8,000 Apr-06 3.8 s 20% 30 Issuance s

May-06 5. 1 N r I 5,000

-y Jun-06 6. 7 10% 0

3 2,000 Jul-06 8 .9 % of F

N 0% F (1,000) (1.00) (0.50) 0.00 0.50 1.00 1.50 2.00 (10)% (1.00) (0.50) 0.00 0.50 1.00 1.50 2.00 Relative Coupon Relative coupon

Source: JPMorgan, FNMA Source: JPMorgan, FNMA Note: Monthly issuance of FN 5s through 6.5s plotted as relative Note: Since September 2005 coupons vs CC, since September 2005 FN 6.5 Issuance Projection ($bn) S E

I Apr-06 3.8 D May-06 5.1

STU Jun-06 6.7

SE Jul-06 8.9

CA MBS 101 Case Study #1: FNMA 6.5s – Fundamentals

Short WALA 6.5s have tight OAS and low SATO •FNMA 6.5s: New OAS (left axis, bps) and SATO (right axis, bps) on issue pools have FN 6.5s by WALA (in months) worse loan

04812 16 20characteristics than 0 110 (2) 100 more seasoned (4) 90 pools (6) 80 (8) 70 (10) 60 (12) 50 (14) 40 OAS SATO Source: JPMorgan, FNMA

FNMA 6.5s are fair fundamentally LIBOR static spread of FNMA 6.5s by CPR, in bp S E I

D CPR 10 20 30 40 50 STU 101—01 73 60 44 25 2 SE

CA MBS 102 Case Study #2: DW 4.5s – Regressions

The Dwarf 5 / 4.5 is cheap historically Rich / cheap of the Dwarf 5 / 4.5 swap Dwarf 5/4.5 swap (y-axis, in ticks) versus 15-year CC yield Residual of Dwarf 5 / 4.5 swap vs 15-year current coupon yield, in (x-axis, %) ticks

70 10 8 65 6

60 4 2 55 0 -2 50 -4 -6 45 -8 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 Feb06 Apr06 Jun06 S E I D STU SE

CA MBS 103 Case Study #2: DW 4.5s – Fundamentals

Outstanding balance and 1-month CPR Outstanding balance and 1-month CPR of Dwarf 4.5s by WALA of Dwarf 5s by WALA

12 20 14 35 1mo CPR 18 1-mo CPR 10 16 12 30 B)

Cum bal ) $ 14 e ($ B) Cum Bal (

8 ) 10 25 e

12 c n a l 6 10 R (% 8 20 Balanc CPR (% e

8 Ba CP mo

o 6 15 ve - 4 lativ ti 1 6 u a l 1-m 4 m 4 10 2 u mu C 2 u

2 5 C 0 0 1 3 5 7 9 11131517192123 0 0 WALA (mos) 1 3 5 7 9 11131517192123 WALA (mos)

Dwarf 5s are fundamentally cheaper than Dwarf 4.5s .LIBOR static spread on Dwarf 4.5s and 5s by CPR

S LIBOR Static Spread (bp) E

I Price 6 CPR 7 CPR 8 CPR 9 CPR 10 CPR

D Dwarf 4.5s 95-08 -7 -3 1 6 10 Dwarf 5s 96-31 6 9 11 14 17 STU SE

CA MBS 104 Case Study #3 (Seasoned pricing): Lower the dollar price, higher the price spread between seasoned and new WALA pools

Payups by WALA vs. TBA Price

14

12 $96 $97 $98

10

8

6

Payup, 32nd 4

2

0 0 5 10 15 20 25 30 35

S WALA E

I Pricing assumptions: D 1. 24-month aging ramp 2. Peak speed of 9% CPR for $96 TBA 3. Peak speed of 10% CPR for $97 TBA STU 4. Peak speed of 11% CPR for $98 TBA SE

CA MBS 105 Case Study #3 (Seasoned pricing): Peak speeds versus the length of the ramp Payups by WALA vs. peak speed „ The peak speed plays an ($97 TBA, 24-mo ramp) insignificant role when pricing seasoned pools 12.0 9CPR 10CPR 11CPR 10.0 8.0 32nd 6.0 (Payup versus TBA assuming up, constant LIBOR static spread) Pay 4.0 2.0 0.0 Payups by WALA vs. aging ramp 0 5 10 15 20 25 30 35 ($97 TBA price) WALA

