Teeing Up a Strong Third Quarter

Despite what we see as mounting headwinds on the consumer, this week’s personal income and spending report for June and Q2 GDP solidify our view of Q3 resilience. While the headlines from the financial press touted the slowdown in Q2 GDP growth to 1.5% a.r. from 2.1% in the first quarter, the focus really should have been on robust private domestic demand. To get the best picture of the demand being generated by the private sector, one should not look at overall GDP growth, but a far narrower measure: real final sales to private domestic purchasers. Many Fed officials consider this measure one of the best ‘core demand’ indicators because it strips out several volatile or policy-driven components from the overall GDP figure, including inventory swings, government spending, and net exports, that can sometimes distort the message from the data.

Here, the Q2 data did not disappoint. Real final sales to private domestic purchasers, essentially comprised of real consumer and business investment subtracting out changes in private inventories, came in at a whopping 3.9% annualized rate. This was the strongest growth in private sector demand in more than three years (Chart 1). For their part, consumers bolstered their real spending across all major categories. Durable goods spending was particularly strong, rising at a 6.8% annualized pace, unmatched since before Donald Trump took office in his second term (Chart 2). The bottom line is that despite handwringing around rising prices and concern over the Iran war with commiserate declines in consumer confidence, domestic demand continued to roll along, and indeed gained momentum, in the second quarter.

This tees up a strong third quarter as the business inventory swing works in favor of stronger GDP growth, and the drag from net exports diminishes. We bumped up our forecast for Q3 real GDP growth to 2.2% annualized from 1.9% following this week’s data releases. We do expect a slowdown in real consumer spending growth in Q3 to around 1.8% a.r. from the second quarter’s frantic 3.2% pace. The strong spending in Q2 depleted private savings as real disposable income growth failed to keep pace (-1.5% a.r.) even as consumer spending jumped. The silver lining is that real disposable income growth re-accelerated in May and June, breaking a negative pattern of declines in six of the seven previous months (Chart 3). This should help put a floor under spending and prevent a serious deterioration in Q3. Friday’s measurable upward revision in the final University of Michigan Consumer Sentiment Index for July (to 55.2 from 49.5 in June) points to this renewed momentum from the consumer and puts even more upside risk on our still-conservative Q3 estimate.

What could hold us back? Resilient demand has a downside as it will make the Fed’s job of bringing inflation back to target a much tougher lift. One of the big negatives to come out of the Q2 GDP report was the sharp spike in the GDP Price Index to 6.2% annualized from 3.6% in Q1. This was much worse than the consensus forecast of 4.0%, underlining the risk that inflation will remain far too high and the Fed may need to follow up its tough talk with action in September to tamp down on inflation expectations and keep long-term Treasury yields from jumping even higher. The 10-year yield is up almost 27 basis points over the last 30 days with fed funds futures pricing at 67% chance of a quarter-point hike in September. The AI trade in the equity market has shown some signs of wobbling; if that turns into a broader rout, the remarkable Q2 resilience in domestic demand could rapidly fade.

Fed Policy Confab: 3 Dissenters, 1 Perplexing Presser

The FOMC left policy rates unchanged on July 29, with the fed funds target range at 3.50%-to-3.75% for the fifth consecutive meeting. The vote was 9-to-3 with Cleveland President Hammack, Minneapolis’ Kashkari, and Dallas’ Logan dissenting in favour of a 25 bp rate hike. On July 31, with the communications ‘blackout period’ ended, we started hearing from the dissenters.

Hammack said, with respect to inflation’s now 63-month run above 2%, “I am not confident it will return to our objective on its own. Supply-side factors, including energy prices, have boosted inflation this year, but I see inflationary pressures coming from the demand side of the economy, as well.”

Kashkari is also seeing “a new demand element” to the sticky inflation story, specifically mentioning the AI infrastructure buildout. And with respect to the recent run of supply shocks (the pandemic’s supply chain disruptions and the trio of wars… Ukraine, trade, and Iran), he said, “I increasingly believe that monetary policy does have an important role to play in addressing a series of successive supply shocks that might lead to entrenched higher inflation.”

Logan said, “Even after accounting for productivity gains and temporary supply shocks, inflation appears to be trending toward the mid-2’s, not all the way to 2 percent, and the risks are to the upside.” She added: “Labor, consumption and financial market conditions indicate that monetary policy is not restraining the economy. Without any policy restraint, inflation will likely continue to trend above target until there’s an unanticipated shock.”

If the data and developments that drove this trio to vote for tightening were to continue for the next 6½ weeks, we reckon more voters will jump on the tightening train. It takes at least seven on board (so four of the remaining nine) to switch policy onto a tightening track. The market is currently pricing about two-thirds odds this happens in September and is fully priced in by year-end.

