Estados Unidos

Overweight While we were a tad early lifting the broad S&P health care sector to an overweight stance mid-year, quarter-to-date, health care stocks are the best performing GICS1 sector up 10%, besting the SPX by 450bps. Such consistent health care outperformance is significant and we reiterate our overweight stance on this defensive sector. Elizabeth Warren’s moves in the polls are a key driver behind the recent health care stellar returns and we would continue to lean against any increase in her probabilities of both winning the Democratic nomination and the 2020 Presidency (see chart). Importantly, Warren’s toning down of her “Medicare For All” speedy adoption to a more pragmatic phase-in approach over a number of years fueled the stampede into health care stocks in general and managed health care (which we are also overweight) in particular. Bottom Line: Any selloff in health care equities due to Warren’s rise in the polls presents an excellent buying opportunity. Stay overweight the S&P health care sector.

Highlights Duration: Incoming data are consistent with our view that global growth is at an inflection point, and will improve during the next few months. As this plays out and recessionary fears fade into the background, we expect the 5-year/5-year forward Treasury yield to settle near 2.5%, 57 bps above its current level. High-Yield: Caa-rated debt has underperformed the duration-matched Treasury index so far this year, despite strong performance for junk bonds overall. We document that weak Caa returns often precede negative returns for the overall junk index. High-Yield: We show several ways in which this year’s Caa underperformance is unique compared to prior episodes. All in all, we conclude that we should not take too strong a signal from the recent Caa spread widening. Remain overweight high-yield in US bond portfolios. The Way Back To 2.5% Chart 1Target 2.5% Worries about a looming US recession peaked in late August when the 2/10 Treasury curve inverted and the 10-year yield hit 1.47%. Since then, some better economic data and the prospect of a “phase 1” US/China trade deal have lifted yields and un-inverted the curve. But the bond market is not yet sending the all-clear. Once recession fears completely fade into the background, we would expect the 5-year/5-year forward Treasury yield to settle near 2.5%. This is the FOMC’s median estimate of the longer-run fed funds rate, and also where the 5-year/5-year forward yield peaked during the last two global growth upturns (Chart 1). We expect that the 5-year/5-year forward Treasury yield will reach 2.5% in the first half of 2020, but global growth needs to rebound for that to happen. At present, we detect some positive signals from our preferred global growth indicators. The Global Manufacturing PMI troughed at 49.3 in July and came in at 49.8 in October (Chart 1, bottom panel). Then last week, Flash PMI data showed further gains in November for the US, Eurozone and Japan (Chart 2). Only the UK saw its manufacturing PMI drop in November, and it accounts for a mere 2% of the global index. There is no Flash PMI estimate for China. We detect some positive signals from our preferred global growth indicators.  More signs of economic optimism are found in regional manufacturing PMIs, which continue to diverge positively from the national number (Chart 3). November data have already been released for New York, Philadelphia, Kansas City and Dallas. All four surveys point to a stronger national print. Chart 2A Bottom In Global PMIs Chart 3Regional PMIs Hooking Up Other data released last week include the Conference Board’s Leading Economic Indicator, which held flat at just above zero in year-over-year terms (Chart 4). The Leading Index is at a key inflection point. A rebound from here would be consistent with the 2015/16 episode (our base case expectation), while a dip into negative territory would sound some alarm bells. Chart 4Keep A Close Eye On Jobless Claims October existing home sales and housing starts came out last week (Chart 4, panels 2 & 3). Both series continue to rebound sharply from the depressed levels seen earlier in the year. This should not be too surprising, given this year’s large drop in mortgage rates. It will be more interesting to see what happens to the housing data as bond yields move higher and the stimulus from low rates fades. We have previously argued that the housing market will provide important clues about where bond yields will peak for the cycle. It will be critical to monitor the housing data as bond yields move higher in 2020.1 One note of caution comes from initial jobless claims, which printed at 227k in each of the past two weeks, slightly above recent levels (Chart 4, bottom panel). Claims remain roughly flat on a 6-month basis, consistent with continued economic recovery. However, a sustained increase would send an important warning sign about the labor market. We will be watching claims closely during the next few weeks. Bottom Line: Incoming data are consistent with our view that global growth is at an inflection point, and will improve during the next few months. As this plays out and recessionary fears fade into the background, we expect the 5-year/5-year forward Treasury yield to settle near 2.5%, 57 bps above its current level. The Puzzling Underperformance Of Caa-Rated Junk Bonds Chart 5The Puzzling Case Of Caa-Rated Junk Bonds Overall high-yield returns have been solid in 2019, but oddly, the lowest-rated junk bonds have not participated in the rally. So far this year, Ba and B-rated