34mo ramp 24mo ramp 14mo ramp 14 12 „ However, the length of seasoning 10 ramp is crucial

32nd 8 p, u 6 „ A longer seasoning ramp leads to S Pay E I 4 higher payups for seasoning D 2

STU 0 0 5 10 15 20 25 30 35

SE WALA

CA MBS 106 Case Study #3: Aging ramp on moderate discount collateral

„ Moderate discounts have two peaks: 14 months and 24 months „ The 24 months peak is driven by „ Tax-advantaged capital gains treatment of two-year old primary residences „ Cash-out refinancing has been front-loaded

Aging curve: moderate discount, -25 to -75bps incentive

16

12

8 o CPR, % 1m 4 2000~2001 2005~2006 S

E 0 I

D 0 5 10 15 20 25 30 35 40 Age STU

Note: Average prepayment aging curves observed from 2000 to 2001 and from 2005 to March 2006 SE

CA MBS 107 Case Study #3: Specified Pools WALA- Actual versus theoretical payup

„ Seasoned payups suggest that the market is priced to a 24 months ramp

„ Potential lengthening of the aging ramp should translate to higher payups for seasoned discounts

„ Relative value advantage in the’04 discounts versus the ’03s

FNMA 5: Actual vs. theoretical FNMA 5.5: Actual vs. theoretical

’05–’06 ramp (24-mo.) ’00–’01 ramp (34-mo.) 34-mo. ramp 24-mo. ramp 14-mo. ramp 18 16 16 14 14 12 12 10 32nd 10 p, 8 u 8 6 Payup, 32nd

Pay 6 4 4 2 2 0 0 S 5 101520253035 0 5 10 15 20 25 30 35 E I WALA WALA D

Note: As of: 5/3/2006; FNMA 5s are priced to 9% CPR terminal speed; FNMA 5.5s are priced to 10% CPR terminal speed STU SE

CA MBS 108 Case Study #4: Specified Pools LLB – Call/extension protection

The trend points to faster LLB discount speeds similar to 1999-2000 Prepayment S-Curve differences (LLB (<$100k) minus generics ($>150k)), in 1999-2000, 2004, 2005 and since fall 2005: 1mo CPR vs. Rate Incentive,bps.

1moCPR 2 1999~2000 2004 2005 2005 Oct~2006 Apr 1

0

-1

-2

-3 Incentive, bps -125 -100 -75 -50 -25

Low loan balance discounts are priced at a small fraction of their combined call and extension value.

Payups and model valuations of MLB ($110k max) pools Call TBA Market Protection 10% Faster Turnover 20% Faster Turnover

S Coupon Price Payup Payup Ext. Protection Total Ext. Protection Total E I

D FNCL 5 94-05 0.5 2 9 11 17 19 FNCL 5.5 96-22+ 1.5 3 6 9 11 14 STU FNCL 6 99-05 4 7 4 11 7 14 SE

CA MBS 109 Market Overview and Origination 1

Demand 11

Mortgage Cashflows and Intro to Prepayments 23

Valuation and OAS 32

Prepayments Analysis and Reports 46

TBA Market and Specified Pools 62

Relative Value Trading 74

Case Studies 99 R E RIM

P ARMs 110 S B M CMOs 116 AN G R O

M MBS Index 163

JP MBS 110 ARM share has remained close to half of total applications

ARM share of total applications by $ volume (%)

60

50 ) % ( e m u l 40 $ Vo y b s

p 30 Ap f o e

ar 20 h S M AR 10

0 Jun- 97 Jun- 98 Jun- 99 Jun- 00 Jun- 01 Jun- 02 Jun- 03 Jun- 04 Jun- 05 Jun- 06

Source: MBA S

ARM MBS 111 Hybrids contributed to most of the growth of the non-agency market

Year-end outstandings ($ billions)

