In the week just passed, the bond market had a modest correction that, unlike its bearish predecessor, strikes me as being as much about data and the Fed as oversold technicals and sentiment. In other words, while the weakness of the prior couple of weeks didn't come with any especially strong data per se, the gains in the last few days had more sustainable, fundamental elements culminating in the CPI and retail sales releases after the correction was already under way.
True, Yellen's appearance before Congress was a bit more dovish than other recent Fedspeak—a bit more careful on confidence that inflation would rise—but then if yields rose in the wake of 1.4 percent year-over-year PCE inflation figures and a tame gain in average hourly earnings, I don't think Yellen told us anything we didn't already have on hand. December Fed fund futures are at 1.22+ percent, shaving the odds of a December hike from 53 percent after NFP to 43 percent now.
And the curve flattened, with the belly outperforming. The oversold technicals preceded Yellen's testimony, the Treasury auctions, the headlines about Trump Jr. meeting with the Russians, and the bulk of the week's most important economic data. If my decades observing market behavior have taught me anything, it's that prices often change more than the facts. However, we need more justification than that —saying something hit resistance simply doesn't wash with most investment committees. Friday's data offered that justification. Not that a slightly more dovish, or circumspect, Yellen necessarily works with a flatter curve, but never let facts interfere with theory.
Speaking of which, I was asked last week by a portfolio manager where 10-year Treasury yields should be over the long haul. Good question, impossible to answer. However, in the course of the discussion I said that the closest proxy likely would be nominal GDP, currently running around 3.4 percent. The mean spread is 35 basis points since 2000, or simply nominal GDP has been roughly 35 basis points more than the 10-year yields over that period. Narrow that to the last seven years and the nominal GDP has been more like 140 basis points more, on average, than the 10-year notes. And in the '90s, 10-year Treasuries averaged 95 basis points more than nominal GDP.
Aren't statistics wonderful? I could manipulate this sort of stuff to come up with any sort of conclusion. Suffice it to say, what the historical "norms" show has clearly evolved over time, and the spread to find a standard is dependent on the economic cycle. Thus, to come up with a yield requires assumptions wrapped in contingencies surrounded by what Churchill might have called enigmas. The narrow spread of today seems rather too narrow and either presages weakening in nominal GDP or a rise in yields. I like the mean idea of 10-year notes trading 140 basis points under nominal since 2010, but would tweak that to consider the depths of the last recession. The mean since 2000 is 35 bps under. Given that nominal GDP has eased over the last few decades, perhaps the new normal of this would be 50 to 75 basis points going forward.
CHARTS AND THEMATICS: I must say that I've gotten a lot of use out of my demographic piece from a few months ago, and borrowed another piece from that for this particular section. What I'm looking at is who, demographically speaking, has been getting jobs. The answer is largely baby boomers, the youngest of whom was born in 1964 and the oldest in 1946. This is, of course, the generation that should be retiring if they had the money to afford it and interest in things other than work. I wonder which of those weighs in most heavily?
In the latest Beige Book, Atlanta mentioned that some firms were “seeking out retirees” to return to work. Out of St. Louis, high-tech firms noted that rising benefit costs had limited wage increases. Interesting stuff, both of which might help understand the graph below, which looks at the employment gains by the age 55+ cohort. It's one of my favorite charts because the narrative is so rich. To wit, the bulk of job gain since the trough of recession has been to this age group. Depending on when you start counting from, between 56 percent and 68 percent of the jobs have gone to this group.
There are those who will rightly point out that since this group is becoming a larger percentage of the overall population it's inevitable that they would be getting a lot of jobs. But the bulk? I don't think so. The motivation is rather cleaner in that they lost so much in home values and retirement money that they need to work, which assumes they had retirement savings to start with. That's a leap; I've relayed that something like 50 percent of Americans over 50 have less than $50,000 in retirement savings. Too, this baby boomer generation has more debt than similar retiring populations in recent history and, of course, at once have the benefit of rising longevity along with the need to pay for that as well as, in many cases, aging parents.
