View the First Principles Quarterly

SPRING 2015

FIRST PRINCIPLES QUARTERLY

FPCM is a privately held investment management firm with expertise across the global fixed income securities and derivatives markets. Founded in 2003, FPCM is a Registered Investment Advisor. FPCM is a fixed income investment manager with over $10bn in assets under management. We provide customized investment portfolios for our institutional clients, which include endowments and foundations, commercial banks, insurance companies and corporations. To our private clients - which include single- and multi- family offices, trusts and high net worth individuals - we offer various investment management and wealth

maximization strategies.

SPRING 2015 FIRST PRINCIPLES QUARTERLY

Contact: In This Issue

First Principles Capital Management, LLC CIO PERSPECTIVE ...... 2

140 Broadway, 21st Floor RATES, INFLATION AND MUNIS ...... 5 New York, NY 10005 Tel: 212-380-2280 MORTGAGE-BACKED SECURITIES ...... 9 Fax: 212-380-2290 www.fpcmllc.com CORPORATE CREDIT ...... 13

EMERGING MARKETS DEBT ...... 17

The information contained herein has been prepared solely for informational purposes and is not an offer to buy or sell or a solicitation of ASSET-BACKED SECURITIES ...... 20 any offer to buy or sell any security. First Principles Capital Management, LLC (“FPCM”), or any of its affiliates, do not make any representation or warranty, express or implied, as to the accuracy or completeness of the information contained herein, and the information contained herein should not be relied upon as a promise or representation whether as to the past or future performance. The information contained herein includes estimates and projections that involve significant elements of subjective judgment and analysis. These statements are not purely historical in nature, but are “forward-looking statements”. They may include, among other things, projections, forecasts, targets, sample or pro forma investment structures, portfolio composition and investment strategies. These forward-looking statements are based upon certain assumptions. Actual events may differ from those assumed. FPCM or any of its affiliates do not make any representations as to the accuracy of these forward looking statements or that all appropriate assumptions relating thereto have been considered or stated and none of them assumes any duty to update any forward-looking statement. Accordingly, there can be no assurance that estimated returns or projections can be realized, that forward-looking statements will materialize or that actual results will not be materially lower than those presented. FPCM and its affiliates disclaim any and all liability as to the information contained herein or omissions here from, including, without limitation any express or implied representation or warranty with respect to the information contained herein. The information contained herein is confidential and proprietary to FPCM and its affiliates. This material and information should be treated as strictly confidential and cannot be disclosed to any other party other than the recipient and its advisers. Notwithstanding anything to the contrary contained herein, the recipient (and each employee, representative, or other agent of the recipient) may disclose to any and all persons, without limitation of any kind, the tax treatment and tax structure of the transactions described herein and all materials of any kind that are provided to the prospective investor relating to such tax treatment and tax structure (as such terms are defined in Treasury Regulation section 1.6011-4). This authorization of tax disclosure is retroactively effective to the commencement of discussions with prospective investors regarding the transactions contemplated herein. By accepting this information, the recipient agrees to be bound by all of the limitations described herein.

SPRING 2015 FIRST PRINCIPLES QUARTERLY

CIO PERSPECTIVE

Doug Dachille Get Real! CEO & CIO In recent weeks, a number of investment managers have offered their views on European interest rates, with a [email protected] particular focus on the yields on German Bunds. One fixed income manager described the ten-year German bund as 212.380.2281 “the short of a lifetime.” A prominent hedge fund manager added it’s crazy to hold nominal bonds with a negative yield, while another fixed income manager suggested a 100 times leveraged short of the German two-year note could produce returns of 20 percent. My response to these comments is to Get Real!

Negative Nominal Yields – The Practical Reality of the Theoretically Infeasible There is little controversy regarding the rationality and economic fortunes of investors purchasing nominal bonds with negative yields, assuming these investors have an investment alternative that enables them to earn a nominal rate of zero for a comparable credit risk in their base currency. Since bank depositors are afforded the option of redeeming deposits in exchange for paper currency, it would seem irrational for investors to ever accept a negative nominal yield on deposits or government bonds. Or is it?

Essentially, paper currency is the economic equivalent of a government bond with a zero nominal yield and a stable asset value in local currency terms. This currency conversion option has long served as the source for the assumed zero rate boundary for nominal interest rates. This option also strikes fear into the hearts of central bankers since it enables depositors to adversely impact banking system reserves and potentially confound expansionary monetary policy.

The traditional monetary policy response to economic weakness typically involves the purchase of government bonds, which causes a commensurate increase in banking system reserves. The expansion of bank reserves serves to increase lending capacity and to lower interest rates. However, bank depositors can elect to redeem deposits in

FIRST PRINCIPLES QUARTERL Y CIO PERSPECTIVE exchange for paper currency and drain the banking the prepayment option is actually a perfectly rational system of expansionary reserve balances targeted by the decision by an individual borrower when her unique central bank. Thus, the action of depositors could serve to constraints are considered. neutralize monetary policy. A similar phenomenon exists with respect to the exercise of Additionally, the seemingly logical presumption that the currency conversion option. Redeeming a bank deposit investors would efficiently and rationally exercise this for paper currency transforms an electronically stored asset conversion option, placing a floor on nominal interest rates into a physically stored one. There are a number of friction at zero, has been another longstanding concern of central costs associated with holding paper currency, including bankers who fear they would lack any policy tools to explicit storage and insurance costs, as well as implicit costs stimulate an economy experiencing deflation. associated with the inconvenience and the reduced timeliness in making payments and transfers. The Since nominal interest rates were presumed to be floored at magnitude of the friction costs will vary by holder zero, the key objective of stimulatory monetary policy was depending upon scale and transaction frequency. to drive the real yield (nominal yield minus inflation) to as Although paper currency theoretically enjoys a zero negative a rate as possible, as quickly as possible. For once nominal yield, in reality the friction costs of managing paper inflation turns negative, it was presumed to be impossible currency holdings can result in an equilibrium negative for central bankers to maintain stimulatory negative real nominal yield – a rational economic reality to what is yields if nominal yields were floored at zero. seemingly ludicrous theoretically.

