US.fixed.income.markets.weekly

US Fixed Income Strategy 2 March 2012

US Fixed Income Markets WeeklyCross Sector Srini Ramaswamy, Kimberly HaranoWe expect the impact of tailwinds from the improving situation in Europe to dominate the negatives, and we retain a positive stance in most risky markets. We maintain our bias for higher rates expecting 10-year rates to move modestly higher into mid year. Bill net issuance will drop but stay positive in the coming weeks, keeping front-end Treasuries cheap versus OIS. Switch out of the 6-year sector into a combination of 4s and 8s. Stay bullish on TIPS breakevens. Stay overweight 10-year Agencies versus Treasuries. Stay biased towards wider spreads at the long end of the curve, but neutral on front-end spreads. Pay in 3.25% June 2016 maturity matched swap spreads versus receiving in 3.25% July 2016 spreads. Position for a cheapening of the belly of a ED6/7Y/25Y weighted butterfly via Eurodollar short-Sep midcurve puts and expiry matched payer swaptions. Stay neutral on rate vol but sell curve vol.ContentsCross Sector Overview Economics Treasuries Interest Rate Derivatives Agencies Agency MBS Non-Agency MBS CMBS ABS Corporates High Yield Short Duration CDO Municipals Emerging Markets Special Topic Forecasts & Analytics Market Movers 2 6 11 18 24 28 36 41 49 53 59 65 70 74 85 86 91 96

Governments Terry Belton, Meera Chandan, Kim Harano, Renee Park

Interest Rate Derivatives Srini Ramaswamy, A. Iglesias, P. Korapaty

MBS and CMBS Matthew Jozoff, Ed Reardon, John Sim, Brian Ye

Remain neutral on the mortgage basis owing to narrower valuations and a somewhat reduced outlook for bank / money manager sponsorship. Be down in coupon in conventionals. Be overweight 30-years vs 15-years. The rally in new issue credit CMBS will continue; single-As are our top pick. The ABS market is gathering more steam as credit investors continue to add risk. We continue to see value in subordinate auto and card ABS.

ABS and CDOs Amy Sze, Rishad Ahluwalia, Maggie Wang

Investment-Grade Corporates Eric Beinstein

High Grade credit fundamentals remain solid, but some credit metrics deteriorated in 4Q11 after several quarters of very strong results.

High Yield Peter Acciavatti, Nelson Jantzen, Tony Linares

The fundamentals for high-yield bonds and leveraged loans remain a positive catalyst in the context of valuations. Given the improved market sentiment post the second 3-year LTRO, we revise our 3-month Libor forecast from 45bp to 40bp at the end of 1Q12. However, we think it will be pressured higher before 2Q12 driven by Moodys review of bank and broker-dealer ratings and money fund reform. We discuss costs associated with holding capital on the sidelines awaiting higher yields and provide specific yield targets in a breakeven analysis. Inflows remain robust and valuations still reasonable. Higher oil prices shape our recommendations.Terry BeltonAC(1-312) 325-4650 [email protected]

Short-Term Fixed Income Alex Roever, Teresa Ho

Municipals Peter DeGroot, Josh Rudolph

Emerging Markets Joyce Chang, Ben Ramsey

Special Topic: Carry and Rule-Based Investing Ruy Ribeiro

Carry strategies in bonds posted strong performance in 2011 with Sharpe ratios above 1. Bond momentum in Europe was also strong.

Srini Ramaswamy(1-212) 834-4573 [email protected]

AC

Indicates certifying analyst. See last page for analyst certification and important disclosures.

US Fixed Income Strategy US Fixed Income Markets Weekly March 2, 2012 Srini RamaswamyAC (1-212) 834-4573 Kimberly L. Harano (1-212) 834-4956 J.P. Morgan Securities LLC

Cross Sector Overview The ECBs sizeable provision of term liquidity in this weeks 3-year LTRO, as well as the potential for additional long-term operations in the future if needed, support a reduction in European tail risk, which in turn has been a key driver of the rally in risky assets We expect the upcoming release of the Feds 2012 Comprehensive Capital Analysis and Review (CCAR) to also be positive for markets Economic data has generally remained supportive, though this weeks durable goods and consumer spending reports have added downside risk around our GDP report, and high energy prices are a concern For now, however, we expect the impact of tailwinds from the improving situation in Europe to dominate the negatives, and we retain a positive stance in most markets We remain biased toward modestly higher Treasury yields as we head toward mid-year

Exhibit 1: Risky assets rallied again this week

Current level,* change since 2/24/12, quarter-to-date change, and change over 4Q11 for various market variablesCurrent Chg from 2/24 QTD chg 4Q11 chg Global Equities (level) S&P 500 E-STOXX FTSE 100 Nikkei 225 Sovereign par rates (%) 2Y US Treasury 10Y US Treasury 2Y Germany 10Y Germany 2Y JGB 10Y JGB 5Y Sovereign CDS (bp) Spain Portugal Italy Ireland Funding spreads (bp) 2Y EUR par swap - par gov't spd 2Y USD par swap - par gov't spd EUR FRA-OIS spd USD FRA-OIS spd 1Y EUR-USD xccy basis Currencies EUR/USD USD/CHF USD/JPY JPM Trade-weighted USD Spreads (bp) 30Y CC MBS L-OAS 10Y AAA CMBS spd to swaps JULI portfolio spd to Tsy JPM US HY index spd to worst EMBIGLOBAL spd to Tsy MAGGIE (Euro HG spd to govies) US Financials spd to Tsy Euro Financials spd to govies 10Y AAA muni/Tsy ratio (level) 0.288 2.045 0.130 1.868 0.102 0.975 376 1239 361 598 91.2 23.9 35.6 30.5 -58.3 1.321 0.913 81.63 80.50 40.2 215.0 193.2 619.7 347.7 164.2 241.5 241.5 93% -0.035 0.023 -0.085 -0.075 0.006 0.029 7 78 -29 20 -2.7 -5.6 -14.8 -9.0 3.6 -0.025 0.018 1.06 -0.15 2.9 -15.0 -5.9 -9.0 -17.2 -11.5 -13.3 -18.8 0.9% 0.025 0.103 0.029 -0.025 -0.027 0.069 -6 61 -127 -154 -29.9 -22.7 -35.5 -25.2 41.6 0.029 -0.031 3.92 -1.75 -5.9 -60.0 -43.6 -104.4 -78.6 -63.7 -89.2 -110.7 -0.8% -0.032 -0.080 -0.406 -0.046 -0.017 -0.089 -1 -36 13 12 21.8 18.6 7.7 12.3 -27.3 -0.051 0.037 0.95 0.23 -0.6 -90.0 -17.6 -83.9 -38.6 16.6 -3.4 25.1 -15.6% 1369.6 2546.2 5911.1 9777.0 3.9 22.5 -24.0 129.7 112.0 229.6 338.9 1321.7 126.2 136.9 443.8 -244.9

Market viewsThe rally in risky assets largely continued this week, with stocks rising in the US and Europe and credit spreads tightening across the board (Exhibit 1). Funding spreads also showed improvement, with EUR and USD FRA-OIS spreads narrowing sharply. A key driver of the rally was this weeks 3-year LTRO by the ECB, which attracted higher subscriptions than expected. Eight hundred banks took down 530bn of funding from the ECB, of which we estimate 313bn consists of new funding (see Overview, Global Fixed Income Markets Weekly, 3/2/12). This sizeable provision of term liquidity, as well as the potential for additional long-term operations in the future if needed, support a reduction in European tail risk, which in turn has been a key driver of the rally in risky assets. As Exhibit 2 shows, two key metrics of the crisis we have usedsemi-peripheral CDS spreads and French bank CDS spreadshave retraced to early August 2011 levels, and the potential for some further retracement

30Y AAA muni/Tsy ratio (level) 105% 0.0% -17.1% 1.6% Commodities Gold futures ($/t oz) 1722.20 -64.10 181.30 -74.60 Oil futures ($/bbl) 106.70 -3.07 7.87 19.63 * 3/1/12 level for Europe and US corporate credit spreads, and the J.P. Morgan trade-weighted USD index; 3/2/12 level for all others.

remains. Evidence of reduced tail risk can also be seen in the pickup in primary market activity. This week, the second CMBS deal of the year (the $1bn COMM 2012LC4) priced well inside of initial guidance with all classes (including the AA and single-A subordinates)

2


Exhibit 2: Metrics of the European financial crisis have retraced to August 2011 levelsSemi-peripheral spreads* versus French bank CDS**; bp bp

Exhibit 3: High grade issuance in February was nearly double the 10-year averageAverage gross issuance by calendar month over 2002-2011, versus 2012; $bn

450 400 350 300 250 200 150 100 Jun 11

Semi-peripheral spreads

400 350 300 250 200 French bank CDS 150 100

90 80 70 60 50 40 30

Average

2012

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Jul 11

Sep 11

Oct 11

Dec 11

Feb 12

* Average 5-year sovereign CDS for France, Spain and Italy. ** Average 5-year CDS for BNP Paribas, Credit Agricole and Societe Generale.