Too, anecdotally, we know this group isn't as concerned with income as it is with health benefits and flexible schedules. This is, of course, a rather broad generalization—is there any other type?—but certainly fits the narrative. It also brings up this next chart, related to the most recent NFP report and inspired by something that David Rosenberg wrote in its wake.
Here you'll see a set of colorful lines that describe the following, according to BLS's Current Population Survey: (1) 16- to 19-year-olds got 203,000 jobs, presumably pure summer work; (2) 55+ got 178,000 jobs; (3) 20- to 24-year-olds lost 178,000 jobs, maybe grads taking the summer off; and (4) 45- to 54-year-olds lost 178,000 jobs. Self-employment jobs rose a whopping 188,000, which if taken in context of the overall 244,000-job gain according to this series means that the bulk of job gains were in that category. Let's be honest—this is not how most people look at the jobs story. The NFP data as released portray something rather different, and the difference between the household data and the CPS data confuses me. The volatile, choppy nature of the charts also suggests we're not really seeing a trend. Still, they collect and report these stats for a reason, and they portray a less robust mien to the June employment figures.
And if you want a real "shocker" consider that employment for men between 20 to 54 years old has been slipping. You can see this in the participation rates. Well, Yellen suggested the opioid epidemic might have a hand in this. From her testimony: “I don't know if it is causal or a symptom of long-running economic maladies that have affected these communities.”
IN OTHER NEWS: Jamie Dimon, a fellow Tufts graduate, I might add, and also chair of JPMorgan Chase, spoke about the end of QE and the uncertainties that come with it: “We've never had QE like this before. We've never had unwinding like this before.” Obviously, that should say something to you about the risk that might mean, because we've never lived with it before…could be a little more disruptive than people think.
In the paragraph that concluded the “Current Economic Situation and Outlook” section of Yellen's prepared testimony before Congress, the word "uncertainty" appeared three times and was repeated in terms of the generic sense of an economic outlook. In the spirit of splitting hairs, in her last testimony she only used the word "uncertainty" twice. The uncertainty in reference to fiscal policy was retained. Added, however, was uncertainty regarding “when—and how much—inflation will respond to resource utilization.” I would say this is a bit more circumspect in her inflation confidence.
Getting back to Jamie Dimon, his views about pulling back on QE both here and abroad are particularly astute because we simply don't know very much. The obvious risk is that as the market absorbs more of what the Fed and other central banks don't buy, rates need to accommodate the supply. A flip side to that is, What happens to spreads on the stuff the Fed didn't buy, i.e., corporates? Do those widen to accommodate the filling void of fixed-income product? And if rates do rise, does that stem the buybacks that have supported equities and enhanced P/E ratios?
Mike MacKenzie of the FT put together a chart for a piece called “Charts That Matter: The Importance of Central Banks for Equity Prices.” I've replicated the chart below, which I think helps raise the cautionary note that when balance sheets go the other way there is risk, and at the very least the uncertainty Dimon references, in the process, which is why I think the Fed and others will proceed with more caution than even their gradual demeanor has opined.
Of course, this last week was filled with the news that Donald Trump Jr. met with Russian officials who claimed to have dirt on Hillary Clinton—“I love it”—after he and just about everyone else affiliated with the administration had so vehemently and colorfully denied such a meeting ever took place. Also, there were son-in-law Jared Kushner and then-campaign-head Paul Manafort. With apologies to the Trump Jr. (et al.) apologists, whether the meeting was innocent or wrong, based on political inexperience or something more nefarious, the repeated denials—and vehemence of those denials—doesn't make anyone involved look good.
Whether this was one of the sources of the bond market's recovery can be debated. It did, however, inspire me to look back to how stocks performed into and out of the Watergate era and overlay that experience with where stocks are today. Frankly, I wouldn't make too much of this chart; I could manipulate it in many ways to create any story I want. Still, what I've done is lead the current chart of the S&P 500 to match up with 1971–76 to illustrate at least some of the influence of political scandals. Bear in mind, other things were going on, like the end of the gold standard, the Arab oil embargo, the fall of Saigon, the Arthur Burns Fed (price controls, foot on, foot off)...and let's not forget the "Whip Inflation Now" campaign and "WIN" buttons (which can be had for as little as one cent on eBay).
