Monthly webcast

Our monthly economic webcast for May has now been published, entitled “Soft landing remains in focus, despite stubborn inflation.”

U.S. economy slows

The predominant concern about the U.S. macro-environment for most of 2024 has been that it is generating too much inflation. This has been enabled, at least in part, by too much economic growth.

Those concerns remain valid today, though they are de-intensifying a little. There is reason to think that the recent U.S. inflation surge could be in the realm of peaking. Real-time indicators point to a slight deceleration (see next chart).

Theoretically, shelter inflation should ebb somewhat further in the coming months given the lags built into the Consumer Price Index (CPI) calculation. Similarly, we believe insurance inflation may be close to peaking – discussed later.

U.S. Daily PriceStats Inflation Index points to slight easing

PriceStats Inflation Index as of 05/03/2024. CPI as of March 2024. Sources: State Street Global Markets Research, RBC GAM

On the growth side, it would appear that the U.S. economy is now decelerating. This is reducing the extent of the remarkable U.S. economic outperformance versus the rest of the developed world, especially as other countries have been reporting better-than-expected growth recently. Reflecting this, global economic surprises are now positive, whereas U.S. economic surprises have just tumbled into outright negative territory. Indeed, they are at their most negative level since early 2023 (see next chart).

Global economic surprises reverse course

As of 05/03/2024. Sources: Citigroup, Bloomberg, RBC GAM

Among the recent economic misses, U.S. first-quarter gross domestic product (GDP) rose by just 1.6% annualized. This is a far cry from the outsized 3% and 5% annualized increases in the prior two quarters -- though not as meagre as it first looks given a temporary drag from inventories and the silver lining associated with stronger imports.

The twin Institute for Supply Management (ISM) numbers for April both disappointed expectations by signaling a slight contraction in activity. This is hardly unfamiliar ground for the goods-oriented measure, as the prior month was the first release depicting an expansion in 17 months. Indeed, one can still argue that the ISM Manufacturing Index is rising on a trend basis. However, the ISM Services Index is more clearly ebbing on a trend basis. The contraction signal was the first since December 2022. The previously ebullient service sector is apparently slowing, though we are dubious the biggest sector of the U.S. economy is truly contracting. There are more optimistic signals from other metrics including the U.S. Beige Book and corporate earnings reports.

U.S. payrolls also missed expectations, arriving at +175,000 new positions, versus an expected 240,000. That’s hardly a bad number in absolute terms. But it was nevertheless the slowest pace of job creation in six months. It’s also probably slightly less than is needed on a steady state basis to keep pace with the high level of undocumented immigration. Proving this, the unemployment rate inched higher from 3.8% to 3.9%.

But rather than shudder at these economic misses, the stock market is actually responding with enthusiasm. Given that most measures of the output gap depict a U.S. economy that is running a bit too hot (see next chart), what’s best for the sustained health of the economy isn’t robust economic growth, but rather modest to moderate growth. And that is what was delivered with the GDP and employment numbers.

U.S. economy is running above full capacity

Congressional Budget Office (CBO), GAM model 1 and 2 estimates as of Q1 2024, International Monetary Fund (IMF) estimates as of April 2024, Organisation for Economic Co-operation and Development (OECD) estimates as of Nov 2023. GAM model 1 estimated using CBO natural rate of unemployment; GAM model 2 estimated using HP filter trends. Shaded area represents recession. Sources: Macrobond, RBC GAM

Perhaps the only complication in this is that it can be hard to distinguish a well-executed soft landing from the beginning of a hard landing since both involve a decelerating economy. As such, it would be nice to see further confirmation of modest to moderate growth in the months to come, so as to be able to rule out the possibility that this is instead the beginning of a more protracted trend downward.

What is a sustainable unemployment rate?

There was a time before the global financial crisis when it had appeared that the U.S. could possibly sustain an unemployment rate of as little as 3.5% to 4.0% without incurring elevated inflation. This was unusually low by historical standards. A “normal” unemployment rate in the 1990s and 2000s was no better than about 4.5% to 5.0%.  In the 1970s and 1980s, normal was arguably 5.25% to 5.75%.