Source: Thomson Financial, Bloomberg, and J.P. Morgan

over-subscribed multiple times. Issuance in corporate markets has been very robust as well. As Exhibit 3 shows, issuance of US high grade corporates in February was nearly double the average over the previous ten years. Similarly, high yield saw $40.4bn of gross issuance in February, making it the third most active month on record (see High Yield). Thus, European developmentsa key catalyst of the recent risk-rallyremain a near-term positive. In the US, the approaching release of the bank stress test results is also likely to prove to be supportive of risky assets. As investors will recall, the Fed announced the 2012 review, known as the Comprehensive Capital Analysis and Review (CCAR), on November 22, 2011. In contrast to prior reviews, all U.S. bank holding companies with total consolidated assets of $50bn or more were required to participate this year, raising the total number of institutions reviewed from 19 to 31. Banks were required to submit their capital plans by January 9, 2012, and will hear back from the Fed by March 15. This years review contains 25 stress-test variables (compared to 9 variables in last years review) and includes a global market shock scenario. In Exhibit 4, we compare the stress case scenario to the baseline scenario and the J.P. Morgan forecast for a few select variables. To be sure, the stress scenario defined by the Fed is fairly extreme, compared to recent economic trends as well as our forward looking projections. However, banks capital positions have been steadily

Exhibit 4: The stress scenario in the Feds 2012 review is fairly onerous compared to our forecast

Stress scenario and baseline scenario for select variables in the Federal Reserves 2012 Comprehensive Capital Analysis and Review, versus J.P. Morgan forecast for those variables*; units as marked Stress scenario Baseline scenario J.P. Morgan forecast Real Unem- Cum HPA Real Unem- Cum HPA Real Unem- Cum HPA GDP ployment from 3Q11 GDP ployment from 3Q11 GDP ployment from 3Q11 (% ) rate (% ) (% ) (% ) rate (% ) (% ) (% ) (% ) rate (% ) 3Q11 2.46 9.09 2.46 9.09 4Q11 -4.84 9.68 -1% 2.33 9.10 0.3% 3.0 8.7 -3.8% 1Q12 -7.98 10.58 -4% 1.92 9.10 0.5% 2.0 8.3 -5.7% 2Q12 -4.23 11.40 -7% 2.22 9.00 0.8% 2.5 8.2 -0.9% 3Q12 -3.51 12.16 -10% 2.43 8.90 1.0% 3.0 8.1 0.0% 4Q12 0.00 12.76 -13% 2.63 8.90 1.3% 2.0 8.1 -2.9% 1Q13 0.72 13.00 -16% 2.69 8.69 1.5% 1.5 8.1 -4.3% 2Q13 2.21 13.05 -18% 2.81 8.48 1.8% 2.2 8.0 1.1% 3Q13 2.32 12.96 -20% 2.90 8.27 2.0% 2.5 8.0 2.0% 4Q13 3.45 12.76 -21% 2.96 8.06 2.3% 3.0 7.9 -0.9% 1Q14 3.36 12.61 -21% 2.94 7.93 2.5% 2Q14 3.71 12.36 -21% 2.95 7.78 2.8% 3Q14 4.64 12.04 -20% 2.95 7.63 3.0% 4Q14 4.64 11.66 -19% 2.94 7.48 3.3%* The Federal Reserves projections for HPA are based on the CoreLogic National House Price Index (seasonally adjusted by Federal Reserve staff), while the J.P. Morgan forecast for HPA is based on the S&P/Case-Shiller home price index.

improving since the financial crisis. Since 1Q09, the banks under review have raised about $270bn of common equity capital, and as Exhibit 5 shows, the largest banks had an average Tier 1 common ratio of 10.7% as of the end of 2011, well above the 5% minimum level for the stress tests. Thus, given banks healthy levels of current capitalization, we think the stress test results will likely be a positive for risky asset

3


Exhibit 5: Given robust capitalization levels currently, we expect the results of the Feds stress tests for banks to be positiveBasel I Tier 1 common ratio, full-year 2012 dividend forecast*, full-year 2011 dividend, and projected change in dividend for the US banks in the Feds 2012 CCAR**; units as indicated Proj. Basel I 2012 chg in Tier 1 Dividend 2011 common forecast Dividend dividend ratio (% ) ($/share) ($/share) ($/share) JPMorgan Chase & Co. 10.10 1.20 1.00 0.20 Citigroup Inc. 11.80 0.20 0.03 0.17 Bank of America Corporation 9.86 0.04 0.04 0.00 Wells Fargo & Company 9.46 0.80 0.48 0.32 Goldman Sachs 12.10 1.40 1.40 0.00 Morgan Stanley 13.00 0.20 0.20 0.00 MetLife, Inc. N/A 1.05 0.74 0.31 PNC Financial Services Group, Inc. 10.30 1.80 1.15 0.65 U.S. Bancorp 8.60 0.72 0.50 0.22 Bank of New York Mellon Corporation 13.40 0.60 0.48 0.12 SunTrust Banks, Inc. 9.22 0.40 0.12 0.28 State Street Corporation 16.90 0.92 0.72 0.20 Capital One Financial Corporation 9.70 0.56 0.20 0.36 BB&T Corporation 9.70 0.80 0.65 0.15 Regions Financial Corporation 8.51 0.04 0.04 0.00 American Express 12.30 0.72 0.72 0.00 Fifth Third Bancorp 9.35 0.40 0.28 0.12 KeyCorp 11.26 0.20 0.10 0.10 GMAC LLC 8.57 N/A N/A Citizens Financial Services Inc. N/A 1.18 1.16 0.02 Comerica Inc. 10.41 0.60 0.40 0.20 Discover Financial Services N/A 0.40 0.20 0.20 Huntington Bancshares Inc. 10.00 0.16 0.10 0.06 M&T Bank Corp. N/A 2.80 2.80 0.00 Northern Trust Corp. 12.10 1.12 1.12 0.00 Zions Bancorporation N/A 0.04 0.04 0.00

Exhibit 6: Brent oil prices are back at their 2011 highs, and our Middle East crisis index remains elevated, suggesting ongoing risk of a supply shockMiddle East crisis index* versus rolling front Brent oil futures contract price; level $/bbl

10 8 6

Middle East crisis index Brent oil prices

130

120

110 4 2 0 Dec 10 100

Mar 11

Jun 11

Sep 11

Jan 12

90

* Defined as exp[0.5*(Brent oil prices 0.1075*S&P + 3.6274*JPM Trade-weighted dollar index + 0.0959*EM stocks -348.34)/4.2]. EM stocks are defined as the weighted average of Brazil, Mexico, Russia, India, China, Hong Kong, Taiwan, Korea, Singapore, Malaysia, Thailand and Indonesia benchmark stock indices in USD terms. Weights determined by market values of the indices (in USD terms) as of year-end 2011. Source: Bloomberg, J.P. Morgan

* For banks that pay quarterly dividends, full-year 2012 dividend forecast estimated as forecast for upcoming quarterly dividend*4. ** BBVA USA Bancshares, BMO Financial Corp, HSBC North America, RBC USA Holdco Corp. and UnionBanCal Corp were excluded given that their parent companies are domiciled outside of the US. Source: Bloomberg

markets. This view is somewhat supported by analyst projections: as Exhibit 5 shows, most banks that are subject to the stress test are expected to either maintain or increase dividends in 2012. Since capital plans will need to be approved by the Fed, this reflects the expectation of banks passing the stress tests and gaining approval to raise their dividends. On the US economic front, data remains supportive, although some downside risks have materialized

recently. On the positive side, labor market data continue to improve, with the four-week averages for initial jobless claims and continuing claims making fresh lows for the cycle in the latest reading. Although the ISM manufacturing survey ticked down from 54.1 to 52.4 in February, the regional activity surveys have strengthened, with the Dallas Fed, Richmond Fed and Chicago purchasing manager surveys rising more than expected. In addition, total vehicle sales rose to a 15.0mn annualized pace in February, the highest level since February 2008. On the other hand, this weeks durable goods report was disappointing, with core capital goods orders down 4.5% and core capital goods shipments down 3.1% in January. In addition, real consumption was flat in January, and past data were revised down to reflect a softer trajectory. Both of these reports add downside risk to our GDP growth forecast for 1Q12. One additional area of concern is energy prices. Brent oil futures prices have already reached their 2011 highs, and our Middle East crisis index remains elevated, suggesting ongoing risk of a supply shock (Exhibit 6). Sustained high energy prices would pose a threat to economy, though the persistent rally in equities despite the rise in energy prices suggests that investors are not yet concerned about the threat to the economy.

4


Thus, markets for now remain driven by tailwinds from lowered European tail risks; as we noted last week, we expect this impact to mostly outweigh the risk from the still-fluid situation in the Middle East, and we retain a positive stance in most markets. We remain overweight high grade, high yield and emerging markets. In CMBS, we believe the rally in new issue credit bonds will continue; given the outperformance of AAs this week, single-As are our top pick in the space (see CMBS). Similarly, in ABS, the most risky and esoteric bonds have benefited from the increase in risk appetite; we continue to see value in subordinate auto and card ABS, as well as certain off-the-run sectors such as US$ UK RMBS and subprime auto ABS (see ABS). In Treasuries, we remained biased toward modestly higher rates as we move toward mid-year. Technicals are now set to turn more negative, with Treasury supply increasing mid-month and investors likely to de-risk further in front of Fridays payrolls report. In addition, sponsorship from China has become more uncertain (see Treasuries).

5

Economic Research US Fixed Income Markets Weekly March 2, 2012 Robert MellmanAC (1-212) 834-5517 Michael Feroli (1-212) 834-5523 J.P. Morgan Chase Bank

Economics 1Q12 GDP forecast is still 2.0% but with downside risks after soft reports on consumer spending and capex Real consumer spending has stalled for three straight months, but a return to growth is expected soon Forecast looks for nonfarm payroll employment to be up 220,000 and the unemployment rate to decline to 8.2% Fiscal stimulus dollarsboth 2009 Recovery Act and other measuresare beginning to fade

Exhibit 1: Real consumer spending%ch saar over 3 months 54 3 2 1 0 -1 2010 Total ex motor vehicles and parts

Total spending

2011

2012

Exhibit 2: Nominal disposable income and wage and salary income%ch saar over 6 months, 3mma 8

The busy economic calendar this past week provided a mix of weak and strong data that is broadly consistent with the forecast for moderate growth this quarter. But the source data used to estimate real GDP have generally been weaker than expected and highlight downside risks to the forecast of 2.0% growth this quarter. In particular: Real consumer spending was unchanged in January for the third consecutive month. Available data suggest that real consumer spending this quarter will increase less than 1.0% saar, down from a forecast of 1.5% a couple of weeks ago. Core capital goods shipments plunged 3.1% samr in January, more than reversing the December surge. Even accounting for the tendency for capital goods shipments to be unusually weak in January, real business spending on equipment and software looks unlikely to reach the prior forecast of 9.0% saar growth this quarter. Private nonresidential construction declined 1.5% samr in January following gains averaging 1.2% per month through 4Q11. Preliminary figures on construction spending are often revised substantially, but for now data on business investment in structures is also tracking below the prior 10.0% saar growth forecast for this quarter.

6 4 2 0 2010

Wage and salary income

Disposable income

2011

2012

Exhibit 3: Consumer confidence and equity pricesSa 80Consumer confidence, Conference Board 70 S&P 500

Nsa, index 14001330 1260

60 1190 50 1120 1050

However, other incoming data are the strong side of expectations, including February readings on auto sales and the Conference Board measure of consumer confidence. And our favorite high-frequency indicator,

40 2010

2011

2012

6


initial jobless claims, declined again in the latest week. The trend in claims has been falling rapidly through February and the employment detail of business and consumer surveys has also generally surprised on the high side. Consequently, the forecast looks for February payroll employment growth of 220,000, up from an average 201,000 over the past three months. The upcoming calendar also includes the February ISM nonmanufacturing survey (Monday) and the January foreign trade report (Friday) that will help refine the current-quarter growth forecast.