Central bankers around the globe likely breathed a While zero may not be the lower boundary for nominal collective sigh of relief when nominal yields breached the interest rates, central bankers should not assume there is no theoretical barrier. The ability of central bankers to drive lower boundary. As central bankers attempt to drive the and maintain nominal interest rates at negative levels is nominal yield more negative, the threshold of breakeven now presumed to be a new monetary policy tool to fight cost and inconvenience of holding physical currency deflation by many market participants. However, bankers should begin to be breached for more and more should not set off on their victory lap just yet. individuals, making it economically rational to exercise the conversion option. This would serve to limit the degree of As discussed earlier, efficient exercise of the currency negativity. conversion option is central to flooring nominal interest rates at zero. However, there are numerous examples in financial Central bankers should also not underestimate the impact markets where individuals do not, or cannot, exercise of ingenuity. In the U.S., there are a number of easy ways to options that are seemingly economic to exercise. The best circumvent negative yields on deposits and government example of this is found in the U.S. mortgage market. bonds with reduced storage and inconvenience costs. One Borrowers are afforded the option to refinance their loans simple strategy involves pre-paying future estimated federal without incurring a penalty. Yet, it is routine to find tax liabilities. Doing so would be the economic equivalent mortgage loans with interest rates above the prevailing of purchasing a portfolio of government bonds with market rate that are not refinanced. laddered maturities at yields of zero.

For some subset of these borrowers, the decision to not In lieu of redeeming deposits for paper currency and refinance is irrational. For most, the decision is completely incurring significant storage costs, a depositor could drain rational. Individual borrowers each face a unique set of her account by writing a handful of bank checks made out friction costs that include taxes, transactions costs and to herself. When the bank check is issued, the deposit appraised home equity. These borrower specific factors account is effectively drained, allowing the former serve to change the mortgage rate threshold for which it is depositor to avoid the application of negative interest rates economically rational for a borrower to execute a to her account prospectively. The storage costs for $10 refinancing. In other words, seemingly irrational exercise of

FIRST PRINCIPLES QUARTERL Y CIO PERSPECTIVE million held in the form of a handful of bank checks is much of the difference in these yields can be accounted certainly a lot lower than storing 100,000 one hundred dollar for by the differences in the respective inflation rates. After bills. adjusting nominal yields in Europe and the U.S. for their respective levels of expected inflation, real yields in Europe To reduce the associated costs of holding physical turn out to be only modestly more negative than U.S yields. currency to a bare minimum, the “securitization” of cash If European bonds are considered the “short of a lifetime” would be the ultimate innovation, let’s call it “Cash for while real yields are only modestly more negative than CUSIPs.” Rather than having individuals separately store those of U.S. bonds, shouldn’t these managers also be physical currency and suffer the inconvenience of questioning the sanity of investing in U.S. bonds with accessing it periodically to transact, a securitization would negative real yields? consolidate currency storage to reduce costs through scale pricing. The issuance of traded securities backed by the While the ECB is likely to continue to provide support to stored cash would ease its use for transactions. Allowing the European bond markets, the Federal Reserve is much closer security to be redeemed in physical currency at the to embarking on a policy of “normalizing” interest rates. demand of the owner would ensure the traded price of the Persistent negative real yields are certainly not normal, so security could not be lower than the principal amount. the process of policy normalization should result in an Paying a bill with securitized cash would be easily increase in real yields toward levels that make economic accomplished by selling a security holding. sense – zero would be a good start for short and intermediate term real yields. Getting Real In order for it to be rational for an individual to make an Since an unanticipated rise in real yields could adversely investment and sacrifice current consumption, there must impact the valuation of many financial assets, shorting be a reasonable expectation that the inflation-adjusted intermediate maturity U.S. TIPS at negative real yields may expected return (real rate of return) will be positive. In this be the true short of a lifetime, while offering the most cost case, the individual is rewarded for her willingness to forgo effective downside protection for risky financial assets. current consumption with a greater amount of consumption in the future. However, if the real rate of return is negative, the decision to defer consumption to make investments leads to a reduction in the individual’s future standard of living. Sacrificing current consumption to impair your future standard of living does not sound like an attractive investment proposition.

While a prominent hedge fund manager declared it to be crazy for investors to purchase nominal bonds with negative yields, nominal yields should not be the metric used to evaluate investor sanity. Should bonds with negative nominal yields in a country experiencing deflation be considered irrational investments, while positive yielding nominal bonds in a country experiencing high inflation are considered to be rational? The arbiter of this debate is not the nominal yield. Instead it is the real yield (nominal yield less inflation).

Though nominal yields on European bonds are lower than U.S. nominal yields, and even negative in some cases,

SPRING 2015 FIRST PRINCIPLES QUARTERLY

RATES, INFLATION AND MUNIS

David Ho Liquid Rates MD, Asset Management The nominal liquid rates market rallied in the first quarter of 2015. For the quarter, the Treasury market rallied 21 bp, [email protected] LIBOR swap rates decreased 21 bp, and TIPS rallied 58 basis points. Global central bank easing and lackluster economic 212.380.2292 data both contributed to the rally.

Treasury Yields 3/31/2015 QUICK READ 3.0% 2.5% 2.5%  Nominal rates rally for the quarter due in part 1.9% 2.0% to global central bank easing 1.4% 1.5% 1.0% 0.6%  Fed has been preparing the market for rate 0.5% hikes later this year 0.0% 2Y 5Y 10Y 30Y  Muni new issue supply has picked up as 2Y 5Y 10Y 30Y issuers take advantage of low rates. QoQ change [bp] (11) (28) (25) (22) Expected to continue

 Credit stress to the municipal markets could 2015 has seen tremendous activity from global central come from several areas, both high-profile banks: 24 central banks eased monetary conditions in the and less discussed quarter. The European Central Bank alone announced an €1.1 trillion asset-purchase program. Some actions are based on burgeoning growth and disinflation concerns while others are motivated by preemptive strikes to devalue currency. These quantitative easing programs have put tremendous downward pressure on yields globally.

Disappointing economic data domestically also helped depress yields. Negative surprises happened frequently on many fronts: GDP, housing, inflation, etc. Some negative surprises could be attributed to volatile energy prices, the strong dollar, and weather conditions. One bright spot economically had been the labor market, but as seen in the latest employment report, even that strong pillar is starting to show some signs of cracks.