What is a sustainable unemployment rate today? It is probably higher than before the global financial crisis:

• Inflation is still too high with the current 3.9% unemployment rate, so the “normal” level is probably above the current reading.

• Most measures of economic slack (refer back to the prior chart) argue that the U.S. economy is overheating alongside a 3.9% unemployment rate. Thus, a higher unemployment rate is theoretically needed to equilibrate the economy on this basis as well.

• The unemployment rate consistent with normal inflation during the 2010s may have been unnaturally low due to deflationary forces such as bank deleveraging and the last gasps of globalization. As those forces faded, they have been replaced by structural inflationary pressures including de-globalization, climate change and scarring effects from the recent inflation shock. Dealing with these requires slightly higher interest rates than otherwise, arguing for a slightly higher unemployment rate as well.

• A slightly higher average inflation rate on a structural basis may induce a slightly higher normal unemployment rate via other channels as well. When inflation is low, it is close enough to zero to be ignored by workers, resulting in wage growth that is slightly lower than otherwise, enabling lower unemployment. But workers have now been awakened to inflation given its recent spike, with the opposite effect on wages and unemployment.

• Conversely, the rise in Internet job listings allows for more efficient job matching between companies and workers, reducing frictional unemployment. That said, this is not a new force. The downward effect on unemployment is likely already significantly incorporated into what we think of as normal unemployment.

• In theory, artificial intelligence could increase structural unemployment as computers supplant workers. However, this has not historically happened at the economy-wide level when new general-purpose technologies arise, and this is not (yet) factored into our forecasts.

Taking into account all of this, we believe it is reasonable to think that a “normal” U.S. unemployment rate in the future will be between 4.0% and 4.5%. The implication is that the unemployment rate can continue to rise modestly further without imperiling the economy, and even in a useful way to further tame inflation.

The Canadian “normal” unemployment rate is probably higher, at 5.75% to 6.25%. This is due to a mix of definitional differences, a different industry mix, and more generous unemployment insurance. Canada’s 6.1% unemployment rate is already within this range.

What’s driving insurance inflation?

While the biggest driver of U.S. inflation is shelter costs, other service-sector inflation pressures also exist. In some cases, these are actively accelerating and causing concern. The biggest example among these is auto insurance (+22.2% YoY), with home insurance costs also rising significantly (4.6% YoY). Insurance costs are only about 3% of the CPI basket, but they are nevertheless responsible for about a third of the current CPI overshoot of the 2.0% target.

What explains the acceleration? There are several forces:

• Homes and cars have become substantially more expensive over the past several years, raising the cost of insurance.

• Cars have become more costly to repair given the inclusion of ever-more electronics, hybrid vehicles with two engines, rising mechanic wages and some lingering parts shortages.

• There has been a surge in car thefts, which insurance companies must compensate customers for.

• There were an unusual number of natural disasters in 2023, presumably at least in part due to climate change.

• As a result of the prior factors, insurance companies have been paying out $1.10 in claims and expenses for every $1.00 in premiums they collect. This certainly incents them to raise prices substantially.

• Insurance pricing is regulated at the state level, but insurance companies have been actively pulling out of some unprofitable states and threatening to exit others, forcing regulators to allow large price increases.

Will things keep getting worse from here? It’s not impossible, but more likely is that insurance inflation is nearing a peak. State governments are unlikely to allow another major round of price increases in such short order. It can be argued that 2023 was an unusually bad year for natural disasters in the U.S., even in the context of climate change. Subsequent years shouldn’t be quite as bad. Car thefts are still elevated, but now appear to be in decline.

Prior premium increases are still appearing in each monthly CPI print given that contracts renew over a rolling 12-month window. This should begin to fade with time. In fact, historically, auto insurance CPI lags overall CPI by an average of 16 months. This is to say, auto insurance CPI usually spikes after CPI has spiked, and also comes down after (see next chart). It would be in keeping with history if auto insurance CPI starts to decelerate quite soon. It is already a bit behind the usual schedule.