800 Jumbo Fixed Jumbo ARM Alt-A Fixed Alt-A ARM 700

600

500

400

300

200

100

0 1998 1999 2000 2001 2002 2003 2004 2005 2006

Source: JPMorgan, Loan Performance. 2006 data as of April. S

ARM MBS 112 Hybrid ARM Structure and Valuation

Example Structure Example

„ Typically 30-year terms ƒ 5/1 Hybrid ARM borrower with an initial coupon rate of 4.5% and a 5/2/5 cap „ Fixed Rate Cash-Flows structure pays 4.5% for 5 years „ 3, 5, 7,10 Year Fixed-Rate ƒ The highest the borrower’s rate can reset after the 5 year fixed rate period ends is (4.5%+5%), or 9.5% „ Hybrid ARM Tails „ Libor/CMT Floaters ƒ The on-going periodic cap restricts the borrower from resetting up more than 2% at each yearly reset „ Resets subject to Caps ƒ The life cap also protects the borrower (initial/periodic/life) by ensuring that the coupon cannot „ 2/2/6 reset above 9.5% for the life of the loan „ 5/2/5 S

ARM MBS 113 Hybrid ARM Key Terminology

ƒRate caps (initial, periodic, and lifetime) offer protection from large interest rate movements by providing a cap and a floor, limiting the amount the resetting contract rate can increase or decrease on each adjustment date ƒThe initial adjustment cap provides a cap and floor on the interest rate at the first adjustment date ƒPeriodic adjustment caps restrict upward and downward movements at each subsequent reset date ƒ Lifetime caps dictate the maximum interest rate of the mortgage loan at any given time ƒMTR (months to reset) is defined as the number of months until the Hybrid ARM resets off its specified Index. In other words, the number of months until the fixed rate portion of the bond ends ƒOnce a Hybrid ARM loan reaches reset, the borrower’s new coupon is determined off a specified index ƒThe most popular indices are LIBOR and CMT ƒThe gross margin is the spread added to the Index that determines the mortgage holder’s new rate ƒ The net margin is what is passed on to the investor. The net margin is the spread added to the Index that determines the coupon the investor receives. Typically, unless otherwise specified, LIBOR-indexed ARMs have approximately 175bps net margin, while CMT-indexed bonds have approximately 225bps margin S

ARM MBS 114 How are Non-Agency MBS Valued and Traded

„ Non Agency MBS do not trade ‘in the screens’ thus valuation and pricing levels reflect this uncertainty

„ Trading is Negotiated around structure, settlement and collateral composition

„ Relative Value and performance analytics help drive trading levels

„ Pricing Conventions „ N-A Fixed Rates; Pass throughs relative to agency pass throughs „ N-A ARMs : Swaps, Treasuries

„ Prepayment Analytics „ OAS Methodology : Prepayment Modeling

„ Credit Analytics

„ Liquidity Premiums S

ARM MBS 115 Market Overview and Origination 1

Demand 11

Mortgage Cashflows and Intro to Prepayments 23

Valuation and OAS 32

Prepayments Analysis and Reports 46

TBA Market and Specified Pools 62

Relative Value Trading 74

R Case Studies 99 E RIM P ARMs 110 S B M CMOs 116 AN G R O

M MBS Index 163 JP MBS 116 FNMA Current Coupon Yield vs. UST 10 Year Yield

Current Coupon 10 yr

6.5

6

5.5

5 d l e Yi 4.5

4

3.5

3 6/2/2003 12/1/2003 5/31/2004 11/29/2004 5/30/2005 11/28/2005 5/29/2006

CMOS MBS 117 Why Agency CMOs?

„ To broaden the investor base by customizing cash flows for investor needs while providing key advantages over other instruments: „ Excess Returns „ Greater Liquidity „ Virtually Zero Credit Risk

„ Can address the following specific needs of investors: „ Enhanced Yields/Spreads „ Targeted Average Life Profiles „ Targeted Duration Profiles „ Customize the risk/reward profiles for investor’s views on: Î interest rates Î yield curve shape Î prepayments Î volatility

CMOS MBS 118 CMOs as % of the Fixed Rate MBS Market

CMO Issuance % of Pass-Through

70 70 e

60 60 c n a ssu I s) 50 50 n h o g u lli i ro B 40 40 h $ T ( e ss- c a n P a 30 30 e t a ssu R I - O d M 20 20 e C x i F f

10 10 o %

0 0 Jan-01 Jan-02 Jan-03 Jan-04 Jan-05

CMOS MBS 119 CMO Principal Types

„ SEQ – Sequential

„ PAC – Planned Amortization Class

„ TAC – Targeted Amortization Class

„ AD or VADM- Accretion-Directed/Very Accurately Defined Maturity

„ FFIEC Bonds

„ SUP - Support or Companion

„ Z - Accrual bond

CMOS MBS 120 SEQs - Sequentials

„ Collateral principal payments are reallocated sequentially into a series of short, intermediate and long maturity bonds