junk bonds have bested duration-matched Treasuries by 786 bps and 717 bps, respectively. But Caa-rated bonds have underperformed the duration-matched Treasury index by 87 bps (Chart 5). We usually think of the Caa-rated credit tier as being “higher beta” than the Ba and B tiers. That is, it should perform best in “risk on” environments, and worst in “risk off” environments. With that in mind, this year’s Caa underperformance is puzzling, and raises two important questions that we attempt to answer in this report. Is Caa underperformance a warning sign for the overall junk sector? Can we identify the reasons for this year’s Caa underperformance? And if so, do they suggest a buying opportunity? A Caa-nary In The Coal Mine? To assess whether this year’s Caa underperformance might be a warning sign for overall junk bond excess returns, we ran a few tests using historical data. First, we looked at calendar year excess returns going back to 1996 (Table 1). We then tested the performance of a couple trading rules to see whether Caa performance is a bellwether for the overall index. For the first test, we identified calendar years when junk index excess returns were positive but Caa was the worst performing credit tier. Four years fit this criteria: 1999, 2005, 2014 and 2019. Of the three years other than 2019, two (1999 and 2014) were followed by negative junk index excess returns the next year. Table 1Junk Excess Returns By Calendar Year We also posited that one difference between the Caa and Ba/B credit tiers might be that Caa-rated firms tend to be smaller. We therefore identified calendar years when junk index excess returns were positive but when small cap equities underperformed large cap equities. We identified eight such years. Of the seven years other than 2019, five were followed by negative junk index excess returns the next year. Both rules appear to give a good warning sign for the overall junk index. What if we combine them? We identify three years when junk index excess returns were positive, but Caa was the worst performing credit tier and small cap equities lagged large caps: 1999, 2014 and 2019. Both 1999 and 2014 were followed by negative junk index excess returns the next year. So far the evidence of Caa underperformance being a warning sign for the overall index is quite compelling. But let’s look more closely at the periods flagged by our trading rules. It is only this year that we have seen a large divergence in terms of direction between Caa spreads and overall junk index spreads. Recall that we identified 2019, 2014, 2005 and 1999 as the four years when overall junk index excess returns were positive, but when the Caa credit tier was the worst performer. If we look at the direction of junk spreads in those periods, we see that the direction of Caa spreads tracked the overall index very closely throughout 2014, 2005 and 1999. It is only this year that we have seen a large divergence in terms of direction between Caa spreads and overall junk index spreads (Chart 6). This divergence is odd, and it suggests that this year is unique compared to the other periods identified in our analysis (more on this below). Chart 62019 Is Unique Another reason to doubt the potential relevance of our calendar year analysis is that the decision to use calendar years is arbitrary, and it severely limits our sample size. We therefore run the same analysis using rolling 12-month periods. The results are presented in Table 2. Table 2Predictive Power Of Caa Returns: Rolling 12-Month Periods From December 1996 To October 2019 First, note the baseline result that there are 178 12-month periods of positive junk index excess returns in our sample. Of those 178 periods, 31% were followed by negative excess returns during the subsequent 12 months. If we apply our “Caa Return” filter and look only at 12-month periods when junk index excess returns were positive but Caa was the worst performing credit tier, our 178 examples fall to just 22. Of those 22 episodes, half were followed by negative junk index excess returns during the subsequent 12 months. Our “Small Cap/Large Cap Equity” filter provides a similar 51% hit rate with a larger sample size of 78. In this analysis we also test a “Caa Spread” filter where we scan for 12-month periods when junk index excess returns were positive, but when Caa spreads widened despite tightening in the overall index spread. We identify only 16 such periods, 56% of which were followed by negative index excess returns. We also looked at what happens when we combine two or more of our filters. Using our “Caa Return” and “Small Cap / Large Cap Equity” filters together, we identify only 18 episodes, 61% of which were followed by negative junk index excess returns during the next 12 months. If we take all three of our filters together, we find only 5 episodes, 4 of which preceded a period of negative junk excess returns. Please recall that the most recent 12-month period meets the criteria of all three of our filters. As was the case with our Table 1 results, an important caveat to this analysis is that of the 5 episodes identified by all three of our filters, the direction of Caa spreads never diverged from the direction of the overall index spread. In fact, we could find no historical period other than this year when Caa spreads diverged in direction from the overall index spread for so long. We conclude that Caa underperformance can provide advance notice of negative junk index excess returns, but also that the current period is so unique that it requires further analysis. Can We Explain