アメリカ合衆国

Overweight, High-Conviction We reiterate our high-conviction overweight call on the BCA Defense Index as three key demand drivers remain upbeat and will continue to underpin relative industry profitability. First, the global arms race is alive and well and any governments seeking to augment their defense capabilities have to solicit the U.S. defense manufacturers. U.S. defense spending is rising at a healthy clip representing the major source of revenue growth for the industry (top panel). Second, there is a space race going on with China and India working on manned missions to the moon, but recently President Trump signaled that he would like to beat both of these countries to the moon and in outer space. The defense industry also benefits when global space related demand is on the rise. Finally, cyber security remains a global threat and governments are serious about fighting it off decisively given the sensitivity of the data that cyber criminals are after. While defense stocks are not pure-play software outfits combating cyber criminals, recent industry tuck-in acquisitions include such software companies in order for defense contractors to offer one-stop shop solutions to governments. Bottom Line: The BCA Defense Index remains a secular overweight and a high-conviction overweight. Please see our most recent Weekly Report for more details. The ticker symbols for the stocks in the BCA Defense Index are: LLL, LMT, NOC, GD and RTN.      

Highlights Corporate Spreads: The Fed’s dovish pivot prolongs the period of time before the yield curve inverts, thus extending the window for corporate bond outperformance. Investors should remain overweight corporate bonds, with a preference for securities rated Baa and below, where spreads remain wide relative to our fair value estimates. Yield Curve: Investors should barbell their U.S. bond portfolios, favoring long-maturity (> 10 years) and short-maturity (< 2 years) securities while avoiding the 5-year and 7-year notes. This positioning will boost average portfolio yield and will benefit from any future hawkish re-assessment of Fed policy. MBS: Lower mortgage rates have led to a jump in mortgage refinancings and wider MBS spreads. However, MBS spreads remain quite low compared to history. Maintain a neutral allocation to MBS in U.S. bond portfolios. Feature Last December, we laid out our key fixed income themes for 2019 in a Special Report.1 In that report we also introduced a framework for splitting the economic cycle into three phases based on the slope of the yield curve. Specifically, we use the 3-year/10-year Treasury slope to divide each cycle into the following three phases:2 Phase 1 runs from the end of the last recession until the 3/10 slope flattens to below 50 bps. Phase 2 encompasses the period when the 3/10 slope is between 0 bps and 50 bps. Phase 3 begins after the 3/10 slope inverts and ends at the start of the next recession. Clearly, as is illustrated in Chart 1, we are smack dab in the middle of a Phase 2 environment. This has implications for how we should think about positioning a U.S. bond portfolio. Chart 1Firmly In Phase 2 What Makes The Middle Phase Awkward? Table 1 shows annualized excess returns for Treasuries and corporate bonds (both investment grade and high-yield) in each phase of every cycle stretching back to the mid-1970s. Treasury excess returns are calculated relative to cash, as a proxy for the returns from taking duration risk. Corporate excess returns are relative to a duration-matched position in Treasury securities. Table 1Bond Performance In Different Yield Curve Regimes A look at Table 1 reveals why we call Phase 2 the “awkward” middle phase of the cycle. The excess returns earned from taking both duration and corporate spread risk tend to be underwhelming. On duration, we observe that in three of the four complete cycles in our sample, Treasury excess returns are lowest in Phase 2. This lines up well with intuition. The flatter yield curve means that Treasuries offer a lower term premium in Phase 2 than in Phase 1. Meanwhile, Phase 3 periods tend to coincide with rapid Fed rate cuts, and thus large capital gains. Phase 2 periods, in contrast, often contain Fed tightening cycles. On corporate credit, we observe that excess returns tend to be lower in Phase 2 than in Phase 1, but are usually still positive. Returns tend not to turn consistently negative until after the 3/10 slope inverts and we enter Phase 3. Overall, if we know nothing other than that we are in Phase 2 of the cycle, our results suggest that we should take less duration risk in our portfolio than in Phases 1 or 3. Overall, if we know nothing other than that we are in Phase 2 of the cycle, our results suggest that we should take less duration risk in our portfolio than in Phases 1 or 3. The results also suggest that we should prefer corporate credit over Treasuries, though to a lesser extent than in Phase 1. What Makes The Middle Phase Long? In last December’s Special Report, we argued that the U.S. economy would remain in a Phase 2 environment for a long time, at least until late 2019. Our reasoning was that, in the absence of inflationary pressures, the Fed would be reluctant to tighten policy enough to invert the 3/10 curve. The Fed’s recent dovish pivot, and the resultant steepening of the curve (see Chart 1), only prolongs the current Phase 2 environment. We now think it will be well into 2020, and possibly later, before the 3/10 slope inverts and the economy enters Phase 3. One obvious investment implication of an extended Phase 2 environment is that we should remain overweight corporate bonds relative to duration-matched Treasuries. However, we also need to consider valuation before drawing too firm of a conclusion. Charts 2A and 2B show spreads for each corporate credit tier, encompassing both investment grade and high-yield, along with our