„ Sensitive to prepayments:

Î Prepayments faster than expected: SEQs shorten

Î Prepayments slower than expected: SEQs extend

„ The shorter average life sequentials (that pay before the longer SEQs within the structure) provide prepayment protection for the longer average life SEQs

„ Shorter principal window than collateral

CMOS MBS 121 SEQ Yield Tables

Front SEQ

Last Cashflow (LCF) SEQ

CMOS MBS 122 PACs - Planned Amortization Class

„ Principal repaid according to a schedule within a specified range of prepayment assumptions called PAC bands

„ Principal schedule provides protection from average life volatility and reinvestment risk associated with prepayments

„ Principal schedules are maintained by redirecting cashflow uncertainty to Support bonds

„ Average Life is less volatile with speeds outside the bands because the supports continue to provide stability

„ and agency bond surrogate

CMOS MBS 123 PAC Yield Tables and Cashflow Graph

Yield table

Cashflow Graph

CMOS MBS 124 TACs - Targeted Amortization Class

„ Structured to pay principal according to a schedule determined by one constant prepayment speed - a “one sided” PAC

„ No protection against extension: TACs only have call protection because a TAC provides protection against faster but not slower prepayments

Î Prepayments faster than TAC speed: Excess principal to supports

Î Prepayments slower than TAC speed: TAC and Support extend

„ Offer higher yield than PACs based on increased extension risk

„ More call protection than a SEQ

Yield Table

CMOS MBS 125 AD/VADMs - Accretion Directed / Very Accurately Defined Maturity

„ All cashflows are derived from the interest accretions of the Z bond

„ VADM tranches must mature prior to the start of the amortization on their corresponding Z because when the Z becomes current pay the Z accretion is no longer available to amortize the VADMs

„ Very stable bond since cashflow is from interest accretion which is NOT affected by prepayments

„ VADMs do not extend even under a zero prepayment scenario

„ No whipsaw risk

„ Pay up for the extension protection results in lower yields and better convexity

CMOS MBS 126 VADM Yield Table and Cashflow Graph

VADM yield table in SEQ Z structure

WAL Graph

CMOS MBS 127 FFIEC Bonds

„ Federal Financial Institutions Examination Council (FFIEC) derived guidelines to determine if MBS investments are suitable for US depository institutions

„ Bonds that meet these guidelines have a wider audience, are more liquid, and trade at tighter spreads than comparable non-FFIEC average life bonds

„ FFIEC test:

Test 1) WAL must be less than 10 years

Test 2) +300 shift Æ less than 4 years extension -300 shift Æ less than 6 years contraction

Test 3) +300 shift Æ less than 17% price change -300 shift Æ less than 17% price change

CMOS MBS 128 SUPs - Supports or Companions

„ Supports are cashflow shock absorbers for PACs

„ Faster prepayments - excess cash flow paid to supports providing call protection for PACs

„ Slower prepayments - any shortfall in cash flow is absorbed by supports which may not receive principal until PAC schedule is met, providing extension protection for PACs

„ High average life and cashflow volatility

„ Higher yields compensate for volatility

CMOS MBS 129 Support Yield Table and Weighted Average Life

Sequentially Tranched Support

Weighted Average Life Graph

CMOS MBS 130 Z bonds - Accrual Bonds

„ No interest until principal payment window starts

„ Interest due is added to outstanding principal of the bond = Z accretion

„ Z accretion accelerates the maturity of shorter tranches or more accurately defines the maturity of others (VADMs)

„ Receive principal payments and interest once other bonds are retired

„ Addition of Z bond to a structure can improve the convexity of the other bonds by reducing extension risk

CMOS MBS 131 Sequential Z Bond Yield Table & Cash Flow Graph

Yield Table

Cash Flow Graph

CMOS MBS 132 Support Z Bond Yield Table & Cash Flow Graph

Yield Table

Cash Flow Graph

CMOS MBS 133 Rocket Z Yield Table & Cash Flow Graph

„ Type of support Z that has a possibility of paying off very fast (“rocket”).