Exhibit 4: ISM manufacturing exports orders and merchandise export volumesSa 65

%ch saar, 3m/3m

Export orders Export volumes

20 15 10

60 55

5 50 45 2010 Decline and reb ound reflect supply-chain effects ofJapan earthquake 2011 2012 0 -5

Stall in spending probably a soft patch

The latest report on January real consumer spending, with revisions, shows extreme weakness. Real consumer spending has been unchanged for the last three months, despite an increasing trend in auto sales. And the outlook for February is not much better. Unit auto sales posted another impressive gain in February, rising to 15.1mn at an annual rate from 14.1mn in January and an average 13.4mn in 4Q11. But with the CPI set to increase 0.4% to 0.5% in February, largely reflecting much higher fuel prices, real spending last month probably edged up only 0.1% samr. While actual spending numbers have been very soft lately, most incoming economic data point to a reacceleration before too long. Most important, labor market indicators suggest that growth of employment and wage and salary income is strengthening. In addition, both the Conference Board and Michigan consumer confidence surveys for February reached their highest levels in a year. Equity prices are now 12% above their 4Q11 average. And detail of this weeks minor revision to 4Q11 real GDP growth includes a fairly massive revision to the governments estimate of wage and salary income growth. Current figures based on more complete source data show that wage and salary income increased 6.6% saar in 3Q11 (was 1.5%) and 5.5% in 4Q11 (was 4.0%). But not all spending indicators are positive. Importantly, while the trend in wage and salary income has been accelerating over the past year, the trend in total disposable income has been slowing. The difference between the two series largely reflects the effects of fewer government transfers, lower interest income, and reduced farm income. The forecast for stronger real spending growth in 2Q12 also depends on the view that gasoline prices stabilize, at least in seasonally adjusted terms, before long.

Surprise: ISM mfg slips, export orders soarManufacturing data through January show a sharp split between the strong acceleration in manufacturing IP and the sharp slowdown in durable goods shipments and orders including for core capital goods. And the February ISM report was anticipated as a source of guidance on how manufacturing is holding up. Regional manufacturing surveys for February had generally posted increases, but the national ISM manufacturing survey turned down in February with declines in key activity components including new orders (from 57.6 in January to 54.9) and production (from 55.7 to 55.3). These declines are not dramatic, and the survey is consistent with continued but more moderate production growth ahead. While the decline in the ISM survey was a surprise against the backdrop of regional survey results, it is consistent with broad economic indicators including the recent slowing in real consumer spending and core capital goods orders and the expectation that the contribution to growth from inventories would be much less this quarter than in 4Q11. The one striking anomaly in the latest ISM manufacturing survey is the 4.5pt increase in new export orders to 59.5. This is the fourth consecutive increase in this series and takes export orders to its highest level since last April. This reading is certainly at odds with the view that real exports will slow this quarter in response to weaker demand conditions in most foreign markets. The next major test of this view is the January trade report (Friday). Based partly on the ISM survey, and partly on data on the aircraft, auto, and oil industries, the forecast looks for a 1.4% samr increase in export

7


volumes and a narrowing in both the nominal and real trade deficits.

Exhibit 5: Fiscal stimulusFY, $ bn ARRA non-ARRA total stimulus funds Change in stimulus funds as % of GDP

2009 177 0 177 177 1.3

2010 323 92 415 238 1.7

2011 171 266 437 22 0.1

2012 39 247 286 -151 -1.0

2013 44 35 79 -207 -1.3

Fiscal stimulus just beginning to waneFederal fiscal policy has been quite accommodative throughout the expansion. Some well-advertised policy choices are arriving at year-end that could change all that, most notably the expiration of the Bush tax cuts. Regardless of those year-end decisions, however, there are two other factors that will cause fiscal policy to become more restrictive in the meantime. First, stimulus dollarsnot only from the original February 2009 Recovery Act, or ARRA, but also from the many subsequent smaller stimulus packagesare beginning to fade this year. Second, though smaller in magnitude, defense spending is also set to experience a meaningful contraction. Together, these two factors will present a headwind to growth even before we arrive at the larger year-end decisions.

ARRA: American Recovery and Reinvestment Act of 2009

Exhibit 6: Fiscal thrust/drag from stimulus policiesChange in deficit impact due to fiscal stimulus, % of fiscal year GDP 2.0

1.5 1.0 0.5 0.0 -0.5 -1.0 -1.5 2009 2010 2011 2012 2013

What is stimulus?Fiscal support for an ailing economy can come in one of two forms: automatic stabilizers and discretionary fiscal actions to support the economyor what is commonly called stimulus. Automatic stabilizers include, for example, the automatic reduction in average income tax rates that takes place as incomes and effective marginal tax rates move lower. This category also includes automatic increases in outlays for income support programs such as regular unemployment benefits. The magnitude of automatic stabilizers varies from countryto-country based on institutional arrangements, but in the US, for every percentage point that the economy is operating below its capacity, the deficit will automatically increase by about 0.35%-point. By this rule of thumb, about 2.0%-points of the current deficitto-GDP ratio is due to automatic stabilizers. Stimulus occurs when policymakers actively make choices to change policy in order to provide more fiscal support to the economy. The most notable example in the current cycle is the 2009 Recovery Act, or ARRA. Not all of ARRA should be considered stimulus, however, as some componentssuch as patching the Alternative Minimum Tax to prevent a large middle-class tax hike occur every year. The only reason the AMT is not permanently extended is so that Congress can preserve an accounting mirage. For this reason, we exclude these extenders in the table on this page. More important for understanding the outlook is to remember that not all

Exhibit 7: Deficit decompositionFY, % of GDP Deficit Automatic stabilizers Discretionary fiscal policy TARP Structural

2009 -2.1 -1.3 -1.1 -5.7

2010 -8.9 -2.3 -2.9 0.8 -4.5

2011 -8.6 -1.9 -2.9 0.2 -4.1

2012f -7.7 -1.7 -1.8 -0.1 -4.0

2013f -5.6 -1.7 -0.5 0.0 -3.4

-10.2

Note: structural includes changes in defense outlays

stimulus measures undertaken in the current cycle were contained in ARRA. Perhaps the most prominent such measure is the payroll tax holiday, which was initially conceived to offset the tax hike that took place when the Making Work Pay tax creditwhich was part of ARRAexpired at the end of 2010. However, in the years since the recession there have been numerous other smaller legislated changes to tax and spending policy to support the economy, including extending unemployment benefits, COBRA health coverage benefits, state fiscal relief, etc. A recent tabulation by the Center for Budget and Policy Priorities places the bill for these non-ARRA stimulus measure over the 2009-13 period at $640bn, almost as much as the ARRA stimulus measures impact on deficits over the same period, which is about $754bn.

8


Even with the recent extension of the payroll tax holiday, the amount of support provided to the economy through discretionary fiscal actions is set to contract this year, which will be a headwind to growth. The total amount of stimulus-related outlays and tax expenditures will contract by about 1% in fiscal year 2012. The exact multiplier effect of this financial measure on GDP is uncertain, but using a fairly standard multiplier of 1.0 would imply that the waning of stimulus could subtract about 1%-point from growth this year. Changes in defense outlays have been dominated by noneconomic considerations, and so they are not included in the above calculations which measures the fiscal policy response the Great Recession.

Exhibit 8: Defense outlays% of GDP5.0 4.5 4.0 3.5 3.0 2.5 CBO, assuming troops in Mideast Forecast OMB

00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15

A farewell to armsMost discussions of how fiscal policy relates to the outlook rightly focus on the discretionary stimulus measures. However, there are other moving parts when it comes to fiscal policy. One that will become increasingly important for the US economy is defense spending. The winding down of military operations in Iraq and Afghanistan could contribute to a decline in defense spending of about 1/4%-point of GDP this year, and then contracting about 1/2%-point in both 2013 and 2014. The regional recessions in California and New England in the early 1990s serve as a reminder that welcome foreign policy developments can sometimes be accompanied by painful economic transitions. Estimating the impact of these developments is made a little more difficult by the fact that the CBOs baseline estimate for these Overseas Contingency Operations (OCOs) must assume a continuation of the troop levels in the prior fiscal year, even though most observers expect the ongoing drawdown of troops in the Mideast will continue. To allow for this possibility, CBO also constructs an alternative scenario whereby troops in that region are gradually decreased, and in that scenario defense spending declines in a similar manner to that in the OMB forecastwhich is allowed to make more realistic assumptions concerning defense policy. One potential offset is that some of these expenditures go to foreign nationals for services provided in the theater of operations. In the economic accounts, these purchases are considered imports. Such imported defense expenditures are less than $30bn, whereas the decline in defense spending over the next few years will likely amount to over $100bn. So even if defense imports fell to

Exhibit 9: Change in defense outlays% of GDP

1.0 0.5 0.0 -0.5 -1.0

Forecast

OMB CBO, assuming troops in Mideast operations down to 45,000 by 2015 80 85 90 95 00 05 10 15

zero, this offers only a partial offset to the domestic drag that is set to occur from reductions in defense outlays.

And now, the elephant in the roomSo far, this note hasnt discussed the fiscal issues that loom at the end of the year: most important of which is the expiration of the Bush tax cuts, whichif realized would increase the tax burden of the household sector by about $250bn, or about 1.5% of GDP. The J.P. Morgan forecast does not assume that this tax increase will be implemented in its entirety. It is difficult to say what will replace current law, in part because that will depend on how the November elections play out. Our uncertainty about the future tax regime is likely shared by many households and businesses, and is why we believe activity could slow around year-end as decisions are deferred until there is more clarity on the tax outlook. Even absent these year-end considerations, however, federal fiscal policyas measured by changes in discretionary fiscal actions and changes in the structural

9


federal deficittightened by about 0.4%-point of GDP in 2011, and looks set to tighten by 1.2%-points in 2012. The 2013 outlook is particularly hazy but our best guess is that fiscal policy will tighten by another 1.9%-points.