FIRST PRINCIPLES QUARTERLY RATES, INFLATION AND MUNIS

TIPS Yields 3/31/2015 to show signs of weakness. Such a scenario would put the Fed in a bind given all the rhetoric and careful nudging of

1.0% 0.7% the market.

0.5% 0.2% Municipal Bonds -0.4% 0.0% Municipal bonds, for the second quarter in a row, under- 2Y 5Y 10Y 30Y performed the liquid rates. For 1Q 2015, the Bloomberg AAA -0.5% -0.8% general obligation yield curve decreased by 4 bp. In comparison, Treasuries rallied 21 bp during the quarter. -1.0% Supply, reduced dealer balance sheet, and increasing

credit concerns are the main cause for the under- 2Y 5Y 10Y 30Y performance. QoQ Change [bp] (114) (69) (32) (17) In the first quarter, new issue supply picked up 63.3% from the year prior at $105.4 billion vs. $64.6 billion from 1Q 2014. Talking the Talk Issuers took advantage of low interest rates to issue new Throughout the first quarter, the Fed has been busy trying to bonds or to refinance existing ones. From conversations prepare the market for a rate hike later this year. In a with dealers, the supply picture will be very robust into the speech to the Economic Club of New York on March 23rd, second quarter as well. For the year, many dealers are Fed Vice Chairman Stanley Fischer said, “It is widely projecting over $400 billion of supply, which would be an expected that the rate will lift off before the end of the increase of 18.5% over last year’s issuance of $338 billion. year.”

And Fed Chair, Janet Yellen, broadcast in a speech in San AAA GO Muni Yields 3/31/2015

Francisco on March 27th she “expects conditions may 4.0% warrant an increase in the federal funds rate target 2.9% sometime this year.” 2.0% 2.0% 1.3% However, Fed members were also careful to emphasize 0.5% that such normalization would be data dependent, and 0.0% the pace would be gradual. Fischer hedged, “We 2Y 5Y 10Y 30Y anticipate that it will be appropriate to raise the target range when there has been further improvement in the labor market and we are reasonably confident that 2Y 5Y 10Y 30Y inflation will move back to our 2 percent objective over the QoQ change (bps) - (7) (11) 3 medium term.”

Yellen also clarified further tightening in monetary policy after the first rate increase will be warranted “provided that In years past - when dealers were allowed to take more risk inflation shows clear signs over time of moving up toward 2 - there was less market volatility when supply and demand percent in the context of continuing progress toward became out of sync. In the past, dealers took down a maximum employment.” larger portion of unsold bonds onto their books when supply was high, warehoused them, and then sold them when The take away? As long as inflation doesn’t materially demand materialized from coupon payments, maturity deteriorate and as long as the labor market remains where redemptions, etc. it is, the Fed will move. What’s not so clear, however, is what happens if inflation doesn’t pick up while labor market starts

FIRST PRINCIPLES QUARTERLY RATES, INFLATION AND MUNIS

Now, banks have to satisfy various liquidity and asset ratios; which translated roughly to 29% of Detroit’s median family their available balance sheets have shrunk. The impact income of $31,800. percolates down to all municipal bond dealers in general, and the willingness of dealers to warehouse deals has been Chicago will need some significant one time measures - greatly curtailed. Most new issues have been aggressively such as sales tax hikes - to prevent it from falling further into priced to reduce the amount of unsold bonds. financial abyss. Without such measures, restructuring of debt or pension benefits are very likely. We believe this On the other hand, liquid rates product continues to find scenario would impact the general muni market greater reserve or flight-to-quality bids from overseas central banks than a Puerto Rico default. However, given the well- and private investors. Therefore, munis will likely publicized and long history of the debt woes of Chicago underperform taxable liquid rates products again in the and Illinois, the magnitude of general market impact from second quarter based on supply/demand dynamics alone. default or restructuring would still be muted.

Beyond second quarter, however, coupon payments, What could be worse? maturing bonds and call redemptions are expected to A bigger shock to the market would occur if another state ramp up. A lot of cash will be chasing after bonds once such as New Jersey or Connecticut gets further supply subsides. This bodes well technically for the downgraded due to their liability burdens. municipal bond market in the third and fourth quarters. By some financial measures, these two states are close to or High Profile Credit Stress even worse off than the financial quagmire called Illinois. On the credit front, there continued to be high profile stress For comparison, we: in the muni market. Puerto Rico continued its struggle to buy time to delay the inevitable. Their Electric Power  Looked at the liabilities of each state, defined as Authority (“PREPA”) appears in a perpetual state of bonded debt + unfunded pension liabilities negotiations with creditors to extend debt obligations. A big  Divided those liabilities by population to obtain hurdle is a $416 million payment of interest and principal on liabilities per capita July 1. Will PREPA avoid default or be forced into To incorporate the wealthiness of residents into the analysis, restructuring? we divided the liabilities per capita by median household The dominant players in Puerto Rican credits are mostly income to see what percentage of the annual income hedge funds (and a couple of well-publicized mutual required to pay off the liabilities. funds). Therefore, whatever the outcome of Puerto Rico, we believe the immediate impact to the larger muni market will be muted.

Chicago and Illinois continued to garner a fair amount of attention last quarter. The woes facing the city and the state have been well publicized. By some measures, the financial condition of Chicago is worse than Detroit before Detroit filed for bankruptcy protection. For example, for Chicago, the sum of general obligation debt, property tax supported bonds, and unfunded pension obligations amount to about $19,300 per capita, or roughly 41% of Chicago’s median family income of $47,300. In comparison, Detroit’s general obligation debt plus unfunded pension obligation was about $9,300 per capita,

FIRST PRINCIPLES QUARTERLY RATES, INFLATION AND MUNIS

By this measure of financial burden, Connecticut is not far behind Illinois. Including unfunded other post-employment benefits (“OPEB”) as part of the states’ liabilities, both New Jersey and Connecticut are in worse shape than the Prairie State.