The market was much on Chair Warsh’s mind during the press conference. He highlighted the intermeeting increase in bond yields, both nominal and real, with some along the Treasury curve “among the most significant in the last two decades.” And in terms of tightening financial conditions, “we haven’t done much… The markets have done quite a bit.” Warsh said the latter “has provided us some comfort that we’ve got the ability and capability to deliver.”

We don’t understand what comfort he was taking. Assuming market expectations for economic growth over the long run and liquidity premiums didn’t change much over the preceding seven weeks, the rise in real yields (nominal yields less the market’s inflation expectations) largely reflected increased expectations for average fed funds rates. If Warsh is truly itching to raise rates (like the three dissenters), the market move would indeed be comforting. But if he isn’t burning to boost rates, which we believe is the case, wouldn’t the move be discomforting? Furthermore, a small part of the rise in real yields could be reflecting increased risk premiums, with market participants perceiving more uncertainty around their Fed and inflation expectations. Put another way: while still likely low, the risk of the Fed not ‘delivering’ might have nudged up in the market’s mind, denting Fed credibility.

Crunch Time for Chipmakers?

The sell-off in global technology stocks intensified this week and was led by chipmakers. Korea’s Kospi index was down nearly 40% before bouncing on Friday to close around 28% below its June peak, while the Nasdaq 100 briefly slid into correction territory. Beyond ongoing concerns about valuations, circular funding, AI infrastructure financing (which is now contributing to wider corporate credit spreads), and the sustainability of AI-related spending (Meta shocked investors this week by reporting a large decline in free cash flow due to heavy AI spending), investors are now fretting about a new risk: rising competition from China. This stems from:

• China’s memory-chip maker CXMT more than quintupled in value during its Shanghai trading debut this week. Meanwhile, South Korea’s leading producers, Samsung Electronics and SK Hynix, plunged on concern that CXMT’s production could lead to oversupply and weaker prices.

• China’s state-backed microchip equipment manufacturer Aishengna announced plans to begin mass production of deep-ultraviolet lithography (DUV) machines used in the manufacture of advanced semiconductors. The company will compete with industry-leading ASML of the Netherlands, though questions remain about the reliability and processing speed of the microchips produced with its equipment.

• The growing adoption of China’s low-cost, open-source AI models, which are good enough to meet many consumer and business needs while requiring less computing power to run.

Increased competition from China is adding to investors’ gnawing concern that future AI-related revenues may not justify the massive spending on AI infrastructure. The emergence of lower-cost, yet capable, alternative AI models and chip suppliers could erode the advantages of first movers. The Economist estimates that the largest U.S. technology companies will ultimately need to generate more than $2 trillion in combined revenue per year to earn an adequate return on their AI capital investments, a figure that exceeds their combined revenue today.

Achieving that level of revenue would require a substantial increase in paid AI adoption, yet many consumers and businesses are unwilling to pay for these services. At the same time, some studies suggest that workplace adoption is rising only gradually. For example, the U.S. Census Bureau’s biweekly survey found that AI usage among U.S. businesses hovered between 17% and 20% in the six months to May and was expected to increase by only three percentage points in the next six months. For AI adoption to increase materially, productivity will need to rise as well. However, recent studies and executive surveys suggest the gains have been limited to date.

Time will tell whether the froth is truly coming off the AI capex and equity boom, but recent developments suggest we are at least moving past the mania phase.

Brave New Yields

Bond yields almost everywhere have ratcheted higher this year, driven primarily by higher real yields. Does this mark a normalization or the start of a trend?

Even as the Fed held rates steady this week, there is no mistaking the sustained upward pressure on global borrowing costs. One measure of long-term global interest rates has pushed above 4% in recent days for the first time since the summer of 2008 (Chart 1). Economies as diverse as France, Germany, the U.K., Australia and the U.S. are seeing yields close to, or at, their highest level in years, with Japan recently crossing a 30-year high. What’s remarkable is that yields are swooping higher regardless of whether the local central bank is knee-deep in a tightening campaign (BoJ), just starting out (ECB), or only pondering hikes (the BoE and the Fed). Let’s consider the causes, the implications, and the outlook for the upswing in yields. Much of the focus below is on the U.S., but many of the drivers fully apply elsewhere.

Treasury yields famously broke out of a four-decade downtrend in early 2022, right around the time that the Fed first began to hike interest rates from the pandemic extremes (Chart 2). And, it’s no coincidence that the trend break happened just as inflation was hitting a 40-year high above 9% and core CPI was moving above 6%. But notably, even as inflation backed off from those highs, and the Fed began easing short-term rates, longer-term yields just kept on going. For example, the 30-year yield is above 5.2%, the highest since June 2007—i.e., just before the start of the Global Financial Crisis undercut rates everywhere. While not quite back to the recent highs in 2023, 10-year yields of over 4.7% are also nearing levels not seen since before the GFC.