10

US Fixed Income Strategy US Fixed Income Markets Weekly March 2, 2012 Terry BeltonAC (1-212) 834-4650 Meera Chandan (1-212) 834-4924 Kimberly L. Harano (1-212) 834-4956 J.P. Morgan Securities LLC

Treasuries Technicals are now set to turn more negative with Treasury supply increasing mid month and investors likely to de-risk further in front of Fridays payrolls report and sponsorship from China has become more uncertain; we maintain our bias to higher rates expecting 10-year rates to move modestly higher into mid year Despite the upcoming supply, we advise against initiating long-end steepeners and prefer to take advantage of any supply-related cheapening to initiate long-end flatteners Bill net issuance will drop but stay positive in the coming weeks, keeping front-end Treasuries cheap versus OIS Switch out of the 6-year sector into a combination of 4s and 8s Stay bullish on TIPS breakevens

Exhibit 1: Ten-year yields have held a narrow trading range since November10-year Treasury yields*; %

2.15 2.10 2.05 2.00 1.95 1.90 1.85 1.80 1.75 Dec 11 Jan 12 Feb 12 Mar 12*Horizontal lines indicate 1 standard deviation move around 4-month average

Exhibit 2: Supply matters: Treasuries continue to sell off into supply and rally out of supply10-year par Treasury yields averaged in the business days around 5 and 10year auctions; 9/11-2/12; %

2.09 2.08 2.07 2.06 2.05 2.04 2.03 2.02 2.01 2.00 -8 -6 -4 -2 0 2 4 6 Business days around 5- and 10-year auctions 8

Market viewsTreasury yields were mixed this week with 2- and 5-year Treasury yields falling 4 bp and 5 bp, and 10- and 30-year yields rising 1 bp and 2 bp, respectively. The directionless market reflects the competing forces of strong performance of risky assets, and strong technicals in the Treasury market driven by month-end buying, Fed purchases at the long end, and a longer than normal period (two weeks) without coupon supply in Treasuries. With this weeks modest sell-off, 10-year yields are at 1.98%, still close to the middle of the narrow trading range they have held since November (Exhibit 1). Economic data were mixed this week with labor markets continuing to show strength but activity data disappointing. January durable goods orders declined a larger than expected 4% with core capital goods and shipments now falling in three of the past four months. The ISM manufacturing survey was also weaker than expected (falling 1.7 points to a still-solid 52.4) while consumer spending data disappointed with real consumption unchanged in January following flat readings in November and December. The soft trajectory suggests

1Q12 real consumer spending is tracking below 1% this quarter. On the positive side, initial jobless claims fell again with the 4-week average reaching 354,000, a new low for the cycle. Consumer confidence was also strong with the February Conference Board index reaching its highest level in a year. Finally, motor vehicle sales were stronger than expected (15.03mm) and pending home sales data was solid rising 2% in January and 15% saar in the last six months. On balance however, the spending data are tracking weaker than expected and provide downside risk to our 1Q12 GDP forecast of 2%. With month-end behind us, technicals are now set to turn more negative with Treasury supply increasing mid month

11


Net non-commercial longs in FV, ED and TU futures contracts averaged around the number of business days around payrolls release day over the past two years; $bn

Exhibit 3: We look for investors to unwind duration longs in the front end ahead of payrolls

Exhibit 4: Commentary from the Fed has been less dovish since JanuaryFed sentiment index*; %

-10 -15 -20 -25 -30 -35 Nov 11 Jan 12 Mar 12* The Fed sentiment index is computed as the cumulative sum of yield changes in 5Y Treasury futures in the 15-minute period following the first Fedspeak headline on Bloomberg, since 7/3/06. Fedspeak is defined as any speech, FOMC statement, or FOMC minutes.

255 250 245 240 235 230 225 220 215 210 -20Source: CFTC

TU, ED and FV TU and ED

155 150 145 140 135 130 125 120 115 110 105 100 20

-15

-10 -5 0 5 10 Business days around payrolls release

15

and investors likely to de-risk further in front of Fridays nonfarm payrolls report. Intermediate Treasury yields continue to exhibit reasonably strong cyclical patterns around coupon supply with yields generally rising going into auction weeks and falling coming out of auctions. Indeed, 10-year Treasuries have rallied in 10 of the last 12 weeks out of supply and sold off in 7 of the last 12 weeks leading into supply (Exhibit 2). These supply technicals, combined with increased sensitivity to the employment report, bias 10-year rates to the upper end of their trading range next week. As discussed last week, the sensitivity of forward OIS rates to downside surprises in the unemployment rate has increased sharply in the last few months (see US Fixed Income Markets Weekly, 2/24/12) . This heightened sensitivity should lead to further de-risking next week and cause investors to reduce yield curve carry trades and unwind duration longs in the front end of the yield curve (Exhibit 3). A very subtle change in tone from the FOMC is also supportive of modestly higher rates as recent Fed speeches have suggested reduced prospects for QE3. Our Fed sentiment index, which tracks changes in 5-year Treasury yields in response to speeches by FOMC members, has increased 10 bp in the last few weeks suggesting commentary from the Fed has been slightly less dovish than during January (Exhibit 4). These speeches include testimony this week from Chairman Bernanke indicating

Exhibit 5: Recent comments by FOMC membersSpeaker Comments Looking ahead, we may need to do more if the recovery falters or if inflation stays well below 2 percent. In answer to a question, Williams said another round of monetary stimulus will depend on how the economy performs but is Williams definitely not off the table. Because interest-rate changes take time to have their desired effect, Williams said he wouldnt wait until the economy reaches full employment to raise borrowing costs. "Our current expectation is the short term rate will stay low for a good bit more time. Eventually the economy will strengthen, inflation will begin to rise and the Fed will have to raise short-term interest rates," Mr. Bernanke said, cautioning that the Fed wants "to make sure banks understand their risks and they are well protected and hedged against whatever course interest rates might take." More bond purchases wouldnt be appropriate now, Lockhart told reporters after his speech. The onset of recessionary conditions and movement in the direction of deflation would be conditions that could lead to more easing, Lockhart said. The current policy is already quite accommodative both in terms of interest rates and the size of the central banks balance sheet, she said. Additional monetary stimulus could spur inflation, while less action by policy makers could risk weakening an already slow expansion and cause disinflation, she said. St. Louis Federal Reserve Bank President James Bullard expressed optimism over the economic recovery, saying he saw a real gain in the recent decline in the unemployment rate. The Fed official also distanced himself from the specific language in the central banks latest policy statement saying that he would have preferred not to include a calendar date in the statement. Way out at the end of 2014, we dont know what the economys going to look like and it over-emphasizes our forecasting ability, which isnt that good.

Date

3/2

Bernanke

3/1

Lockhart

3/1

Pianalto

3/1

Bullard

2/24

St. Louis President James Bullard said he doesnt favor additional asset purchases since inflation is above the central banks target. Inflation numbers would have to indicate further disinflation or the threat of deflation, Bullard Bullard said at the U.S. Monetary Policy Forum hosted by the University of Chicago Booth School of Business. Inflation risks are to the upside currently. Source: Bloomberg, Wall Street Journal

2/24

12


the need for banks to protect against higher rates, comments by Lockhart and Bullard indicating further asset purchases are not appropriate right now, and comments from SF Fed President Williams noting that further stimulus is data-dependent and that the Fed will begin raising policy rates before the economy reaches full employment (Exhibit 5). While the main thrust of communications from the FOMC remains that they expect policy rates to remain low until 2015, the subtle change in tone highlights the importance of upcoming data in maintaining that view. Finally, we note that this weeks newly released data on Treasury holdings of large foreign investors highlights a significant longer-term risk to the Treasury market. As discussed in more detail below, these data show China sold $155 bn of Treasuries during 2H11 with sales accelerating in the fourth quarter. To the extent this represents a more permanent desire on the part of the single largest holder of US Treasuries to diversify into other assets, the longer-term risks to the Treasury market are potentially quite large. In sum, with technicals now more supportive, risks asymmetric around payrolls, Treasury rates rich to fair value, and less-certain sponsorship from China, we maintain our bias to higher rates expecting 10-year rates to move modestly higher into mid year (Exhibit 6). To be sure, next Fridays payroll report will be key in determining the near-term direction of the Treasury market. A weaker than expected report, especially one that moves the unemployment rate higher, will result in investors reentering carry trades with rates likely to move back below the middle of the trading range they have held in the last few months. At the long end of the curve, we expect the upcoming mid-month auctions to generate increased selling pressure with many investors initiating back-end steepeners into supply. Despite the upcoming supply, we advise against initiating long-end steepeners and prefer to take advantage of any supply-related cheapening to initiate long-end flatteners. Auction cyclicals leading into 30-year auctions have become much more muted recently as Operation Twist has increasingly reduced the available float at the long end of the curve (Exhibit 7). This has meant that pre-auction steepeners have performed much more poorly while amplifying postauction flattening. As a result, we advise patience in positioning around long-end supply and are biased to

Exhibit 6: J.P. Morgan interest rate forecast%Actual Rates Effective funds rate 3-mo LIBOR 3-month T-bill (bey) 2-yr Treasury 5-yr Treasury 10-yr Treasury 30-yr Treasury 1Q12 2Q12 3Q12 4Q12

2 Mar 12 31 Mar 12 30 Jun 12 30 Sep 12 31 Dec 12 0.10 0.48 0.07 0.28 0.85 1.98 3.11 0.10 0.40 0.04 0.27 1.10 2.25 3.30 0.10 0.45 0.02 0.30 1.25 2.50 3.60 0.10 0.50 0.02 0.30 1.25 2.50 3.60 0.10 0.50 0.02 0.30 1.25 2.50 3.60

0.46 * 30-year Treasury yields minus 5-year Treasury yields averaged in the business days around the 30-year auction before and after Operation Twist was announced*; % 0.40 0.53 Pre Post Operation Operation Twist 0.52 Twist

Exhibit 7: Long-end supply concessions have fallen after Operation Twist

0.51 0.50 0.49 0.48 0.47 -10 -5 0 5 10

0.35

0.30

0.25

Business days around Bond auction* Pre Operation Twist includes Bond auctions between April 2011 and September 2011. Post Operation Twist includes Bond auctions held starting October 2011

initiating long-end flatteners as we approach the bond auction. This month in particular is more supportive than usual of post-supply long-end flattening. Four out of the seven purchase operations in the 25- to 30-year sector are scheduled to occur after March 20. Moreover, purchases this month are likely to be of longer duration than typical given the Fed schedule released this week. Even if the Fed purchases $4.25bn per operation in the 6- to 8-year sector, which is the upper end of the announced range (of $3.5 to $4.25bn) for this part of the curve, purchases in the sector will comprise only 29% of total purchases in the month, lower than the target of 32% and the prior 3-month average of 33%.