Yield Table

Weighted Average Life Graph

CMOS MBS 134 CMO Interest Types

„ Floater

„ Inverse Floater

„ IO - Interest Only (Trust and Structured)

„ PO - Principal Only (Trust and Structured)

„ Inverse IO

CMOS MBS 135 Floaters

„ Coupons reset periodically, usually monthly, at a rate of an index, usually 1mLibor, plus a spread, know as the margin

„ Created with an inverse or inverse IO such that the weighted-average coupon of the pair is always equal to the underlying fixed-rate bond

„ Shorter and less negatively convex than underlying fixed rate

„ Offers protection against interest rate risk up to the cap

„ The cap adds some duration and negative convexity. Without a cap the duration would be negligible and convexity neutral

„ Shorter durations= less price volatility than fixed-rate CMOs unless rates rise and the coupon reaches its cap

CMOS MBS 136 Floater Yield Table & Coupon Graph

Yield Table

Coupon Graph

CMOS MBS 137 Support Floater Yield Table & Coupon Graph

Yield Table

Coupon Graph

CMOS MBS 138 Inverse Floaters

„ Pays down simultaneously with their corresponding floater

„ Coupon falls when the index rate rises and are typically levered positions in the underlying fixed rate cash flow

„ High yields frequently compensate for the increased risks

„ Floored inverses: a “baby” inverse floater - coupon moves inversely but can never drop below a designated fixed coupon

„ Provide a way to leverage MBS if you are bullish on the MBS sector

„ Provide a way to leverage views in one package if you disagree with Î forward rate curve Î FED expectations Î prepayment forecasts Î volatility views

CMOS MBS 139 Inverse Yield Table & Coupon Graph

Yield Table

Coupon Graph

CMOS MBS 140 Floored Inverse Yield Table & Coupon Graph

Yield Table

Coupon Graph

CMOS MBS 141 IO - Interest Only

„ Receive only interest cash flow from the notional amount of the underlying bond

„ Since IOs do not pay principal, cashflows exist only if principal remains outstanding

„ The amount of cash flows received is highly dependent upon the amortization and prepayments of the underlying class

„ Benefit from slowing prepayments Î Faster prepayments reduce the notional balance more rapidly leading to smaller interest payments

„ Bearish security that usually have negative durations

„ Interest stripped from designated pools of collateral are known as STRIPS and are traded in liquid broker markets

„ Interest stripped from CMOs are known as Structured IO and can be customized for investor needs on: Î lockout Î PAC bands Î underlying collateral STIPs

CMOS MBS 142 IO Yield Tables

Structured PAC IO

Trust IO

CMOS MBS 143 PO - Principal Only

„ Stream of principal payments purchased at a discount

„ Hedge for prepayment risk since POs benefit from faster prepayments: Î Principal is returned at par at a faster rate Î Lower discount rates boost the price

„ The amount of cash flows received is highly dependent upon the amortization and prepayments of the underlying class

„ Bullish security with large, positive duration and positive convexity

„ Super PO’s provide a more levered prepay bet

„ Principal stripped from designated pools of collateral are known as STRIPS and are traded in liquid broker markets

„ Principal stripped from CMOs are known as Structured PO and can be customized for investor needs on: Î lockout Î PAC bands Î underlying collateral STIPs

CMOS MBS 144 PO Yield Tables

Structured SUP PO

Trust PO

CMOS MBS 145 IO/PO Reports

„ Trust IO/PO reports from JPMorgan provide daily price data:

CMOS MBS 146 Inverse IO

„ Pays down simultaneously with their corresponding floater like an inverse except the Inverse IO does not pay principal

„ Coupon falls when the index rate rises and are typically levered positions in the underlying fixed rate cash flow

„ Faster prepayments reduce the notional balance more rapidly leading to smaller interest payments

„ Provide a way to leverage views in one package if you disagree with Î forward rate curve Î FED expectations Î prepayment forecasts Î volatility views

CMOS MBS 147 Inverse IO Yield Table and Coupon Graph

Yield Table

Coupon Graph

CMOS MBS 148 Recent Innovations in CMO Market

„ AS/NAS – Accelerated Security/Non-accelerated Security

„ TTIB – Two Tiered Index Bonds

„ Super-Floater

„ Customized Floater: FHR 3069 CF

„ RELO – Relocation collateral deals

„ Pre-pay Linked Notes/Interest Accrual Notes (IANs)

„ Freddie Mac Reference Notes

CMOS MBS 149 AS and NAS - Accelerated and Non-Accelerated Securities

„ AS security receives principal payments more quickly than its respective collateral.