Liability Burdens of Illinois, NJ and CT

Median Liabilities/Capita/ Liabilities Population Liabilities/Capita Household Income Median Household Income Illinois 144,574,452,000 12,880,000 $11,225 $56,796 19.76%

NJ 92,924,716,390 8,899,339 $10,442 $71,637 14.58%

CT 41,739,198,000 3,574,097 $11,678 $64,032 18.24%

Liability Burdens, including OPEB

Liabilities Median Liabilities/ Capita/ (including OPEB) Population Liabilities/ Capita Household Income Median Household Income Illinois 179,062,452,000 12,880,000 $13,902 $56,796 24.48%

NJ 159,724,716,390 8,899,339 $17,948 $71,637 25.05%

CT 63,739,198,000 3,574,097 $17,834 $64,032 27.85%

SPRING 2015 FIRST PRINCIPLES QUARTERLY

MORTGAGE-BACKED SECURITIES

Mattan Horowitz Tracking the American Dream VP, Asset Management Pick up any newspaper over the past several years, and you are likely to see an article about the impact of the [email protected] Fed’s purchasing $1.7 trillion in agency mortgage-backed securities (“MBS”) on the housing market. Let’s take a look 212-324-6018 at something which receives much less fanfare - the impact the housing market has had on the agency MBS market. In order to do any analysis, we first need to find a suitable way QUICK READ to quantify the housing market. A number of different approaches paint very different pictures:  Different measures of the housing market Percentage Change in Housing from 2000 paint different pictures of performance 120% Case-Schiller CPI NAR  Home price appreciation has had a Zillow FHFA 100% significant impact on MBS

80%  Recent drivers of home prices will likely be different going forward 60%

40%

20%

0% 3/2000 9/2003 3/2007 9/2010 3/2014

 Zillow – Zillow collects a variety of data, including prior sales, county records, tax assessments, real estate, and data homeowners enter about their home on Zillow’s website. Zillow then runs a suite of automated valuation models (“AVM”) on over 100 million homes in the US to estimate home values which ultimately gets used to create a national home price index. While this index incorporates a lot of data, it has a lot of model risk and may not be the most parsimonious approach to quantifying the housing market.

FIRST PRINCIPLES QUARTERLY MORTGAGE-BACKED SECURITIES

 National Association of Realtors (“NAR”) – The NAR This is no small task, which is why there are 83,700 real estate calculates a median home price index using appraisers in the US, according to the Bureau of Labor existing home sales. The NAR index is more volatile Statistics. An immense amount of effort is given to valuing than other measures because the composition of homes today – projecting home prices into the future is homes sales changes over time. For example, the exponentially more difficult. index could increase simply because a larger amount of high priced homes are being sold. This HPA and MBS methodology is useful for realtors who typically With a relevant housing index in hand (FHFA), let’s explore receive commissions as a percentage of home how the recent appreciation in home prices has impacted price sales as compensation. the mortgage market.

As home prices appreciate, mark-to-market loan-to-value  Case-Schiller Index – This is the most well-known (“LTV”) ratios fall. The lower a borrower’s LTV, the lower their index and is similar to the also popular CoreLogic mortgage rate will be (they pay less in risk-based pricing Index (they are produced by the same company). adjustments), and the easier it will be to qualify for a The Case-Schiller Index only looks at repeat-home mortgage. This effect can be clearly seen in the cohort of sales. This has the benefit of creating a more 30 year fixed rate loans originated in 2011 with high LTVs apples-to-apples comparison of home prices, but it (over 105% LTV) and mortgage loan rates of around 5% excludes some useful data: non-repeat home sales. (investors in MBS only get 4.5%). These borrowers have very high LTVs because they refinanced through a special  Federal Housing Finance Authority (“FHFA”) – This program, the Home Affordable Refinance Program index is calculated using repeat transactions, both (“HARP”), which allows underwater borrowers to refinance sales and refinancings, on all mortgages backed their mortgage, but they can only go through the HARP by Fannie Mae and Freddie Mac. While this index is program once! Therefore, even though mortgage rates restricted to only loans guaranteed by Fannie Mae have fallen since they took out their mortgage, it is still and Freddie Mac, it has increased coverage by difficult for these borrowers to refinance again – unless the including refinancings in addition to home sales; value of their home goes up significantly. there were a lot more refinancings than home sales during Quantitative Easing!

 Consumer Price Index (“CPI”) - For the purpose of calculating CPI, an entirely different approach is used to calculate housing costs, which shows that housing costs have gone up steadily since 2006. CPI calculates the cost by looking at owners’ equivalent rent – which ignores housing prices because owning a home is a capital investment and not a consumed good.

Out of all of these indices, clearly the FHFA housing index is the most relevant for the agency MBS market. Also note: whenever a mortgage is underwritten for Fannie, Freddie, FHA, VA, etc., a simple index is not used to price the home, rather the house must get valued by an appraiser (there are exceptions for some streamline refinancing programs).

FIRST PRINCIPLES QUARTERLY MORTGAGE-BACKED SECURITIES

2011 High LTV FH 4.5s: Voluntary CPRs by HPA Buckets 2012 FH 3.0s: Voluntary CPRS by HPA Buckets

25 30 8 30

CPR % in Bucket CPR % in Bucket 7 25 25 20 6

20 20 5 15

15 4 15

10 3

10 10 Bucket HPAin Cohort of % % of Cohort in HPA Bucket HPAin Cohort of % 2 5

Voluntary Prepayment Speed [6 mo. CPR] mo. [6 Speed Prepayment Voluntary 5 5 Voluntary Prepayment Speed [6 mo. CPR] mo. [6 Speed Prepayment Voluntary 1

0 0 0 0 <5 5-10 10-15 15-20 20-30 30+ <5 5-10 10-15 15-20 20-30 30+

Prepayments on loans which saw less than a 10% increase A 1% increase in prepayment speeds (1 CPR) might sound in home price appreciation (“HPA”) are very slow. Given trivial, but in a market where mortgages trade well above the original weighted average LTV on these loans is 116%, par, each additional CPR comes at a serious cost to these borrowers are still underwater on their mortgage and investors. not eligible to refinance. Meanwhile, as HPA increases beyond 10%, prepayment speeds increase lock step. Back to the Future Given the current dynamics in the housing market, what An obvious question is why are borrowers who have seen could be the implications for a further increase in home little HPA prepaying at all? The reason is turnover – typically prices? Data from the Census Bureau indicates the people selling their house and using the proceeds to pay homeownership rate has been steadily declining from 2006 off the mortgage. To get a better handle on turnover and rental vacancies have been declining. People are prepayments, let’s look at loans originated with 3.5% 30 renting more and buying less. This implies that the recent year fixed rate mortgages in 2012. These borrowers have increase in home prices was likely not driven by an increase some of the lowest mortgage rates in history, so they are in demand from people looking for primary residences. not incented to refinance. Therefore, all of the prepayments for these loans must be due to turnover – not refinancings. If we break out their prepayments by HPA buckets, we see that borrowers who have seen substantial appreciation in their house, over 15% increase, turnover their loans (likely sell their house) about 1% more annually (1 CPR), than borrowers who have seen less HPA.