13


Exhibit 8: 4s and 8s look cheap versus 6s0.06 0.05 0.04 0.03 0.02 0.01 -0.00 -0.01 -0.02 -0.03 -0.04 -0.05 Sep 11 Dec 11

T 2.25% Nov-17 yield 0.69 * 3.25% May-16 0.4 * 3.375% Nov-19; %

Exhibit 9: Even though the pace will fall, net issuance of bills will stay positive and average $11bn/ week for the next six weeksProjected net issuance of bills and Treasurys operating cash balance; $bn 40 38 35 Projected bill $bn 140

20 0 -20 -40

10

19

18

14

20 8 8 8

net issuance

8

120 100

-10

-5

80 60

Mar 1202 Feb-60

Proj. cash balance

-36

40 20

01 Mar

08 Mar

15 Mar

22 Mar

09 Feb

16 Feb

23 Feb

29 Mar

On the curve, we recommend switching out of the 6year sector into a combination of 4s and 8s. The 4-year sector has cheapened significantly on the curve over the past week, and the 3.25% May-16s in particular is now one of the cheapest issues on the curve (using the yield error relative to the fitted curve as a metric). In addition, the yield error of 2.25% Nov-17s and 3.375% Nov-19s are near the lows and highs the past year, respectively. Thus, we recommend switching out from the Nov-17s into a combination of May-16s and Nov-19s on a curve and level neutral basis (see Trade Recommendations). As shown in Exhibit 8, this weighted trade currently appears around 5bp mispriced. Treasury/OIS spreads are currently close to the wider end of their 1-year range. Despite attractive valuations, we are cautious on positioning for a richening of front-end Treasuries versus OIS rates, waiting instead for better entry levels in the coming weeks. Two reasons motivate this view. First, we expect net issuance of bills to stay positive in the coming weeks, with outstanding bills expected to increase another $60 bn over the next five weeks. As shown in Exhibit 9, although the pace of net issuance will slow from the levels in February, we expect it to stay positive for the next six weeks. After mid-April, however, bill net issuance should turn negative, which will then be supportive of tighter spreads between front-end Treasuries and OIS rates. Second, dealer positions in Treasuries have stayed elevated, which will likely keep GC repo and thus Treasury/OIS spreads wide. As discussed earlier this year, (see US Fixed Income Markets Weekly, 1/20/12), Treasury/OIS spreads have been highly correlated with dealer Treasury holdings, with spreads

Exhibit 10: Foreign buying of Treasuries was strong thru 3Q11 before declining in 4Q1180 3Q11

Average monthly Treasuries (coupons and bills) purchased by foreigners; $bn

60 Year-ended June 2011 4Q11

40

20

0 Jun 06

Jun 07

Jun 08

Jun 09

Jun 10

Jun 11

All data is for the year ending June 30 except 2H11, which is quarterly data. Source: Treasury TIC

narrowing as dealers grow their short base in Treasuries. Currently, with mortgage and corporate dealer inventories extremely light, dealers are net long Treasuries, with their net position close to the longest of the last five years. The lack of a dealer short base in Treasuries combined with positive bill supply should keep GC repo rates and Treasury OIS spreads elevated in the near term.

Foreign buying of TreasuriesTreasury released details of its annual survey on foreign holdings of US securities this past week. This data is released with a significant lag, and the latest release

14

03 May

05 Apr

12 Apr

19 Apr

26 Apr


relates to the holdings at the end of June 2011. In addition, Treasury also introduced new holdings data which will be released in a more timely manner going forward, and is expected to provide more accurate country-level data. Currently this is available for Treasuries for select major holders as of September and December 2011, but will be made available on a monthly basis going forward. These two releases are important since they provide a countrylevel breakdown of holders of Treasuries, which is not otherwise available. There are three key takeaways from the release. First, the data shows that the overall pace of foreign buying was solid in the year ending June 2011, averaging $52bn/month in line with the previous two years. Foreign buying increased to $72bn/month in 3Q11, before weakening in 4Q11 (Exhibit 10). As discussed previously (see US Fixed Income Markets Weekly, 11/18/11), the pickup in 3Q11 was largely driven by a flight-to-quality bid for Treasuries from Europe as the sovereign debt crisis in Europe escalated. Second, China became a large net seller of Treasuries in 2H11. Exhibit 11 shows that China went from buying $16bn Treasuries per month to selling $26bn per month in 2H11, selling a total of $155bn Treasuries in 2H11. This is notable not only because China is the largest holder of Treasuries, but also because Chinas foreign exchange reserves were little changed over the same period (only $16bn lower), suggesting a diversification away from Treasuries into other asset classes. Finally, the data highlight that weak demand from China was more than offset by the heavy currency intervention from Japan, which more than quadrupled its pace of Treasury buying during 2H11. This buying, however, is now likely to fade with the BoJ growing its balance sheet and the Yen weakening this year. Further lackluster demand from China is therefore likely to be a significant negative for the Treasury market, potentially requiring significantly higher yields for the Treasury market to clear.

Exhibit 11: The US Treasury lost its largest client in 2H11Country China, Mainland Japan Brazil Taiwan Luxembourg Russia Switzerland Belgium Hong Kong United Kingdom Jun'09-Jun'10 Jun'10-Jun'11 16 16 8 7 1 4 3 0 0 2 2 -1 2 1 1 4 3 -2 1 4 3Q11 -12 34 3 7 2 -1 13 14 0 -6

Average monthly net purchases of Treasuries (bills and coupons) for the top 10 Treasury holders at the end of 2011; $bn

4Q11 -39 25 0 4 8 0 -5 1 3 -2

Source: Treasury TIC

Exhibit 12: Continued declines in tail risk in Europe should support wider TIPS breakevens10-year TIPS breakevens versus Euro Stoxx 50 index; bp level

280 260 240 220 200 180 10-year TIPS breakeven Euro Stoxx 50

3200 3000 2800 2600 2400 2200 2000

160 1800 Mar 11 Apr 11 Jun 11 Jul 11 Sep 11 Nov 11 Dec 11 Feb 12

TIPSThis week the rally in breakevens hit a speed bump as Brent oil prices declined 1.5% and shorter-maturity nominal yields fell. All in all, 5-, 10-, 20-, and 30-year breakevens narrowed 6bp, 4bp, 4bp, and 5bp, respectively.

Our views on breakevens are unchanged; we continue to look for breakevens to widen. As we highlight in the Cross Sector Overview, declining tail risks in Europe are supportive of the rally in risky assets, and we expect TIPS breakevens to widen in sympathy. As Exhibit 12 shows, TIPS breakevens have tracked European equities very closely over the past year and are likely to widen as the European situation improves. In addition, escalating tensions in the Middle East could also send oil prices higher and push breakevens wider. Over the next week, the market is likely to focus on the Fed sales operation on Monday and the employment report on Friday. First, we expect the Fed to sell around $1.34bn TIPS on Monday, which is the average sale size

15


over the past five operations. Given that the Fed owns only five issues in this sector of the curve (with holdings totaling only $5.2bn), we continue to expect the Fed to attempt to sell issues that it holds more of. Exhibit 13 shows the TIPS in the sales bucket ranked by Fed holdings. Although the employment report has less direct relevance for TIPS than, say, the CPI report, in recent months, breakevens have become more sensitive to surprises in payrolls, with 10-year and 30-year breakevens now more sensitive than front-end breakevens (Exhibit 14). Exhibit 15 shows the projected change in breakevens for various surprises in payrolls based on a 1-year regression. The analysis suggests that if payrolls surprise sharply to upside, the breakeven curve is likely to steepen, and 10year breakevens are likely to outperform if the surprise is large enough.

Exhibit 13: We expect the Fed to sell issues that it holds more ofFed holdings ($bn Cpn 3.000 2.000 1.625 1.250 0.625 Maturity 07/15/12 01/15/14 01/15/15 04/15/14 04/15/13 of real not'l) 1.7 1.6 1.5 0.2 0.1 Fed holdings (% ) 7% 8% 8% 2% 1%

Statistics for TIPS in the sales bucket for the upcoming Fed operation; $bn

Exhibit 14: Longer-maturity breakevens have become more sensitive to payroll surprises in recent monthsRolling 12-month beta of change in breakevens (bp) on employment report release dates regressed against surprise in payrolls* (000s)

0.04 0.03 0.02 0.01 0.00 -0.01 Feb 11

5Y

10Y

30Y

Trade recommendations Buy 4s and 8s versus 6s May-16s have cheapened dramatically over the past week, and 6s now appear very rich versus 4s and 8s on a curve and level neutral basis. We recommend switching out of 6s into a combination of 4s and 8s. Buy 69% risk, or $151.7mn notional of T 3.25% May16s (yield: 0.682%; bpv: $442/mn). Sell 100% risk, or $168.7mn notional of T 2.25% Nov17s (yield: 1.041%; bpv: $576/mn). Buy 40% risk, or $50mn notional of T 3.375% Nov-19s (yield: 1.511%; bpv: $778/mn). Weighted spread is 3.4bp. One-month weighted carry is 0bp and roll is 0.2bp. Stay in 2s/7s steepeners hedged with reds/ greens flatteners Stay overweight 100% risk, or $170.1mn notional of T 0.25% Feb-14s (yield: 0.276%). Stay underweight 98% risk, or $50mn notional of T 1.375% Feb-19s (yield: 1.381%). Stay underweight 98% risk, or 1,327 EDM3 contracts (price: 99.43). Stay overweight 100% risk, or 1,354 EDZ4 contracts (price: 98.85). (US Fixed Income Markets Weekly, 2/24/12). P/L since inception: -2.1bp of yield.

May 11

Aug 11

Dec 11

* Defined as actual release minus median of Bloomberg forecast survey for change in total nonfarm payrolls. Source: Bloomberg, J.P. Morgan

Exhibit 15: with 10-year breakevens likely to widen 3bp if payrolls surprise by +100KProjected change in breakevens on payrolls release date based on 1-year regression of change in breakevens on employment report release dates regressed against surprise in payrolls*; bp

5Y 10Y 30Y

-200 -3.8 -7.7 -5.2

-100 -2.3 -4.1 -2.5

Payrolls surprise (000s) -50 0 50 -1.5 -0.7 0.0 -2.3 -0.5 1.3 -1.2 0.2 1.6

100 0.8 3.1 2.9

200 2.3 6.7 5.6

* Defined as actual release minus median of Bloomberg forecast survey for change in total nonfarm payrolls. Source: Bloomberg, J.P. Morgan

Maintain exposure to belly cheapening trades Stay overweight 94% risk, or $307.1mn notional of T 0.25% Feb-15s (yield: 0.393%). Stay underweight 100% risk, or $157.4mn notional of T 2.75% Feb-18s (yield: 1.092%).