A key question is: Have yields simply shaken off the dual depressants of the GFC and the pandemic, and are now finally back to a more normal level, or are these the early days of a sustained upward march? The answer will largely depend on where inflation goes over the medium term, and on that front there is one encouraging sign. The knee-jerk explanation of this year’s rise in yields is to attribute it to concerns about higher inflation, arising from the conflict with Iran and costlier energy. Yet, that’s not really the obvious driver, according to market pricing. The implied breakeven inflation rate from 10-year yields at around 2.3% is actually a touch lower than in late February, and is right in line with its two-decade average (Chart 3). The same goes for even the shorter-term five-year rates. In other words, the market is quite confident that the latest spike in energy costs will not spill over into other prices and/or more long-lasting inflation.

In turn, that market confidence is driven by the view that the Fed will respond to the current inflation bump with higher short-term rates, which is partly expressed in the sustained rise in real yields—a move Chair Warsh applauded this week. For all the focus on inflation concerns, it’s been higher real yields that have been the big mover of nominal yields in the past year. For example, the 10-year real yield is approaching 2.5%, a level it only surpassed once since 2008, and up from negative readings as recently as 2022 (Chart 4). In a similar vein, the 30-year real yield is fast approaching 3%, a level that hasn’t been seen since 2002 (aside from a spike in the fall of 2008)—again, a dramatic shift from the negative real rates of barely four years ago. Beyond a reversion to norms for the term premium and shifting expectations on Fed policy, what else could explain this bounce in real rates? We consider three possible drivers:

1) Fiscal concerns. It’s always difficult to draw a statistical link between government debt & deficits and long-term yields. The main reason that the correlation is weak is that government borrowing is just one of an array of factors affecting demand and supply for fixed-income, and deficits can be ballooned by a weak economy (which at the same time undercuts yields). Moreover, the old saw is that government deficits don’t matter, until they do—in other words, it’s not a nice, neat continuum of debt driving yields, it’s more of a sudden, step-function, where markets will grow concerned at a certain threshold level of debt. Finally, government finances in one economy likely will not act in isolation on yields, as borrowing authorities have access to capital from many diverse sources around the world. However, when government debt globally is moving in one direction in near unison, that’s when it can become a broader issue for yields everywhere (Chart 5).

2) Quantitative Tightening. The entire point of quantitative easing (QE) was to add an extra layer of monetary easing, by helping drive down long-term interest rates. So, now that these policies are slowly going into reverse, it’s not exactly controversial to suggest that QT is subtly driving yields higher. What is controversial is the degree to which QT is playing a role in the back-up. Looking at the smoothed change in securities on the Fed’s balance sheet (inverted in Chart 6), there does appear to be at least a loose relationship with yields, but it’s not a good leading indicator. Not unlike government borrowing trends, it’s but one factor driving the demand-supply balance. However, it is fair to conclude that QT has at least played a role in the normalization of real yields in the past three years.

3) AI impacts. A more recent factor driving real rates higher is a stronger growth outlook arising from AI—both the wave of spending on the build-out (as well as the associated borrowing), and the potential for a productivity boom. While Warsh stated earlier this year that AI productivity gains could restrain inflation and help lower rates, other Fed officials lean against that view. From AI itself: “Governor Barr noted that the massive capital expenditure and productivity surge tied to AI could instead push up the economy’s equilibrium interest rate.” And, “Officials including Governor Cook and Minneapolis Fed President Kashkari point out that surging demand for AI infrastructure—such as data centers straining the power grid and driving up electricity and chip costs—creates immediate inflationary pressures.” We tend to believe that a productivity boom can open the door to higher growth without inflation pressures, but R-star would still be higher in that world (Barr’s point). Finally, while it appears that productivity has been trending higher, it’s far from breaking out (Chart 7).

Bottom Line: Taking these factors together, there are plenty of reasons to believe that the upswing in long-term yields won’t be reversed soon. The term premium has normalized, government debt ratios are still rising, while QT and the growth boost from AI will keep the pressure on yields for some time yet. However, as long as underlying inflation is broadly stable—and the energy price spike doesn’t spread—we suspect that this is close to the peak in yields. To answer the earlier question, the recent move in yields appears to be more of a full return to “normal” after the twin shocks of the GFC and Covid, and less of a start of a lengthy upward trend.

Finally, we’ll readily admit that pinning down the neutral interest rate is incredibly difficult. Chicago Fed President Goolsbee joked that R-star should be called “R-Sasquatch”. But note that the average spread between the 10-year Treasury yields and overnight Fed funds, going back 60 years, has been almost precisely 100 bps—the current spread is very close to 100 bps. And, 10-year Treasuries are still closely tracking the 10-year growth rate in nominal GDP, as theory would suggest (Chart 8). Together, these very basic indicators drive home the point that current yields are “about right”.