FIRST PRINCIPLES QUARTERLY MORTGAGE-BACKED SECURITIES

Home Ownership [%] and Rental Vacancy [%] Distressed Sales as Percentage of Total Sales

70 12 Home Ownership (LHS) Rental Vacancy (RHS) 69 11 68 10 67 9 66 8 65

64 7

63 6

Source: CoreLogic, January 2015

Putting this all together, value based investors utilizing low We can corroborate this assertion with data from leverage were a key component in the recent increase in CoreLogic which shows a historically high percentage of home prices. If we are to see a repeat performance in housing purchases are all-cash - roughly 40% currently. home prices, it will likely not be driven by low leverage value investors given that prices are higher now and there Cash Sales Percent are less distressed opportunities. The next big leg up in housing will need to be driven by leverage and speculative investors. Speculative/leveraged investors are going to face headwinds in terms of access to mortgage credit. The mortgage products (option-arm, negative amortization loans, subprime, etc.) and loose underwriting standards (low-documentation, no-documentation, stated income, etc.) which fueled speculative/leveraged investors in the past are not currently available in the market. It is uncertain if we are going to see speculative/leveraged investors meaningfully return to the market.

Source: CoreLogic, January 2015 Given this view, we like buying bonds backed by loans with very high LTVs (over 125%) which were refinanced through It is probable investors are the ones buying homes with all- HARP in the past 15 months when they are priced similarly cash, not owner-occupiers. Typically, when people buy to older high LTV bonds. The mark-to-market LTV on these homes as their primary residence, they take out a loans is still high – they have not seen as much HPA as their mortgage. One of the draws for investors to the housing more seasoned brethren and they will need to see market has been a large supply of distressed sales – which substantial HPA to be able to refinance and drive are usually sold at a discount. Looking at another dataset prepayment speeds higher. Therefore, these bonds will from CoreLogic, we see that as home prices increased from perform well if HPA moderates. 2011, the percentage of distressed sales decreased.

SPRING 2015 FIRST PRINCIPLES QUARTERLY

CORPORATE CREDIT

Mark G. Alexandridis Unknown Unknowns … MD, Asset Management The expectation of the Federal Reserve firming policy at some point in 2015 will most certainly have an impact on [email protected] the credit markets. The conventional approach is to examine the behavior of the asset class under previous 212.380.2293 episodes of Fed tightening and make a determination as to whether or not there are discernible patterns that may apply prospectively – as Mark Twain averred: “history

doesn’t repeat itself, but it does rhyme.” QUICK READ The data broadly reveal the following:  Will the historical behavior of the credit 1. There is little homogeneity in the pace, magnitude, and markets be relevant as the Federal Reserve subsequent impact on the economy of monetary proceeds to normalize? tightening over the last five cycles. Cumulative increases in rates have ranged between 135 bp and 425 bp and  Fundamentals of the corporate credit market induced recessions on three occasions (1981, 1999, and and spreads do not indicate froth in the 2001) on average 18 months after the first rate rise. market 2. In aggregate, corporate credit spreads tend to widen as the Fed tightens. However, in the last three significant  Historically, credit cycles appear to be instances (1994, 1999, and 2004) spreads were roughly getting shorter and defaults more intense unchanged through the first 6-8 months of the cycle and then subsequently widen, e.g., +84 bp in 1994 and +430  Unprecedented initial conditions, waning bp in 1999, for the next two years. liquidity, and current market structure augur 3. Corporate defaults typically peak two to three years for a greater skew to higher spreads and after the first rate rise. acceleration of credit deterioration

Chart 1: HY Spread [bp] and Fed Funds Rate Target [%]

2,000 20

1,500 15

1,000 10

500 5

0 0

Recession HY spread (LHS) Fed funds rate target (RHS)

FIRST PRINCIPLES QUARTERLY CORPORATE CREDIT

Chart 2: Fed Funds Rate [%] and HY Default Rates [%]

20 Recession Default Rate (LHS) Fed Funds Rate (RHS) 20 16 16 12 12

8 8

4 4

0 0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013

Given the current state of the financial world (level of rates, Chart 3: U.S. GDP YoY [%] health of the economy, credit fundamentals, and structural differences in the credit markets) can one reasonably infer 6 that the historical behavior of the credit markets will be 5 relevant as the Federal Reserve proceeds to normalize the 4 rate regime after a prolonged period at the zero bound? 3

2 Fed Tightening and the Credit Cycle The confluence of when rates begin to rise and where in 1 the business cycle the economy is at that time (early, full, or 0 late in the expansionary phase) is likely to have wide -1 ramifications on the prospects for the credit markets. Of -2 course, the identification of the business cycle is more art -3 than science and is only feasible post hoc – in fact the NBER only published peaks and troughs with a lag of -4 approximately twelve months. 1985 1989 1993 1997 2001 2005 2009 2013 Note: Light blue bars indicate recession years Presently, the cycle is nearly 72 months from the last trough in June 2009. The range (trough to trough) of the last three State of the Corporate Credit Market and Historical wavelengths has been 91 – 128 months. And the expected Sensitivity to Rate Hikes pace (2.0 – 2.5 years) and magnitude (300 – 400 bp) of the The fundamentals (e.g., leverage ratios and current default program are consistent with previous tightening cycles. rates) and market perception of credit risk (spreads) do not However, the absolute level of growth over the post-crisis ostensibly indicate any froth in the market. Although ‘expansion’ has been anemic when compared with leverage is rising, the pace and absolute level is not previous periods that warranted monetary intervention. particularly alarming given that interest coverage ratios are quite high and the term structure of high yield debt has Moreover, other important macro differences prevail been extended over the last four years. And corporate today: (1) the low absolute level of rates, (2) the fragile defaults are either at or near historical lows. nature of the global economy, and (3) the epic rise that the USD has experienced prior to any tightening whatsoever.