Motor vehicle insurance prices lag inflation

As of March 2024. Sources: U.S. Bureau of Labor Statistics (BLS), Macrobond, RBC GAM

Sifting through the latest buzz words

A few economic buzz words have been used recently to describe the current economic situation, and we ultimately reject both of them.

Some analysts are describing the economic environment as one of ’stagflation.’ The term refers to the combination of economic stagnation and too much inflation. This would be the worst of both worlds.

But the word doesn’t really describe the U.S. economic environment right now. The inflation part is arguably correct (though one might debate whether sub-5% inflation truly suffices given that the term was popularized in the 1970s when inflation was persistently higher than today). But the U.S. economy is still growing materially, even if it is decelerating. That can’t be called stagnation.

It is tempting to use the term stagflation to describe some other developed countries that experienced similar inflation to the U.S. while delivering virtually no economic growth in 2023. But even there, it isn’t quite right. A secondary requirement of stagflation is that there must be persistence: in that context, a few quarters of stagnating output just isn’t enough to be a period of economic stagnation – and growth has since begun to pick up.

Another term thrown around is the idea that the U.S. economy might be on its way to a ‘no landing’ outcome, a wrinkle on the usual debate between a ‘soft landing’ and a ‘hard landing’. A ‘no landing’ would involve the economy continuing to power forward, versus a ‘soft landing’ that involves some deceleration in growth without succumbing to a recession.

But it isn’t clear that a ‘no landing’ is really an option. Yes, growth could be strong for a period of time. But further strong growth, barring a positive productivity shock, would mean that the U.S. economy overheats even more. The overheating would need to be eventually undone, and it is unlikely that inflation could normalize in that environment. In turn, rates would have to go higher, and the economy would slow later.

As such, you can argue that a ‘no landing’ scenario is really just a delayed soft or hard landing. For our part, we don’t see evidence of a ‘no landing’ taking hold given the recent economic deceleration in the U.S. We still think a ‘soft landing’ is more likely than a ‘hard landing’.

Tracking the business cycle

The RBC GAM quarterly U.S. business cycle scorecard has just been updated. It is unusually scattered in its assessment, with significant votes cast for practically every possible phase of the cycle (see next chart).

Business cycle is most likely at ‘mid’ or ‘late cycle’

As of 05/03/2024. Calculated via scorecard technique by RBC GAM. Source: RBC GAM

But we can still make a few useful observations.

Arguments that the cycle is ending have weakened since last quarter. ‘End of cycle’ and recession signals have both diminished. This is consistent with our view that the risk of recession has fallen somewhat and is no longer the most likely outcome.

There remains a subset of indicators that believe a new cycle is beginning (‘start of cycle’). But after enjoying earlier momentum, this group failed to rally further support over the past quarter. So ‘start of cycle’ remains a less plausible interpretation than the final group on the chart (‘mid and late cycle’).

This last group enjoys both the largest increase in vote share and also the most votes. ‘Mid cycle’ and ‘late cycle’ have the best claim to capturing the economy’s current position within the business cycle. This argues for a few years of further growth, but not an endless runway.

This is actually a fairly exciting (if tentative) conclusion. It aligns well with our view that if a recession has indeed been avoided, this is not the start of a new cycle but instead the continuation of the prior cycle. See our discussion on the subject in an earlier #MacroMemo.

Gauging an economy’s positioning within the business cycle is actually quite an important step in assessing the stock market outlook. The stock market historically does best in the first few phases of the cycle and worst in the final few phases (see next chart). If this is indeed a ‘mid cycle’ or ’late cycle’ moment, then the expected equity returns are positive, but only moderate.

Annualized S&P 500 returns rise and fall over the course of the business cycle

As of 03/12/2021. Shaded area represents range. Based on data from 1949 business cycle onwards. Sources: Macrobond, RBC GAM

How have higher interest rates affected the economy?

High interest rates continue to inflict pain, and by some measures the pain is still mounting.

U.S. consumer loan delinquencies continue to rise at a brisk pace (see next chart). The credit card default rate is now twice as high as it was at its 2021 trough. In fact, it is at its highest reading since 2011.