spread targets. The spread targets are the median levels observed in prior Phase 2 environments, adjusted for changes in the average duration of the bond indexes over time.3 The charts reveal that Aaa-rated bonds already look expensive, while Aa and A-rated bonds are close to fairly valued. Baa-rated bonds are 13 bps cheap relative to our target, while the high-yield credit tiers offer significantly more value. Chart 2AInvestment Grade Spread Targets Chart 2BHigh-Yield Spread Targets As discussed in last week’s report, the Fed’s dovish pivot will cause corporate spreads to tighten in the near-term, but it will take longer before Treasury yields respond by moving higher.4 For Treasury yields to move higher, investors must first become convinced that the Fed’s reflationary efforts are translating into stronger global economic growth. Ultimately, we expect this will occur in the second half of this year and Treasury yields will be higher 12 months from now, as the Fed will fail to deliver the 92 bps of rate cuts that are currently priced. The flat yield curve means that the yield give-up is small, and we expect global growth to improve in the second half of the year. Bottom Line: The Fed’s dovish pivot prolongs the period of time before the yield curve inverts, thus extending the window for corporate bond outperformance. Investors should remain overweight corporate bonds, with a preference for securities rated Baa and below, where spreads remain wide relative to our fair value estimates. Investors should also keep portfolio duration low. The flat yield curve means that the yield give-up is small, and we expect global growth to improve in the second half of the year. Barbell Your Portfolio Chart 3Barbell Your Portfolio For those unwilling or unable to deviate portfolio duration significantly from benchmark, there is another way to bet on the Fed delivering fewer cuts than are currently priced into the market. Investors can run a barbelled portfolio, favoring short-maturity (< 2 years) and long-maturity (> 10 years) securities, while avoiding the belly (5-year/7-year) of the curve. This sort of positioning has a few advantages. First, since the financial crisis, the yield curve has tended to steepen out to the 5-year/7-year point and flatten beyond that point whenever our 12-month Fed Funds Discounter rises (Chart 3). Conversely, whenever the market prices in more cuts/fewer hikes and our discounter falls, the yield curve has flattened out to the 5-year/7-year maturity point and steepened beyond that point. This correlation has been very consistent during the past few years, and continued to hold during the most recent decline in rate expectations. Notice that the 5-year yield has fallen by more than either the 2-year or 10-year yields since our Discounter's early-November peak (Table 2). Table 2The Belly Of The Curve Is Most Sensitive To Rate Expectations The upshot is that, if rate expectations rise during the next 12 months, as we expect, the 5-year and 7-year notes will endure the most damage. The second reason why a barbelled portfolio makes sense is that valuation is very attractive. Chart 4 shows that the 5-year yield is below the yield on a duration-matched 2/10 barbell. It also shows that this 2/5/10 butterfly spread is very low relative to our model’s fair value.5  Chart 42/10 Barbell Is Attractive Versus 5-Year Bullet We run similar fair value models for every possible bullet/barbell combination along the yield curve, and barbells appear universally cheap (see Appendix). Bottom Line: Investors should barbell their U.S. bond portfolios, favoring long-maturity (> 10 years) and short-maturity (< 2 years) securities while avoiding the 5-year and 7-year notes. This positioning will boost average portfolio yield and will benefit from any future hawkish re-assessment of Fed policy.   MBS & Housing: The Implications Of Lower Mortgage Rates Alongside bond yields, mortgage rates have fallen sharply during the past few months, a trend that has important implications for both MBS spreads and future housing data. We consider the outlook for both. MBS Spreads Lower mortgage rates encourage homeowners to refinance their loans, and any increase in refinancing activity puts upward pressure on MBS spreads. Not surprisingly, as mortgage rates have declined we have seen a jump in the MBA Refinance Index and a widening of nominal MBS spreads (Chart 5). Chart 5MBS Spreads Still Historically Tight While spreads have widened somewhat, they remain low compared to history (Chart 5, top panel). As such, we do not see a compelling buying opportunity in MBS. This is especially true relative to corporate credit where spreads are more attractive. Chart 6Limited Upside For Refis With the mortgage rate now below 4%, our rough calculation suggests that approximately 44% of the Bloomberg Barclays Conventional 30-year MBS index is refinanceable. A regression of the MBA Refi Index versus the refinanceable share suggests a fair value of 2014 for the Refi Index, slightly above its actual level of 1950 (Chart 6). We also calculate that a further drop in the mortgage rate to below 3.5%, where it troughed in mid-2016, would increase the refinanceable share to 77%. Our regression translates this 77% share to a level of 3309 on the Refi Index. It should be noted that when the refinanceable share rose to 77% in 2016, the MBA Refi Index peaked at 2870. This means that our simple regression analysis probably overstates the surge in refis that would occur if mortgage rates fell another 50 bps. In addition, we think it’s unlikely that mortgage rates will actually fall back to 3.5%, as they did in 2016, and as such, we are hesitant to position for further MBS spread widening. The improvement in housing actitivty is not uniform across all indicators. We recommend maintaining a neutral allocation to MBS for now. If mortgage rates drop and spreads widen