FIRST PRINCIPLES QUARTERLY CORPORATE CREDIT

Chart 4: Net HY Leverage Ratio Chart 5: HY Spread Experience After First Fed Hike 5.5 200 Average 5.0 180

4.5 160

4.0 140

3.5 120

3.0 100

2.5 80 1998 2000 2002 2004 2006 2008 2010 2012 2014 Spreads HYNormalized US Barclays 60 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 High yield corporate spreads at 480 bp are well above the Months lows (250 bp) achieved in 2007 but some excess spread can be explained away by the abrupt sell-off in the Energy sector and the technicals created by the CDO/CSO market from 2005-2007 that forced investment grade and high Chart 6: HY Default Rate Experience Following First Fed Hike yield spreads to levels that were at odds with the 3.50 fundamentals. The 400-600 bp spread range is the historical Average norm in the expansionary phase of the cycle. 3.00

Remarkably, the range of outcomes for the change in high 2.50 yield credit spreads and defaults has been quite benign for the first 12 months after the initial rate hike. (Charts 5 and 6 2.00 depict the range of outcomes in HY spreads/defaults months after all rate hikes since 1994. A normalized spread 1.50 of 110 implies that spreads have widened by 10% since the first rate hike. Similarly, a normalized default rate of 2 would 1.00 imply that default rates had doubled since the first rate

Moody's HY Normalized Default Rate Rate Default Normalized HY Moody's 0.50 rise.)

In the worst case spreads increased by 25%, in the best 0.00 1 4 7 10 13 16 19 22 25 28 31 34 case decreased by 10%, and on average were flat over Months the first 12 month horizon. When mapped to total return it is generally hard to envision a loss in a generic high yield portfolio given the ample carry. That range however It is notable that the credit cycles appear to be getting increases decidedly in the successive year for both metrics. shorter and the default intensity more severe. Similarly the Typically, corporate defaults peak somewhere between time to recovery and the strength of recovery in terms of two to three years after the first interest rate hike. default experience also seem to be accelerating. The post- crisis high yield default rate peaked at 15% but fell to 2.5% within three years. Whereas in the previous cycle defaults peaked at roughly 12% and required nearly four years for defaults to recede to 3%.

FIRST PRINCIPLES QUARTERLY CORPORATE CREDIT

Prospective View Unfortunately, it is difficult to suggest that extrapolating the The circumstances are further exacerbated by a well- past experience of how credit spreads/defaults have documented collapse in secondary market liquidity and reacted to Fed tightening cycles in this particular case. As the wholesale change in the composition of corporate in differential equations, the solution/outcome depends bond ownership. US corporate debt has more than trebled critically on the initial conditions. It is precisely those initial since 2008 but dealers have pared their holding by more conditions that differentiate this particular cycle from all than two-thirds. And mutual fund and ETF ownership of the others and – to employ regrettably a well-worn market corporate bond market has soared from 9% in 2000 to more platitude – therefore it will probably be different this time. than 25% presently.

Given six years of extraordinary monetary and fiscal The cocktail of unprecedented initial conditions, waning accommodation coupled with the attendant liquidity, and an arguably less stable market structure consequences to foreign markets, average domestic augurs for: (1) a wider range of outcomes than previously growth, and no inflationary pressure whatsoever, the initial observed tightening cycles with a greater skew to higher conditions preceding liftoff are not only unlike any spreads and (2) an acceleration of the half-life of the credit observed before but were unfathomable pre-crisis. Any hint deterioration (spreads and defaults). In the best case, the or intimation of normalizing policy has exposed the fragile reaction of the credit market may be reminiscent of Ravel’s nature of the market. The taper tantrum and the gripping Bolero – the melody may sound the same but the volume appreciation of the USD well in advance of any rate hike builds through time. underscore the unhealthy risk off/on trading mentality that prevails and the concentration of risk.

SPRING 2015 FIRST PRINCIPLES QUARTERLY

EMERGING MARKETS DEBT

Prasad Kadiyala Strong Dollar Impact MD, Asset Management 1Q 2015 saw a strong rally in Emerging Markets (“EM”) fixed income spreads and rates which mirrored the performance [email protected] in Developed Markets (“DM”). However, the period was 212-380-2297 also highlighted by continued strengthening of the US Dollar resulting in a mixed EM performance.

Emerging Market Bonds Performance QUICK READ Year- 1Q 2015 on-year  EM debt performance was tempered by EMBI+ (hard currency)1 2.01% 5.65% strengthening of the US Dollar GBI-EM Diversified2 -3.96% -11.14% (local currency), US$ return  Fast pace of EM financial markets GBI-EM Diversified (local development mirrored by growth in bond 2.53% 8.97% currency), local currency return fund assets GBI-EM Diversified (local 8.20% 14.04%  Inflation-linked securities have potential to currency), Euro return dampen adverse FX moves

In an environment of intense pressure from FX and commodities selloffs, the relative performance of EM fixed income over the past year is a testament to the increasing maturity and breadth of the EM markets and the vigilance of the respective EM central banks in taking appropriate timely policy actions. The S&P GS Commodity Index is down 40% from its peak in June 2014 while the JP Morgan EM FX Index is down 16.4% (Chart 1) over the same period and down 30% in the last 5 years. In spite of these sharp corrections, the GBI-EM Diversified Index returned 125% over the last 10 years, doubling the performance of the US Treasury Index over the same period. The 10 year returns are comparable to those from the S&P 500 and High Yield index.