The Divergence Between Caa Spreads And The Overall Index? As mentioned above, the current period of sharply widening Caa spreads alongside a rangebound overall index spread is unique historically. This not only raises questions about the relevance of the historical analysis we just presented, but also cries out for an explanation. Fortunately, several things appear to explain the odd behavior of Caa spreads. First, changes in index duration. Junk index duration fell dramatically in 2019, but the decline was much larger for Ba and B rated credits than for the Caa tier. If we control for changes in index duration by looking at 12-month breakeven spreads instead of the average index option-adjusted spread, we see that the spread divergence looks much less dramatic (Chart 7). Controlling  for changes in index duration by looking at 12-month breakeven spreads instead of the average index OAS, we see that the spread divergence looks much less dramatic. Second, it’s possible that credit quality has deteriorated more for the lowest-rated credits than for the rest of the junk index. That would explain the spread divergence. However, this appears to not be the case. Our bottom-up sample of high-yield firms shows that debt-to-assets and interest coverage look similar compared to history for both the median high-yield firm and the worst 10% of firms (Chart 8). Chart 7A Duration Story Chart 8Credit Quality Is Not The Culprit   Finally, we consider the sector composition of the different credit tiers. We look at year-to-date sector contributions to each credit tier’s excess returns and find that the difference between Caa and the rest of the index is concentrated in the Energy and Communications sectors (Chart 9). Caa-rated Communications firms underperformed the Ba and B credit tiers because of two Caa-rated firms – Frontier Communications Corp and Intelsat – that ran into problems. As for Energy, we note that the Caa tier has much more exposure to the Oil Field Services sub-sector than the other credit tiers. This sub-sector captures many of the shale players, who have struggled with falling oil prices. Notice that this year’s decline in the WTI oil price tracks Caa spread widening very closely (Chart 10). Chart 9Contribution To Year-To-Date Excess Returns* (%)   Chart 10Blame Energy The Appendix at the end of this report provides a sector decomposition of the different junk credit tiers. Specifically, it presents three tables. One showing the sector weights in each credit tier. A second showing year-to-date excess returns for each sector by credit tier. A third showing the contribution from each sector to each credit tier’s year-to-date excess returns. Investment Conclusions Overall, we are hesitant to make too much of the recent Caa underperformance. Yes, we find compelling evidence that Caa underperformance can be a bellwether for negative high-yield excess returns. However, the behavior of Caa spreads in 2019 doesn’t resemble the prior periods in our analysis very closely. Specifically, the Caa spread doesn’t tend to diverge from the overall index spread in terms of direction, as it has this year. We are also able to identify two compelling reasons for this year’s divergence between Caa spreads and the overall index. The first is the change in relative index duration, and the second is stress in the shale oil sector due to a falling oil price. Spreads should adjust to changes in duration over time, and the stress in the shale sector should ease if oil prices rise as our commodity strategists expect.2 Given the uniqueness of the current period, and our base case outlook for a rebound in global growth, we are inclined to view Caa bonds (and junk bonds more generally) as an attractive buying opportunity in the current environment. But we will keep an eye on the performance of Caa bonds during the next few months. If global growth recovers and the oil price rises, but Caa continues to lag the overall index, then it may compel us to change our view. Appendix Table 3Sector Weights Within High-Yield Corporate Bond Credit Tiers* (%) Table 4Sector Year-To-Date Excess Return* By High-Yield Credit Tier (%) Table 5Sector Contribution To Year-To-Date Excess Return* For Each High-Yield Credit Tier (%) Ryan Swift US Bond Strategist rswift@bcaresearch.com   Footnotes 1  Please see US Bond Strategy Weekly Report, “Position For Modest Curve Steepening”, dated October 29, 2019, available at usbs.bcaresearch.com 2 Our commodity strategists forecast an average price of $63/bbl for WTI crude oil in 2020. Please see Commodity & Energy Strategy Weekly Report, “Lingering Oil-Demand Weakness Will Fade”, dated November 21, 2019, available at ces.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification

Underweight We reiterate our recent downgrade in the S&P semi equipment index to underweight. Our previous mid-year attempt to fight this rally in chip equipment stocks fell short, but thankfully our prudent risk management metric limited our losses. While still early, there is tentative evidence that our underweight stance is starting to bear fruit as the position has moved in our favor. We expect to harvest more gains down the road as a potential trade tension flare up and sustained capex blues leave no room for error in the perfectly priced S&P semi equipment index. Please refer to this Weekly Report for additional details. Bottom Line: We remain underweight the S&P semi equipment index, and maintain the stop loss at the -10% relative return mark. The ticker symbols for the stocks in the index are: BLBG – S5SEEQ – AMAT, LRCX, KLAC.