further in the near-term, then a buying opportunity may present itself. Housing Activity Chart 7Housing Activity: A Mixed Picture The drop in mortgage rates will also have a significant impact on housing activity data. This is important because, as we have demonstrated in prior reports, housing activity data – particularly single-family housing starts and new homes sales – are reliable indicators of U.S. recessions and interest rates.6 By all measures, housing activity weakened significantly as mortgage rates surged in 2018. But it has improved somewhat now that mortgage rates have declined. However, the improvement is not uniform across all indicators (Chart 7): New home sales jumped sharply early this year, then fell back more recently. The current trend is neutral, with the latest monthly print very close to the 12-month moving average (Chart 7, top panel). Housing starts and permits are both trending below their respective 12-month moving averages, though by less than in 2018 (Chart 7, panel 2 & 3). Existing home sales have popped, and are now exerting upward pressure on the 12-month average (Chart 7, panel 4). Likewise for mortgage purchase applications (Chart 7, panel 5). Homebuilders also report that lower mortgage rates have led to a jump in sales activity (Chart 7, bottom panel).  With mortgage rates still low, the tentative rebound in housing activity data should continue in the coming months. Looking further out, we see significantly more upside in single-family housing starts and new home sales as builders shift construction toward lower-priced properties. The Bifurcated Housing Market Beyond the large swings in mortgage rates, another trend has significantly influenced housing activity in recent years. For the past few years, homebuilders have focused their attention on higher priced homes, and that segment of the market now looks oversupplied. Data from the American Enterprise Institute Housing Center show that the recent deceleration in home prices has been driven by falling prices for the most expensive homes. Homes in the lowest price tier have seen prices accelerate (Chart 8).7 The divergence is also evident in the supply data. New home inventories are roughly consistent with average historical levels, while existing home inventories are incredibly low (Chart 9). In fact, new home inventories now represent 6.4 months of demand while existing home inventories represent 4.3 months of demand (Chart 9, panel 3). Such a wide divergence is historically rare. Chart 8An Oversupply Of High ##br##Priced Homes... Chart 9...And An Undersupply Of Low Priced Homes   The divergence between an oversupply of new homes and an undersupply of existing homes is a result of new construction having focused on higher priced homes in recent years. The median price for a new home used to be only slightly above the median price for an existing home, but the difference shot up to above 75k during the past few years (Chart 9, bottom panel). More recently, the price differential between new and existing homes has started to fall, as builders are starting to recognize that the greater growth opportunity lies at the low-end of the market where demand is strong relative to supply. As this supply-side adjustment plays out, it will provide an additional boost to new homes sales and housing starts going forward. Appendix The following tables present the current read-outs from our butterfly spread models. We use these models to identify opportunities to take duration-neutral positions across the Treasury curve. The following two Special Reports explain the models in more detail: U.S. Bond Strategy Special Report, “Bullets, Barbells And Butterflies”, dated July 25, 2017, available at usbs.bcaresearch.com U.S. Bond Strategy Special Report, “More Bullets, Barbells And Butterflies”, dated May 15, 2018, available at usbs.bcaresearch.com Table 3 shows the raw residuals from each model. A positive value indicates that the bullet is cheap relative to the duration-matched barbell. A negative value indicates that the barbell is cheap relative to the bullet. Table 3Butterfly Strategy Valuation: Raw Residuals In Basis Points (As of June 27, 2019) Table 4 scales the raw residuals in Table 3 by their historical means and standard deviations. This facilitates comparison between the different butterfly spreads. Table 4Butterfly Strategy Valuation: Standardized Residuals (As of June 27 2019) Ryan Swift, U.S. Bond Strategist rswift@bcaresearch.com Footnotes 1 Please see U.S. Bond Strategy Special Report, “2019 Key Views: Implications For U.S. Fixed Income”, dated December 11, 2018, available at usbs.bcaresearch.com 2 We use the 3/10 Treasury slope in place of the more commonly referenced 2/10 slope because it is a close proxy that provides an additional 14 years of historical data. 3 For more details on how we arrive at our spread targets please see U.S. Bond Strategy Weekly Report, “The Value In Corporate Bonds”, dated February 19, 2019, available at usbs.bcaresearch.com 4 Please see U.S. Bond Strategy / Global Fixed Income Strategy Weekly Report, “The Fed’s Got Your Back”, dated June 25, 2019, available at usbs.bcaresearch.com 5 For more details on our yield curve models please see U.S. Bond Strategy Special Report, “Bullets, Barbells And Butterflies”, dated July 25, 2017, available at usbs.bcaresearch.com 6 Please see U.S. Bond Strategy Weekly Report, “More Than One Reason To Own Steepeners”, dated September 25, 2018, available at usbs.bcaresearch.com 7 Low-tier homes are those in the bottom 40% of the price distribution in each metro area. High-tier homes are those that are both in the top 20% of the price distribution and exceed the GSE loan limit by more than 25%. For further details: http://www.aei.org/wp-content/uploads/2019/06/HPA_market_conditions_report_June_2019.pdf Fixed Income Sector Performance Recommended Portfolio Specification