There has also been a large jump in auto loan delinquencies. The increase in mortgage delinquencies has been more muted. These trends haven’t yet interfered problematically with consumer spending, but the threat exists.

U.S. consumer loan delinquency is now rising

As of Q4 2023. Sources: Federal Reserve Bank of New York, Macrobond, RBC GAM

Meanwhile, on the corporate side, the default rate for U.S. high-yield bonds also continues to rise (see next chart). Of course, it has merely increased from a very low level to a more historically normal level at this point, and the rate of increase appears to be slowing. But the line is still rising.

U.S. high-yield default rates have returned to pre-pandemic levels

As of March 2024. Last-12-month default rates of U.S. high yield corporate bonds. Sources: Bank of America, RBC GAM

However, this tale of woe isn’t the entire story. The property market is highly rate-sensitive, yet U.S. home prices appear to be rebounding (see next chart). Similarly, the beleaguered commercial property sector appears to be stabilizing on a price basis after a sharp decline (see subsequent chart). All of this is to say that some aspects of the economy appear to be turning a corner after having adjusted to the new interest rate environment.

U.S. home prices are turning higher

As of February 2024. Shaded area represents recession. Sources: S&P Global, Federal Reserve Bank of St. Louis, Macrobond, RBC GAM

U.S. commercial property prices are stabilizing

As of April 2024. Shaded area represents recession. Sources: Green Street Advisors, Macrobond, RBC GAM

Shifting expectations as Fed recalibrates

The U.S. Federal Reserve (the Fed) left its policy rate unchanged in early May. A superficial reading would have left a hawkish impression: inflation was described as not cooperating and the accompanying press conference pointed away from a first rate cut in June or July.

But financial markets already knew about the upside CPI surprises. As such, they had already priced out all but a single 25 basis point rate cut for 2024. That was down from the more than six cuts that were priced at the start of the year.

Clearly, markets were looking for new information. The key new information from the meeting was as follows:

1. Despite the inflation miss, the Fed believes the labour market is loosening – a requirement for rate cuts.

2. The Fed ruled out rate hikes, which was a relief for markets that had begun to grow nervous about the possibility.

3. The Fed indicated that the timing of the upcoming presidential election is of no consequence to its actions. As a result, whereas the market had been nervous about pricing in a first rate cut in September, with the view that it could be interpreted as political interference, that date has returned to play.

In turn, markets now price a September rate cut, and the possibility of another cut in December. That seems about right based on what we currently know about the decelerating economy, the prospect of a return to decelerating inflation and the signals the Fed is sending.

Putting wage growth in context

The tight U.S. labour market is commonly described as being a reason inflation is struggling to decline. The service sector has been particularly affected given its labour-heavy expense base.

But the situation is more nuanced than it first looks. Yes, U.S. nominal wage growth of approximately 4% year over year (YoY) is running hotter than inflation. But you can’t stop there.

It is also necessary to factor in productivity growth. If a worker is being paid more in inflation-adjusted money but is producing more as well, that sounds like a fair trade. Real unit labour costs adjust wages not just for inflation but also for rising productivity. As it turns out, real unit labour costs in the U.S. have actually been falling over the past few years, arguing that rapid wage growth is seemingly not a central driver of inflation (see next chart).

U.S. productivity-adjusted labour cost has been declining

As of Q1 2024. Shaded area represents recession. Sources: U.S. Bureau of Labor Services, Federal Reserve Bank of St. Louis, Macrobond, RBC GAM

But, inevitably, it isn’t quite as simple as that – for several reasons.

1. Productivity gains tend to come primarily from the development of new technologies and from a rising capital stock rather than from an increase in the quality of labour. As such, one can argue that the gains from these productivity advances should primarily accrue to the businesses that invested in the new technologies and capital, rather than be captured by workers in the form of higher wages. In turn, it isn’t fair to subtract the entirety of productivity growth when assessing the interplay of wages and inflation.