16


Stay overweight 42% risk, or $25mn notional of T 6.625% Feb-27s (yield: 2.522%). (US Fixed Income Markets Weekly, 2/24/12). P/L since inception: -0.2bp of yield. Stay overweight 75% risk, or $191.7mn notional of T 0.25% Jan-15s (yield: 0.381%). Stay underweight 100% risk, or $112.8mn notional of T 1.25% Jan-19s (yield: 1.362%). Stay overweight 54% risk, or $25mn notional of T 6.625% Feb-27s (yield: 2.522%). (US Fixed Income Markets Weekly, 2/3/12). P/L since inception: -4.2bp of yield. Stay underweight May-38s versus May-37s Stay overweight 100% risk, or $48.9mn notional of T 5% May-37s (yield: 2.961%). Stay underweight 100% risk, or $50mn notional of T 4.5% May-38s (yield: 2.992%). (US Fixed Income Markets Weekly, 2/10/12). P/L since inception: +0.2bp of yield. Stay underweight on-the-run 10s versus triple olds Stay underweight 100% risk, or $50.6mn notional of 2% Feb-22s (yield: 1.983%). Stay overweight 100% risk, or $50mn notional of T 3.125% May-21s (yield: 1.824%). (US Fixed Income Markets Weekly, 2/3/12). P/L since inception: -0.9bp of yield. Stay in weighted 2s/3s steepeners Stay overweight 100% risk, or $219.8mn notional of T 0.125% Dec-13s (yield: 0.268%). Stay underweight 70% risk, or $100mn notional of T 0.25% Jan-15s (yield: 0.381%). (US Fixed Income Markets Weekly, 1/27/12). P/L since inception: -0.8bp of yield. Stay in weighted 4s/25s steepeners hedged with reds/10s flatteners Stay overweight 96% risk, or $156.8mn notional of T 2.125% Feb-16s (yield: 0.614%). Stay underweight 78% risk, or $25mn notional of T 4.75% Feb-37s (yield: 2.963%). Stay overweight 100% risk, or $76.4mn notional of T 2% Nov-21s (yield: 1.947%). Stay underweight 100% risk, or 2,732 EDM3 contracts (price: 99.435). (US Fixed Income Markets Weekly, 1/27/12). P/L since inception: -6.6bp of yield.

Closed trades in 2012TRADE Nominal Treasuries 10-year duration shorts

P/L reported in bp of yield unless otherwise indicated

ENTRY

EXIT

P/L -11.1

01/20/12 02/10/12

Curve Underweight 4s versus 3s and 7s Buy 2.375% Jun-18s versus 1.875% Oct-17s 7s/30s flatteners Buy Nov-21 Cs versus Ps 4s/6s steepeners vs. reds/greens flatteners 7s/30s flatteners 3s/7s flatteners

12/09/11 10/14/11 01/06/12 09/23/11

01/06/12 01/06/12 01/27/12 01/06/12

0.9 -2.8 -14.2 -0.2 3.0 3.5 8.5

01/06/12 01/27/12 02/10/12 02/24/12 02/12/12 02/24/12

Misc. Sell 2s versus OIS 12/09/11 01/06/12 Buy 5s versus OIS 01/06/12 02/04/12 Sell 2-year Treasuries vs. JGBs swapped into USD 08/12/11 02/10/12 Buy the March CTD Bond basis 12/09/11 02/24/12 TIPS Sell 5-year TIPS on a breakeven basis

0.8 5.0 44.0 1/32nds

12/16/11 01/20/12

-12.3

Stay overweight 6.5% Nov-26s versus a weighted barbell of Aug-23s/Feb-31s Stay underweight 62% risk, or $39mn notional of T 6.25% Aug-23s (yield: 2.109%). Stay overweight 100% risk, or $50mn notional of T 6.5% Nov-26s (yield: 2.505%). Stay underweight 38% risk, or $16.5mn notional of T 5.375% Feb-31s (yield: 2.741%). (US Fixed Income Markets Weekly, 1/20/12). P/L since inception: -3bp of yield. Stay underweight Apr-14 TIPS versus Jan-14 TIPS Stay long 100% risk, or $25mn notional of TII 2% Jan14s (yield: -1.966%). Stay short 100% risk, or $25.4mn notional of TII 1.25% Apr-14s (yield: -1.878%). (US Fixed Income Markets Weekly, 2/24/12). P/L since inception: +3.4bp of yield.

17

US Fixed Income Strategy US Fixed Income Markets Weekly March 2, 2012 Srini RamaswamyAC (1-212) 834-4573 Praveen Korapaty (1-212) 834-3092 Alberto Iglesias (1-212) 834-5116 J.P. Morgan Securities LLC

Interest Rate Derivatives Our views on swap spreads across the curve remain unchangedstay biased towards wider spreads at the long end of the curve, but neutral on front-end spreads and FRA-OIS The term structure of old 7-year note asset swap spreads is dislocated in April June 2016 maturitiespay in 3.25% June 2016 maturity matched swap spreads versus receiving in 3.25% July 2016 spreads Given our medium-term bias towards higher rates, we continue to favor bearish trades with flat holding costs and/or trades that offer asymmetric exposure to higher rates; position for a conditional cheapening of the belly of a ED6/7Y/25Y weighted butterfly in a sell-off via Eurodollar short-Sep midcurve puts and expiry matched payer swaptions Stay neutral on rate volatility, but sell curve vol

Exhibit 1: In spite of the recent narrowing, there is still room for more retracement in spreads driven by the compressing influence of central bank liquidity

Current value, change over the past week, best and worst levels in 1H11 and since 2H11 in the 2Y, 5Y and 10Y maturity matched swap spreads, 10Y swap yields, 3- and 6-month forward constant maturity FRA-OIS differential, the 3M EUR/USD FX OIS basis and semi-peripheral sovereign CDS spreads*

1-week Worst level Best level Percent Current change since 7/1/2011 in 1H11 retracement 10Y swap spreads (bp) 7.7 -1.5 22.7 3.9 80% 6M FRA-OIS (bp) 31.8 -7.8 72.4 16.7 73% 5Y swap spreads (bp) 25.0 -2.7 44.9 15.8 68% 3M FRA-OIS (bp) 31.4 -6.6 66.6 15.2 68% 2Y swap spreads (bp) 26.0 -4.2 54.8 12.6 68% 3M EUR/USD xccy OIS basis (bp) -44.4 -1.9 -105.7 -4.4 61% Semiperipheral spreads (bp) 290 -30 433 126.6 47% 10Y swap yields (%) 2.12 0.04 1.90 3.83 11% * Average of Italy, France and Spain 5Y CDS spreads.

Exhibit 2: Maturity matched swap spreads in the long end of the curve have lagged the rise in fair valueMaturity matched 30-year swap spread versus fair value*; bp

-24 -26 -28 -30 -32 -34 -36 Nov 11 Fair value Actual 30Y Sswap spread Dec 11 Dec 11 Jan 12 Feb 12 Feb 12

SwapsSwap spreads are sharply narrower over the past week, led by the front end of the curvematurity matched swap spreads are now narrower by 5-6bp in the 2- to 3year sector, by 2-3bp in the 5- to 10-year sector, and flat at the long end of the curve. The key driver has of course been the considerable further easing in funding pressures brought about by this weeks 3-year LTRO conducted by the ECB. European banks took in 529bn of 3-year term funding, with about 59% of that amount representing new funding. Given this sizeable liquidity injection, funding pressures continued to ease, measured by virtually any indicator; FRA-OIS spreads in USD as well as EUR are sharply narrower, the 3-month EUR/USD OIS currency basis is sharply wider (i.e., less negative), and 3-month USD Libor is lower by 1.5bp on the week. Looking ahead, we remain neutral on front-end spreads and FRA-OIS spreads. On the one hand, the powerful compressive influence of central bank liquidity could continue to pressure spreads narrower; despite the considerable narrowing that has already occurred, FRAOIS spreads still have room to retrace further to early 2011 levels (Exhibit 1). In addition, the upcoming

* Fair value based on a regression model of maturity matched 30-year spreads versus the 2s/30s Treasury curve, estimated duration of the aggregate VA universe, and cumulative Fed gross purchases of Treasuries. Regression since August 2010.

release of the bank stress test results in the US by the Fed could also prove to be a positive factor, as was the case in 2009 (when US banks were much less well capitalized than is currently the case). On the other hand, this is balanced by two factorsthe risk of bank ratings downgrades and the fact that FRA-OIS spreads now appear narrow to fair value, both of which we noted last week (see Interest Rate Derivatives, US Fixed Income Markets Weekly, 2/24/2012). Thus, given balanced risks, we stay neutral on FRA-OIS and front-end spreads. We remain biased towards wider swap spreads at the longer end of the curve. As equities have risen off of their late-November lows and the curve has steepened, VA duration has fallen, pushing the fair value of 30-year

18


swap spreads steadily wider. In contrast, actual maturity matched swap spreads in that sector have trended mildly lower over this period (Exhibit 2), but the mispricing relative to fair value has begun to correct in recent weeks. Consequently, we remain biased towards wider swap spreads in the 30-year sector. On a relative value basis, we recommend buying old 7s in the Apr 2016 June 2016 sector versus longer maturities on a swap spread switch basis. As seen in Exhibit 3, the term structure of the old-7s swap spread curve is dislocated in this sector; the spread differential between the 3.25% June 16s and 3.25% July 16s (separated by a mere one month in maturity), for instance, is at an extreme level as seen in Exhibit 4. How much steeper can this spread curve go? This steepening is likely the result of technicals, and in principle there is of course no limit to the mispricing between these two bonds. However, it is worth noting that if the June 16s were to cheapen by a further 1.4bp, the dirty prices of both bonds (the 3.25% June 16s and the 3.25% July 16s) would be the same. In that eventuality, an investor could conduct a cash-neutral switch out of the 3.25% July 16s into the 3.25% June 16s and receive virtually identical coupon and principal cash flows, but each cash flow would occur one month sooner. Thus, we would view this as a soft bound; given over 2.5bp of upside potential in the spread switch, versus a potential downside of 1.4bp, we view the risk reward as favorable and recommend buying the 3.25% June 16s versus the 3.25% July 16s on a swap spread switch basis (see Trade recommendations).