„ NAS Security receives principal more slowly than its respective collateral

„ NAS + AS = SEQ

„ The AS bond receives the “accelerated” principal payments that would have otherwise been allocated to the NAS bonds

„ The NAS bond is locked out until the AS bond is paid off; then the NAS begins receiving its pro-rata principal payments

„ NAS is better than a PAC = no whipsaw risk

CMOS MBS 150 AS and NAS Yield Tables

AS

NAS note the average life of 5.30 at 0 PSA

CMOS MBS 151 TTIB - Two Tiered Index Bond

„ A type of inverse floater that pays a fixed rate as long as 1m Libor stays below a certain threshold

„ Once 1m Libor crosses the threshold the coupon declines on a levered basis within a corridor of rates until it reaches 0%

„ Essentially shorting an option that 1m Libor will not increase beyond a certain threshold

„ Compensating for shorting the option by getting a higher coupon.

Historical Digital TTIBs: Once 1m Libor 1m Libor crosses the graph threshold the coupon declines immediately to 0%

Floored TTIB: Once 1m Libor crosses the threshold the coupon declines immediately to a fixed rate floor

CMOS MBS 152 TTIB Yield Tables and Coupon Graphs

Digital TTIB yield table Digital TTIB coupon graph

1 bp Corridor TTIB yield table 1 bp Corridor TTIB coupon graph

CMOS MBS 153 TTIBs with additional features

„ TTIBs with lock-out: coupon rate is not conditional on 1m Libor for an initial period, ensuring desirable rates over that period.

„ Initial reset date can be several years into bond’s lifetime.

Locked-out Digital TTIB yield table Locked-out Digital TTIB coupon graph

CMOS MBS 154 TTIBs with additional features (cont.)

„ Floating-rate TTIBs: when 1m Libor is below a threshold, bond pays as a floater.

„ Floating-rate TTIBs with lock-out: bond maintains its initial coupon formula for an initial period, regardless of 1m Libor.

Example: FHR 3140 CF - Is a regular L + 165 bp bond for first 7 years, before converting into a regular floating-rate TTIB.

CMOS MBS 155 Super-Floater

„ A type of floater that pays a fixed rate as long as 1m Libor stays above a certain threshold (usually greater than current levels).

„ Essentially shorting an option that 1m Libor will remain below a certain threshold.

„ Compensated for shorting the option by receiving a VERY high coupon should 1m Libor go above the threshold.

„ Example: FHR 3111 HF (receives 66% coupon if 1m Libor > 6.5%)

Coupon Graph

CMOS MBS 156 Customized Floater: FHR 3069 CF

„ Classified on Bloomberg as “Complex” because the formula for calculating the payment is not the standard Libor + discount margin

„ Unique structure

„ Payment Formula: Î If Libor is less than 4.8%, bond pays Libor + 2.35% Î If Libor is greater than 4.8% but less than 7.15%, bond pays 7.15% Î If Libor is greater than 7.15%, bond pays 0%

Coupon Graph

CMOS MBS 157 RELO – CMO Backed by Relocation Mortgages

„ Relocation Mortgage: a mortgage made to a transferred employee to finance a home purchase at a new job location

„ Mortgage usually requires an employer to contribute to mortgage funding

„ Mortgage typically originated by an agreement between the employer and the lender under a relocation program administered by the employer or its agent

„ Prepayment speeds depends on typical prepayment behaviors and other RELO factors: Î Whether the mortgages are made in connection with a permanent relocation of a corporate headquarters Î The likelihood that borrowers will be relocated again Î The frequency with which further relocations may occur

„ Historically this sector has fast prepay speeds

„ CMOs backed by RELO collateral usually trade at a deep discount

CMOS MBS 158 RELO Yield Table and Weighted Average Life Graph

Yield Table

Weighted Average Life graph

CMOS MBS 159 Prepay Linked Notes or Interest Accrual Notes (IANs)

„ Agency debt and MBS hybrid

„ Redemption schedule is based on a pre-selected reference pool

„ Like MBS (unlike agency debt) there is no explicit call date

„ Like agency debt (unlike MBS) there is a stated final maturity

„ Effective duration management tool for those who like MBS sector

Yield Table Recently-priced deals

CMOS MBS 160 Freddie Reference Notes

„ A Pre-pay linked note that trades live on Trade Web; an automated broker

„ Availability by all dealers on Trade Web means Î Better liquidity and Î Better price transparency than pre-pay linked notes that do not trade live on Trade Web