NEAR-TERM THOUGHTS: The Phillips curve has been getting undue attention of late. This is, of course, a model that describes the inverse relationship between the unemployment rate and inflation. When the unemployment rate falls, inflation rises and vice versa. The Fed has brought this up a lot, recently saying they expect the low unemployment rate to precede wage gains and lift inflation to their target. It's one of the justifications, perhaps the most important one, for their hiking so far.
Thing is, it doesn't really work so well in its most simplistic form, and that's what I portray. The academic arguments, and there are many, are rather more arcane than I'd like to discuss (or perhaps you'd care to read about), but do a search and you'll find them. One of the ones that makes the most sense to me is put forth by Lacy Hunt of Hoisington Investment Management. That's an FYI. In any event, take a look at this chart, which shows UNR vs. YoY CPI. I've provided dark arrows to show when UNR was falling and inflation rising, and lighter ones that show UNR falling and CPI well behaved if not sliding as well. I don't think this needs more explanation.
As for my near-term market views, my first line is a technical one. Again, having said that, the selloff that took 10-year notes close to 2.40 percent was about bearish momentum and sentiment; the action last week reversed those various measures even before we got Friday's CPI and retail sales. In context, NFP day was a bearish outside day that then saw some of a quadruple top (not a real technical signal) just shy of 2.40 percent and momentum measures starting to roll over. July 11 started the more dovish Fedspeak and was an outside day down in yields—an important signal effectively neutralizing NFP day's action. Then, even with semisoft auctions, the correction continued, and with the critical data Friday, momentum made a decisive bullish cross.
In short, I'm in a bullish mode. The target in 10-year yields is the zone of the 21-, 40- and 200-day moving averages from 2.23 percent to 2.26 percent, but I don't think that will prove much of a challenge. 2.20 percent is a bigger level in terms of resistance, and then toward the 2.10 percent area. Given the oversold stochastics and MACD, that seems doable, but frankly I'm more in a range frame of mind and wouldn't chase it for more than a trade beyond 2.20 percent.
Support is the NFP day close at 2.36+ percent, which is also the close from the July 11 bullish outside day and the subsequent highs. The fact that in the week ahead data is decidedly second tier gives more impetus to the technical argument. The 2-year/10-year curve is in flattening mode (see chart at end), and I see a move to the 21- and 40-day MAs near 89 basis points. Like the belly vs. the wings and especially like 5-years vs. 2-years/10-years and 2-years/30-years:
EPFR INSIGHTS: I'm taking a different spin on looking at these EFPR flows by making the distinction between ETFs and mutual funds flows into government bond categories. Again, what's before you in this section is my education into the space as I try to determine the directional significance and market implications of flows beyond the simple "If they buy, prices go up."
By separating ETFs from mutual funds, I ponder if I can see a difference between long-term investors and more-active trading accounts. I don't yet have a conclusion but invite discussion/observations if you have them.
Starting with short-term U.S. government funds, while there have been weeks of divergence, for the most part it appears that ETFs and mutual funds run in the same direction flow-wise. When there are outlier weeks, ETFs are clearly the ones making the big moves. For context, the last week saw the second-largest ETF buying flow since the start of 2016. What a difference a yield makes!
Intermediate-term funds show more diversity and divergences. I suppose this makes sense, as ETFs may serve more as a trading or hedge vehicle. In the latest week, for example, ETFs had about $68 million in negative flows—the last two weeks have seen a reversal of what had been pretty steady buying all year. In contrast, mutual funds saw a $271 million bout of selling.
ETF flows show buying in long-term government funds against modest selling in mutual funds. Still, the net flow for ETFs over the last several weeks has been positive and simply slowed while mutual fund flows in the sector are essentially flat.
David Ader is chief macro strategist for Informa Financial Intelligence.