1 JP Morgan EMBI+ Index 2 JP Morgan Government Bond Index-Emerging Markets Diversified Index

FIRST PRINCIPLES QUARTERLY EMERGING MARKETS DEBT

Chart 1: FX Performance [%] Chart 3: Assets in Bond Funds Worldwide [USD Billions]

DXY EMCI Euro Brazil Turkey Mexico S Africa 7,000 0% EM Advanced HY Advanced 6,000 -5% 5,000 -10% 4,000

-15% 3,000

-20% 2,000

-25% 1,000 YTD 12/31/13-03/31/15 -30% 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 DXY: US Dollar Index versus major world currencies; EMCI: JPM Morgan index for EM currencies Source: Lipper and IMF

Amazing Growth This amazing growth in fund investments, while greatly Total EM outstanding debt grew from $2 trillion in 2001 to beneficial to EM, has also increased the exposure of these about $12 trillion in 2013, split equally between sovereigns markets to global financial conditions and potential and corporate issuers (Chart 2). The fast paced contagion. Academic studies have shown retail investors development of EM financial markets is further evidenced are generally more reactive to recent market events than by the sharp growth in EM bond funds managed (Chart 3). institutional investors and tend to momentum trade. Over the past 10 years, EM bond fund assets grew from less Institutional investors, on the other hand, tend to be more than $200 billion in 2004 to around $3 trillion by the end of strategic and have longer investment horizons. Fund flow 2014. data show a sharp drop-off in retail flows over the last few years on weakening EM currencies and the taper tantrum Chart 2: Total Domestic Outstanding [USD Billions] starting in mid-2013 (Chart 4). Considering the proclivity of fund flows to impact asset prices, it behooves investors to 7,000 Sov Corp be knowledgeable of the investor base of the target EM 6,000 country and to know the market proportion of foreign 5,000 investors.

4,000 Chart 4: Cumulative Flows into EM Bond Funds [USD Billions] 3,000 140 2,000 Retail Institutional

1,000 90 0 2001 2003 2005 2007 2009 2011 2013 40 Source: Lipper and IMF

-10 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: EPFR and Barclays

EM governments in general have stayed true to their goal of improving accountability and creating more market

FIRST PRINCIPLES QUARTERLY EMERGING MARKETS DEBT friendly environments. The available EM public sector debt Mexico. Signs of inflation pass-through are evident from continues to compare favorably to that of DM in terms of recent inflation prints in some of the EM countries. measures such as Debt/GDP. EM Central Banks have also demonstrated their independence through their recent Chart 5: Rolling YOY Inflation vs. YOY Change in FX actions, providing further comfort to EM investors. A South Africa potential area of concern is EM non-financial corporate YOY Inflation YOY FX Change (RHS) -50% debt which has doubled since 2009 to $2.4 trillion. It is 13% estimated that about $650 billion of this debt will mature by 2017. Rolling this debt could be problematic if the softness 8% in EM markets persists or gets worse. 0%

3% Dampening Adverse FX Prospectively, investments in public sector debt of selective -2% 50% EM countries with high yields looks attractive on a Dec-99 Dec-02 Dec-05 Dec-08 Dec-11 Dec-14 fundamental basis. At this stage of the U.S. economic cycle

(please refer to the Corporate Credit section of this FPQ), with more frequent market references to imminent Fed Turkey tightening action, the primary risk to EM investors is likely to 14% -40% YOY Inflation YOY FX Change (RHS) come from further adverse FX moves. -20% At a recent lunch meeting, Fed Vice-Chairman Stanley 7% Fischer’s response to a question on concerns about US 0% Dollar strengthening is particularly apt: he believes FX evolution is complex and is certain most predictions will be 20% wrong. For this reason, the Fed broadly ignores FX in making 0% Jan-07 Oct-08 Jul-10 Apr-12 Jan-14 policy decisions. Current consensus indicates the Fed tightening moves are mostly priced in, and a year from now, rates are likely to finish close to their forwards. Brazil

An investment in inflation-linked securities of countries with 9% -40% YOY Inflation YOY FX Change (RHS) sharp FX corrections has the potential to dampen the 7% -20% negative FX impact by participating in the higher FX- induced inflation pass-through (Chart 5). 5% 0%

A recent IMF study on FX sell-off induced pass-through3 3% 20% shows the impact on import prices can be very significant, ranging from 100% in Turkey, to 80% in Brazil, and to 20% in 1% 40% 4 Dec-02 Dec-06 Dec-10 Dec-14 India. A Fed study shows forward inflation compensation has tended to exceed targets by 1.0%-1.5% in Brazil and

3 Exchange Rate Pass-Through in the Global Economy: The Role of Emerging Market Economies; MATTHIEU BUSSIÈRE, SIMONA DELLE CHIAIE, and TUOMAS A. PELTONEN; IMF Economic Review Vol 62, No. 1; 2014 4 Are Long-Term Inflation Expectations Well Anchored in Brazil, Chile and Mexico?; Michiel De Pooter, Patrice Robitaille, Ian Walker, and Michael Zdinak; Board of Governors of the Federal Reserve System, International Finance Discussion Papers, Number 1098, March 2014

SPRING 2015 FIRST PRINCIPLES QUARTERLY

ASSET-BACKED SECURITIES

Richard Dolan ABS Issuance — Robust CFO ABS new issuance ended 1Q 2015 at $59 billion, up 11% year-on-year and up 45% quarter-on-quarter5. Higher new [email protected] issuance was driven by auto, equipment, and other ABS 212.380.2283 (consumer loans, franchise fees, and railcar leases), offset by lower credit card and student loans. Auto continues to dominate new ABS issuance, accounting for 53% of all issuance in the first quarter. At the end of March, total consumer ABS outstanding equaled $715 billion, with student loan ABS accounting for $237 billion, autos $187 Rongfeng “Becky” Li, CFA billion, credit cards $183 billion, equipment $27 billion, and VP, Asset Management others at $81 billion. [email protected] In 1Q 2015, the ABS floating rate index5 had 0.23% excess 212-380-2296 return. Container ABS had the best quarterly excess return with 1.1%, while railcar lease ABS had the worst with -1.54% due to the impact of lower oil prices and retrofitting tank cars to meet safety standards. In general, the longer term trend continued: the lower the quality QUICK READ collateral/enhancement, the higher the excess return, as investors continue to chase yield.  Auto continues to dominate ABS new issuance Consumer Debt — Another Record High Consumer credit (excluding mortgages) extended its four-  Student loan performance is the exception year expansion and reached yet another record high of to strong consumer performance $3.3 trillion in December 2014, according to the Federal Reserve. A majority (73%) of consumer credit continues to  ABS protection still strong despite weakening be in the non-revolving category, mostly auto and student auto collateral loans. Both continuing to break records, with auto loans reaching $956 billion and student loans reaching $1.3 trillion. Rising auto financing has been driven by strong vehicle sales, with annualized auto sales topping 17.05 million in March. By comparison, revolving credit, primarily higher- cost credit cards, was $889 billion – relatively unchanged since 2010.