Exhibit 3: Technical flows appear to have cheapened old 7s in the April-to-June 2016 sector, as evidenced by the dislocation in the term structure of old-7-year note spreadsAsset swap O-spreads* for old 7-year notes with maturities between Feb 2016 and Dec 2016

24.5 24.0 23.5 23.0 22.5 22.0 21.5 21.0 Cheap sector

* O-spreads refer to proceeds asset swap spread, calculated based on an OIS discount curve. For a more detailed discussion, see US Fixed Income Markets 2012 Outlook, 11/25/2011.

Exhibit 4: The June / July 2016 spread differential for old 7s is now at extreme levels3.25% July 2016 minus 3.25% June 2016 O-spread differential*; bp

2.0 1.5 1.0 0.5 0.0 -0.5 -1.0 -1.5 -2.0 Jun 11 Sep 11 Dec 11 Mar 12* O-spreads refer to proceeds asset swap spread, calculated based on an OIS discount curve. A wider spread indicates a richer bond.

3.25% Jun16/Jul16 ASW curve

Swap yield curveOn the yield curve, we continue to favor trades that offer asymmetric exposure to higher yield levels. Recent sharp declines in the unemployment rate may well be a puzzle to economists, but they do imply greater uncertainty around current expectations for the funds rate as of some future date (say, 4Q2014). As our Treasury strategists have pointed out, the corresponding risk premium appears to not be fully priced in, and this factorin addition to othersleaves us biased towards higher rates in the medium term (see Treasuries). That said, outright bearish trades suffer from negative carry, a disadvantage since the eventual timing of any move higher in rates is very unclear at this point. Thus, bearish trades with flat holding costs, and/or trades that offer asymmetric upside

in the event of a rise in yields (relative to their downside in a rally) are both attractive in the current context. One example of a trade that offers both these characteristics is a conditional 6th ED / 7Y / 25Y put butterfly (61:61 weights on the wings) constructed using September Eurodollar short-midcurve puts and expiry matched payer swaptions. As seen in Exhibit 5, the yield spread corresponding to this weighted butterfly has been considerably directional with yield levels in recent months; yet, current option market implied volatilities

19


allow the put butterfly to be constructed at zero net premium. In addition, the weighted yield spread is currently too low relative to the level of yields, as also seen in Exhibit 5. Thus, we recommend buying 9/14/2012x7Y payer swaptions versus selling Sep shortmidcurve puts and expiry matched payer swaptions on 25-year tails (see Trade recommendations).

Exhibit 5: The 61:61 weighted 6th ED/7Y/25Y butterfly spread has been positively correlated with intermediate yield levels in recent months, and appears too low currently6Mx7Y swap yield minus (0.61*(6Mx25Y swap yield plus 6th constant maturity ED yield)), regressed against the 6Mx7Y swap yield; %

-0.10 -0.15 -0.20 -0.25 -0.30 -0.35 -0.40 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.36Mx7Y swap yield; %

OptionsVolatility in the swaptions market dropped in the shortexpiry short-tenor sector, but nevertheless outperformed the expectations derived from the skew (Exhibit 6). The rest of the vol surface was unchanged for the week; the exception was 5-year expiries, which rose modestly. Going into the payroll report next week, we continue to remain neutral on gamma, largely because several key drivers are at levels at which our model suggests returns on selling delta-hedged straddles are likely to be flat. Market depth is at levels where it is not a significant driver of realized volatility. Other drivers are largely unchanged from last week, and while a strong employment report could push 3Mx10Y swap yields higher (aiding long gamma positions in the process), our near-term outlook on intermediate maturity yields is still only modestly higher than current levels, implying that realized volatility is unlikely to pick up significantly. While we are neutral on outright gamma on yields, we are bearish on curve volatility (between short and long term swap yields). This bearish view largely stems from our bullish view on correlation between front-end and longer-term swap yields; the potential for an increase in front-end yields in the event of a strong employment report is an additional supportive factor. First, we note that current realized curve correlations are well above implied correlations, in the 2s/10s sector, for instance. Curve correlations, while especially volatile since last summer, are typically fairly stable. The downward swings observed in realized correlations in the latter part of 2H11 were largely driven by two factors the Feds new communications policy action (initiated at the August meeting, when the Fed first indicated that it expected to hold the funds rate at current levels until mid-2013, a projection that has since been extended even further into the future), and an escalation of the European debt crisis, also in the latter part of 2H11. The first factor naturally causes front-end rates to become more immune to typical drivers of yields, thereby becoming less

Y = 0.3973 X1 - 0.9863 R = 61.45% standard error = 0.0404 period = Sep 02,11 - Mar 02,12

Exhibit 6: Swaption volatility outperformed the skew this week

Actual change in swaption implied volatility levels since 2/24/2012 for options expiring between 1 month and 5 years, on 1- to 30-year underlying swap rates, versus expected change*; abp

2 0 -2 -4 -6 -8 -10 1y tails -12 3m2y

2y tails

3m10y

Actual = Expected line 2

* Derived from the swaption skew as of 2/24/12, evaluated as twice the difference between the implied volatility at a strike equal to the ATMF rate plus the yield change over the week, and the implied volatility at a strike equal to the ATMF rate.

-4 -2 -10 -8 -6 0 Expected change in swaption implied volatility*; abp

volatile as well as less correlated to yields elsewhere on the curve. The second factor also produced a decorrelation between front-end and longer-end swap yields, as wider FRA-OIS spreads and falling longerterm yields went hand in hand. Looking ahead, our outlook is for correlations to rise, with both the factors mentioned above now turning supportive of a rise in correlations between front-end and

20


longer-end swap yields. Should the recent positive tone of economic data continue going forwardparticularly the declining trend in the unemployment raterisks to the Feds soft commitment policy will continue to increase; while a near-term reversal of the Feds soft commitment is unlikely, such risks will nonetheless likely cause markets to de-anchor front-end yields to some extent and produce an increase in correlations between short and long-term yields. With respect to the second factor, the second 3-year LTRO is now behind us, FRA-OIS spreads have retraced much of their crisis widening, and will likely no longer be the dominant driver of front-end volatility; correlations between shortterm and long-term swap yields have tended to rise when FRA-OIS volatility declines relative to front-end swap yield volatility, as might be expected and as seen in Exhibit 7. Secondly, all else being equal, higher front-end rates should they occurwould result in a rise in two-year tail volatility relative to 10-year tails (Exhibit 8). This would in turn be bearish for curve volatility; at current levels of 2s/10s realized correlation and the two individual rate volatilities, curve volatility is effectively short two-year volatility, meaning that a rise in the volatility of front-end yields (which would accompany a rise in front-end yields) should depress curve volatility. Taken together, the above arguments, coupled with the fact that risks to front-end yields are biased asymmetrically higher due to the current proximity to the zero bound, argue for an outright bearish stance on curve volatility. Therefore, we now recommend selling delta-hedged 2s/10s singlelook CMS curve straddles on an outright basis (see trade recommendations).

Exhibit 7: With FRA-OIS having nearly fully retraced, FRA-OIS volatility is likely to decline as a fraction of the volatility of front-end yields, producing a rise in correlations between front-end and longer-term yieldsRolling 3-month correlation between 6Mx2Y and 6Mx10Y forward swap yields, versus rolling ratio of 3-month realized volatility on 6M forward FRA-OIS to the 3M realized volatility on 2-year swap yields

1.0 0.8 0.6 0.4 0.2 0.0 Jun 09 3M realized correlation Ratio of FRA-OIS vol to vol of 2year swap yields (inverted)

0.2 0.4 0.6 0.8 1.0 1.2 Mar 12

Dec 09

Jul 10

Jan 11

Aug 11

Exhibit 8: If front-end yields head higher from current levels, 2-year volatility is likely to increase relative to 10year volatility6Mx2Y minus 6Mx10Y implied volatility (bp/day) versus 6Mx2Y swap rate (%)

Trade recommendations Pay in 3.25% June 16 spreads versus receiving in 3.25% July 16 maturity matched swap spreads The swap spread curve term structure for old 7-year notes is dislocated in the Apr-June 2016 maturity sector, with the 3.25% June 16s / 3.25% July 16s spread curve now at very steep levels. In addition, we note that a further 1.4bp cheapening of the June 16s would cause the two bonds dirty prices to be the same; given identical coupons, this would mean that investors could switch into June 16s on a cash neutral basis, and earn virtually identical cash flows, each of which would occur one month sooner. We believe this should limit the further downside risk in this trade to

0.5 0.0 -0.5 -1.0 -1.5 -2.0 -2.5 -3.0 -3.5 -4.0 -4.5 -5.0 0.4

Y = 4.7392 X1 - 0.8289 X1^2 - 6.2843 R = 87.41% standard error = 0.4001 period = Mar 01,10 - Mar 01,12

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

6Mx2Y swap rate (%)

around 1.4bp, resulting in a favorable risk-reward for this spread switch. Buy $100mn notional of 3-1/4% June 16 and pay fixed in $105.3 notional of a 06/30/2016 swap; sell $98mn notional of 3-1/4% of July 16 and receive fixed in $103.5mn notional of a 7/31/2016 swap @ a matched maturity spread differential of 1.3bp (defined July 16 minus June 16 maturity matched spread differential). Carry on this trade is 1.3bp over a 3-month horizon. Three month slide on the trade is 2.6bp.