„ Trades at slightly lower yields than other pre-pay linked notes due to the advantage of greater liquidity and greater price transparency

Yield Table Trade Web Screen Offering:

CMOS MBS 161 Conclusion

„ As a premier investment bank, strives to be a leader in the CMO market

„ The CMO team’s recent production is growing rapidly: Î In April & May 2006, the #1 FNMA issuer! Î The #3 overall conventional issuer (FNMA + FHLMC) over same period. Î Over $5 billion in deal volume in those two months alone!

„ The CMO team can provide the following client needs: Î Unique trade ideas through structuring capabilities Î Relative value analysis Î Marked-to-market valuations & portfolio analysis Î Liquidity through market making

„ Let the CMO team help you maximize the total return of your portfolio!

CMOS MBS 162 Market Overview and Origination 1

Demand 11

Mortgage Cashflows and Intro to Prepayments 23

Valuation and OAS 32

Prepayments Analysis and Reports 46

TBA Market and Specified Pools 62

Relative Value Trading 74

Case Studies 99 R E RIM

P ARMs 110 S B M CMOs 116 AN G R O

M MBS Index 163

JP MBS 163 JPMorgan MBS Index

„ The JPMorgan Mortgage Index (“the Index”) measures the performance of fixed-rate agency-backed mortgage pass-through securities • Included: – Thirty-year and 15-year fixed-rate pass-through securities issued by FNMA, FHLMC, and GNMA

• Excluded: – ARMs – Non-Agency (whole loan), Jumbo, and 10- or 20-year securities – Balloons, GPMs, and TPMs – FHLMC 75-day delay mortgages – FNMA Mega, FHLMC Giant, and GNMA Platinum pools X E ND I S

MB MBS 164 JPMorgan MBS Index

„Due to paydowns and new pool issuance which are reported monthly, the mortgage universe undergoes monthly transformations

„JPMorgan re-balances the Index automatically to reflect the changing compositions of the mortgage market. Once a month, on the last business day of the month, all the pools represented in the Index are re-aggregated for use next month

„At the end of each business day, JPMorgan trader marks are used to price all the constituent securities in the Index „ Traders mark benchmark issues (TBAs) and specified pools „ Algorithm is used to price securities that are not actively traded X E ND I S

MB MBS 165 Calculation of Index Returns

„ Daily total return of the Index is the market value weighted average of the daily returns of its constituent securities

„ The daily total return of a security is defined as its daily change in market value over its previous market value. The change in market value equals change in price plus change in accrued interest. Interest is accrued daily at the net coupon rate based on 30/360 day-count convention. Using formulas, the relationship is as follows:

„ The Index starts at 100 on January 1st, 2000. On every business day after market close, a daily Index return is computed. A new index value is then calculated as the product of the previous Index value and the daily Index return

„ This methodology assumes that the Index settles daily (on business days) and that returns are reinvested into the Index on a daily basis. Using raw Index values, one can easily calculate the periodic total return of the Index between any two business-days by dividing the ending Index

X value by the starting Index value E ND I S

MB MBS 166 Total return swaps on the JPMorgan MBS Index •Investors can receive (or pay) the total rate of return of the JPMorgan Mortgage Index and pay (or receive) LIBOR – a specified spread

Investor

Index Return LIBOR - spread

JPMorganChase Bank

•Advantages of the TROR Index Swap: – MBS returns competitive and less volatile than corporate debt – Ease of execution – Sector exposure with no security selection required – No MBS delivery / allocation required – Locked-in funding spread – Finite investment term – Paydowns automatically reinvested in Index X

E – Pay or receive Index ND I S

MB MBS 167 Mortgage Index returns X E ND I Source: Bloomberg S

MB MBS 168 Index Composition

Index composition in June 2006, par weighted (%) and month/month changes in 10-year equivalents ($bn) X E ND I Source: JPMorgan S

MB MBS 169 Index Performance

Index duration and convexity profile, as of 5/31/06

Index statistics, May 31st, 2006 Index statistics, May 31st, 2006 X E ND I S

MB MBS 170 Index Performance

Index returns and excess returns by sector and coupon (bps), in May 2006 X E ND I S

MB MBS 171