5 Source: Bank of America/Merrill Lynch

FIRST PRINCIPLES QUARTERLY ASSET-BACKED SECURITIES

U.S. Consumer & Mortgage Debt Outstanding [USD Billions] 1,500 15,000

14,500 1,300

14,000 1,100 13,500 900 13,000

700

$ Billions $ 12,500

500 12,000 Mar-06 Jul-07 Nov-08 Apr-10 Aug-11 Jan-13 May-14 Sep-15 Student Loans Outstanding Auto Loans Outstanding Revolviing Consumer Credit Outstanding Mortgage Debt Outstanding (RHS) Source: Federal Reserve

Strong Consumer Performance, Except Student Loans While consumer credit has been expanding, consumer Unlike other consumer credit, student loan delinquency credit performance remains strong not only due to and default rates have stayed high since the Great Recession. The delinquency ratio for all student loans (both historically low interest rates and falling unemployment– but largely due to better underwriting standards. Delinquency federal and private) 90 days or more delinquent rose from and default performance for autos and credit cards are 7.55% in 2Q 2008 to 11.32% in 4Q 2014. The great majority of student loans are federal loans – subsidized by you and me, actually performing better than pre-crisis levels. Default rates for bank credit cards, auto loans, and first mortgages the U.S. taxpayer. The overall increase in student-loan are near ten-year lows. delinquencies is largely due to poor performance of federal student loans, which are not credit underwritten; while private student loans have actually improved dramatically since 2008 with much tighter underwriting standards.

Student Loan Delinquency Stays High, While Other Consumer Debt Default Has Fallen

14% 12% 10% 8% 6% 4% 2% 0% Jun-04 Jun-05 Jul-06 Aug-07 Sep-08 Oct-09 Nov-10 Dec-11 Jan-13 Jan-14 Feb-15

Mortgage Default Rate Bank Card Default Rate Auto Loan Default Rate All Student Loans 90+ Days Delinquent

Source: S&P/Experian, Federal Reserve

FIRST PRINCIPLES QUARTERLY ASSET-BACKED SECURITIES

Our forecast is for this sector to continue to be a major allowing longer repayment terms, higher LTVs, and lower political story as we approach the 2016 Presidential FICO scores - especially for subprime borrowers. According election. We’re troubled that political jawboning is being to S&P, loans with original maturities greater than 60 months echoed by the business media which is starting to view rose to 79.17% in 2014, up from a low of 73.57% in 2010. debt forgiveness as a political eventuality. This will only Some auto loans now have extended tenors of up to 84 reinforce poor credit trends. But with $1.3 trillion in months, a record length. Longer-term prime auto loans indebtedness outstanding, we would agree that something perform two times worse in terms of default risk than loans needs to change. with traditional repayment terms of 60 months or less, according to Moody’s. Also, the average LTV ratio for subprime auto increased to 114.78% in 2014, up from Weaker Auto Collateral but ABS Protection Still Strong 111.81% in 2011. So we continue to expect auto Overall auto loan default rate was 1.06% in February 2015, delinquency and default rates to rise over the next few near the lowest level in a decade. However, auto loan ABS quarters. has experienced further weakness in delinquency rates, net losses, and recoveries for both prime and subprime auto Further adding to the changing risk profile, used-vehicle sectors as credit quality continues to weaken. Prime auto prices are expected to decline through 2018 due to several ABS performance is still relatively stable, but subprime auto factors, including burgeoning used-vehicle supplies from ABS Subprime ABS has shown some recent deterioration. recent new-vehicle peak sales, increased percentage of leasing, and the reemergence of the role of incentives to Weaker auto performance has been expected as auto move new vehicles. Since reaching a peak in May 2011, lenders (captives and banks) have been relaxing their the Manheim used vehicle value index (a measure of underwriting standards over the past three years by wholesale auto prices adjusted for mix, mileage and

Auto ABS Net Loss Rates

16.0% 3.0% Subprime Auto ABS (LHS) Prime Auto ABS (RHS) 12.0% 2.0%

8.0%

1.0% 4.0%

0.0% 0.0% Dec-05 Dec-06 Dec-07 Dec-08 Dec-09 Dec-10 Dec-11 Dec-12 Dec-13 Dec-14 Source: S&P

Auto Loan Performance – Yearly Averages

2008 2009 2010 2011 2012 2013 2014 Net loss rate (%) 8.71 9.39 5.85 4.65 4.15 4.93 5.76

Recovery Rate (%) 39.07 41.78 46.08 46.9 47.09 48.6 45.59

60+ day delinquencies 4.18 4.62 3.25 2.49 2.53 2.98 3.51

Source: S&P

FIRST PRINCIPLES QUARTERLY ASSET-BACKED SECURITIES season) has fallen 2.6%. However, used vehicle prices are expected to decline 10% by 2018, according to RVI Risk Outlook. Used vehicle loans make up approximately 73% of subprime and 33% of prime auto ABS collateral in 2014.

Nonetheless, while we expect the auto-finance sector’s “raw material” to weaken over the horizon, the ABS structure continues to offer ample protection for investors. For instance, senior tranches of subprime auto ABS now have enhancement levels that are substantially larger than pre-Great Recession structures, and can withstand double digit loss rates. In addition, senior auto ABS tranches offer investors protection via their sequential-pay structure and the gradual build-up of credit enhancement as deals deleverage over time. So while the sector is beginning to experience some modest credit headwinds, the ABS structures are designed to handle severe credit storms.