21


Position for a cheapening of the belly of a EDU3/7Y/25Y butterfly in a sell-off using ED midcurve puts and payer swaptions The EDU3/6Mx7Y/6Mx25Y swap butterfly (61:61 weighted) retains positive correlation with yield levels and should cheapen in a sell-off. Buy $100mn notional of 6Mx7Y payer swaptions (notification 09/14/12, maturity 09/18/19, ATMF 1.787%, strike 1.797%, premium 163.2 bp), sell 1598 0EU2 99.375 puts (midcurve option expiry 09/14/12, EDU3 @ 99.385, strike 99.375, premium 11.5bp), and sell $22.3mn notional of 6Mx25Y receiver swaptions (notification 09/14/12, maturity 09/18/37, ATMF 2.835% and strike 2.845%, premium 527.6 bp). This trade is close to premium neutral, constructed with risk weights of 61% each on the Eurodollar leg and the 6Mx25Y leg, and is expected to be profitable so long as the weighted butterfly yield spread widens as intermediate yields rise. Sell 6M single-look CMS curve straddles on the 2s/10s swap curve We believe that an improving economic outlook and reduced stresses from Europe is supportive of high realized curve correlations. An improving outlook should also result in higher front-end volatility, which, like higher curve correlations, is also bearish for curve volatility. As a result, we look to sell curve volatility. Sell $250mn notional of 6-month curve straddles on the 2s/10s CMS curve (end date 09/06/2012, ATMF CMS 2s/10s yield spread and strike at 1.64%) @ a premium of 48.5bp of notional (implied volatility 5.4bp/day). Unwind longs in the weighted Ultra-long calendar spread We unwind this trade as it was recommended around Treasury futures rollover technicals. Unwind longs in 1000 WNH2 contracts versus covering shorts in 940 WNM2 contracts (US Fixed Income Markets Weekly, 2/10/12) @ a profit of 1.9/32nds. Unwind belly richening position on the EDZ2/EDZ3/EDZ4 butterfly We stop out of this trade as it has steadily underperformed since inception. Unwind longs in 1000 EDZ3 futures versus covering shorts in 650 EDZ2 futures and covering shorts in 550

EDZ4 futures (US Fixed Income Markets Weekly, 1/20/12) @ a loss of 9.9bp of yield. Continue to receive in 10-year swap spreads hedged with 20% risk in September FRA-OIS wideners Continue to receive in $100mn notional of 2% Feb 22 maturity matched swap spreads (stay short $100mn notional of 2% Feb 22s, and continue to receive fixed in $96.2mn notional of a 2/15/2022 swap), and continue paying 20% of the risk (or $18035/bp) in September 2012 3M FRA-OIS spreads (US Fixed Income Markets Weekly, 2/24/12). P/L since inception: 0bp of yield. Maintain 30-year swap spreads wideners Continue to pay in 4.75% of Feb 2041 spreads (stay long $100mn notional of the 4.75% of Feb 2041, and continue to pay fixed in $110.3mn notional of a 02/15/2041 swap). (US Fixed Income Markets Weekly, 2/24/12). P/L since inception: loss of 0.9bp of yield. Maintain 5-year swap spread wideners in a rally Stay long 1,000 FVM2 123.25 calls (FVM2 @ 12304, strike 123.25, expiry 5/25/2012) and stay short $129mn notional of a 05/25/12x8/31/16 receiver swaption (notification 05/25/12, swap start date 07/05/12, maturity 8/31/16, strike 1.073%, ATMF 1.098%) (US Fixed Income Markets Weekly, 2/24/12). P/L since inception: loss of 1.8bp of yield. Stay long 3Mx1Y payer spreads Stay long $100mn notional of a 0.506%/0.756% 3Mx1Y payer swaption spread (1:1 weighted, notification 5/10/2012, maturity 5/14/2013, ATMF 0.506%; buy the ATMF strike and sell the A+.25 strike) (US Fixed Income Markets Weekly, 2/10/12). P/L since inception: loss of 1.2bp of yield. Maintain belly cheapening position on the 1-year forward 2s/5s/10s swap butterfly Continue to pay fixed in $100mn notional of a 1Yx5Y forward starting swap (swap start date 02/14/2013, swap end date 02/14/2018, yield 1.513%), continue to receive fixed in $138.3mn notional of a 1Yx2Y forward starting swap (swap start date 02/14/2013, swap end date 02/17/2015, yield 0.734%) and continue to receive fixed in $29.6mn notional of a 1Yx10Y forward starting swap (swap start date 02/14/2013,

22


swap end date 02/14/2023, yield 2.354%) (US Fixed Income Markets Weekly, 2/10/12). P/L since inception: loss of 0.8bp of yield. Maintain conditional 2s/30s steepeners via 3-month expiry options, versus selling 3-month expiry caps on the 2s/30s CMS curve Stay long $50mn notional 3Mx30Y payer swaptions (notification 5/3/2012, swap start date 5/8/2012, swap maturity date 5/8/2042, ATMF and strike 2.883%) versus staying short $500mn notional 3Mx2Y payer swaptions (notification 5/3/2012, swap start date 5/8/2012, swap maturity date 5/8/2014, ATMF and strike 0.5275%); also stay short $998.5mn notional of 3-month expiry 1-look caps on the 2s/30s CMS curve (end date 5/8/2012, ATMF CMS 10s/30s yield spread and strike at 239.7bp) (US Fixed Income Markets Weekly, 2/3/12). P/L since inception: profit of 0.3bp of yield. Maintain 30-year matched maturity swap spread wideners, hedged with 5-year swap spread wideners and the 5s/30s Treasury yield curve Stay long $50mn notional of 3.125% of Nov 2041s and pay fixed in $51.6mn notional of a 11/15/2041 swap; also, stay long $95mn notional of 0.875% of Jan 2017s and continue to pay fixed in $67.4mn notional of a 01/31/2017 swap (US Fixed Income Markets Weekly, 1/27/12). P/L since inception: loss of 2.8bp of yield. Continue to receive fixed in the belly of a 3-month forward weighted 5s/10s/30s swap butterfly Continue to receive fixed in $100mn notional of a 3Mx10Y forward starting swap (swap start date 04/24/2012, swap end date 04/25/2022) and continue paying fixed in $133mn notional of a 3Mx5Y forward starting swap (swap start date 04/24/2012, swap end date 04/24/2017) and paying fixed in $21.9mn notional of a 3Mx30Y forward starting swap (swap start date 04/24/2012, swap end date 04/24/2042) (US Fixed Income Markets Weekly, 1/20/12). P/L since inception: profit of 0.2bp of yield. Maintain synthetic conditional trade by selling 10s/30s YCSO caps versus 10Y receiver swaptions Sell $500mn notional (i.e., $50K/bp forward DV01) of single-look 6-month caps on the 10s/30s swap curve (end date 6/11/12, strike and ATMF 10s/30s CMS

Closed trades in 2012

P/L reported in bp of yield for swap spread, yield curve and miscellaneous trades, and in annualized bp of volatility for option trades, unless otherwise specifiedTrade Swap spreads 10-year swp sprd widener Front end swap spread wideners via options Pay Aug 20 vs Feb 20 sprds (hedged) Yield curve 5s/10s swp sprd crv steepeners 3m fwd 1s/15s crv flatteners via rec swptns Buy belly of EDZ2/EDZ3/EDZ4 bfly Buy wtd WN calendar sprd PL in 32nds Options relative value Sell 10s/30s YCSO vs 6Mx2Y swptn strddls Buy 6Mx5Y vs 9Mx5Y swptn strddls Entry Exit P/L 01/06/12 01/06/12 01/20/12 Entry 11/04/11 01/06/12 01/20/12 02/10/12 Entry 10/28/11 01/20/12 1/20 2/3 2/3 Exit 2/3 2/3 3/2 3/2 Exit 1/27 2/10 (1.8) (4.4) 1.7 P/L (3.5) (16.1) (9.9) 1.9 P/L 18.2 (0.6)

yield spread at 60.1bp) versus buying $21.3mn notional of 6Mx10Y receiver swaptions (notification date 06/11/12, maturity date 06/13/22, ATMF and strike 2.335) (US Fixed Income Markets Weekly, 12/09/11). P/L since inception: loss of 0.4bp of yield. Stay long 6Mx30Y versus 6Mx10Y swaption straddles Stay long $43.7mn notional of 6Mx30Y swaption straddles (notification date 8/24/2012, maturity date 08/29/2042, ATMF and strike 2.849%) and stay short $100mn notional of 6Mx10Y swaption straddles (notification date 8/24/2012, maturity date 8/30/2022, ATMF and strike 2.220%) (US Fixed Income Markets Weekly, 12/09/11). P/L since inception: loss of 3.0abp.

23

US Fixed Income Strategy US Fixed Income Markets Weekly March 2, 2012 Praveen KorapatyAC (1-212) 834-3092 Renee Park (1-212) 834-7218 J.P. Morgan Securities LLC

Agencies We revise our fair value target for 5-year Agency spreads versus Treasuries to be 5bp narrower at 23bp Treasury released its preliminary annual report on foreign holdings of US securities; China and Japan were the largest net sellers of Agency debt, yet we estimate foreign holdings of Agency debt has increased as a percentage of total debt outstanding FNMA reported its fourth quarter earnings this week, requesting $4.6bn in capital from Treasury

J.P. Morgan 5-year Agency spread target level and model* versus Treasuries, and current spread relative to target (bp, unless indicated otherwise; as of 3/2/2012)Variables Intercept Fed fund expectations (% ) Agency debt foreign holdings, 3M chg ($bn) Agency debt outstanding ($bn) Liquidity factor (bp) 1-year ahead budget surplus ($bn) European debt crisis proxy Coefficient -20.0 22.4 -0.2 0.1 -5.4 0.1 11.6 T-stat -1.8 30.8 -16.8 8.1 -30.2 32.5 22.5 0.13 3.0 760 1.2 -1202 2.2 23 25 1Q12 Est.

Exhibit 1: Five-year Agencies are close to our estimates of fair value versus Treasuries

Forecasted 5-year Agency spread to Treasuries (bp) Actual 5-year Agency spread to Treasuries (bp)

Market viewsOver the past week, Agencies versus Treasuries were largely unchanged with the exception of the 10-year sector, where Agencies cheapened by 2bp versus Treasuries. On the other hand, over the past week the Agency spread curve versus swaps flattened with 2-year Agencies cheapening by 5.5bp, 10-year Agencies largely unchanged, and 20-year Agencies richening by 1bp. Five-year Agencies cheapened about 3.5bp versus swaps amidst a new $4.5bn 5-year FNMA Benchmark Note was issued. The bellwether issue had an unusually low rate of central bank investor participation, and at 7% is the lowest for a new FNMA Benchmark Note since March 2002. Domestic investors took down the majority of the deal at an 88% participation rate. We remain neutral on 5-year Agencies versus Treasuries, given the current level of spreads versus our targets from our revised fair value model (Exhibit 1). Our target for the 5-year Agency versus Treasury spread is now 23bp. In our revised model, we substituted changes in Agency custodial holdings (released weekly by the Fed) with an estimate of changes in foreign holdings of Agency debt as a proxy for demand. The estimated foreign demand uses the higher frequency custody holdings data (which does not separate out Agency debt and MBS flows), as well as monthly and annual TIC data. Previously, we had estimated a relationship between monthly and an

US.fixed.income.markets.weekly

Documents

y us treasury

y germany

y aaa cmbs spd

y cc mbs

year sector

y eur par swap par govt

risky markets

year rates