2.  Keep in mind that the goal is to get inflation lower. Wage growth may be reasonable in the context of current elevated inflation levels, but they are high if the goal is to push inflation downward. It certainly isn’t the sole responsibility of wages to do this. But we need nominal wages and corporate pricing decisions to be trending downward to facilitate this.

3.  A tight labour market is about more than just wage growth. When unemployment is low, it increases consumer confidence, encourages more spending growth, allows companies to raise prices more, and so on. A looser labour market is still a necessary condition for lower inflation, whether operating through the wage channel or other channels.

The main point is still the same: wages may not be driving inflation as much as commonly imagined. But the labour market probably still needs to ease somewhat to get inflation back to normal – alongside other changes elsewhere in the economy.

Quantifying geopolitical risks

There is a great deal of hand-waving that occurs when geopolitical risks are evoked. This slippery topic is notoriously hard to define or quantify.

To be sure, it seems entirely reasonable to say that geopolitical risks are presently high given the grinding war between Russia and Ukraine, turmoil in the Middle East, frictions between China and the U.S., and the upcoming U.S. election (with consequences for international affairs). But there are always tensions and conflicts in the world. Is there truly more than the usual amount of geopolitical risk in the world today?

Helpfully, there exists a new Geopolitical Risk Index that automatically trawls major newspapers for key geopolitical terms, quantifying the level of risk on that basis. And what do you know – yes, there does appear to be substantially more risk than usual (see next chart). It is an admittedly skittish chart, but even setting the choppiness aside, the level of geopolitical risk appears to have been reliably higher than normal over the past few years and is actively trending higher.

Geopolitical Risk Index is rising on concerns in the Middle East

The Geopolitical Risk Index constructed by Caldara and Iacoviello measures adverse geopolitical events and associated risks based on automated text search results of several newspapers covering geopolitical tensions. Sources: Caldara and Iacoviello (2022), RBC GAM

U.S. election nears, but becomes less clear

The U.S. election is now less than six months away and the two candidates – Biden and Trump – are all but locked into place.

But there still isn’t much clarity as to who will win (see next chart). The polls show a dead heat, an academic betting market shows Biden substantially ahead, and conventional betting markets show Trump slightly ahead.

2024 U.S. presidential election likely to repeat Biden vs. Trump

RealClearPolitics as of 04/23/2024, poll averages for Biden vs. Trump matchups only. Others acknowledge possibility of other candidates contesting the election. Predictit as of 04/25/2024, probability of winning derived from prediction markets data. oddschecker as of 04/25/2024, probability of winning derived from median daily betting odds. Sources: oddschecker, Predictit, RrealClearPolitics, Macrobond, RBC GAM

If anything, the race is tightening as Trump had appeared to be leading just a few months ago according to most metrics.

What’s the focus of China’s third plenum?

China’s third plenum was mysteriously delayed last fall and has now been scheduled for July. The policy summit occurs once every five years. It is traditionally a time when the country announces major new political and economic policy directions. The delay may have been motivated by a desire to implement significant changes, though that is speculation.

For context, the third plenum in 1978 marked China’s turn toward capitalism and markets. The third plenum in 2018 set the path for President Xi to govern indefinitely.

The coming third plenum is reported to focus on deepening reforms and promoting the modernization of China. A speech by President Xi in March spoke of plans to continue opening up the country. Such vague language still leaves a lot to the imagination.

Given the recent orientation of Chinese public policy, one might imagine further efforts to spur consumer demand (in contrast to China’s usual supply-side stimulus), more support for the beleaguered housing market, and efforts to encourage the country’s underperforming stock market.

Catching up on productivity and technology developments

There have been a number of recent developments in the productivity and technology space.

Noncompete clauses banned

First, the U.S. Federal Trade Commission recently banned noncompete clauses in employee contracts. This is theoretically quite a positive thing for economy-wide productivity, as it allows knowledge workers to more easily change employers or start their own business in the same industry as their prior employer. This will likely encourage innovation and help to diffuse knowledge more efficiently between firms.

California has long had a similar regulation in place, and this is often cited as one of the reasons why Silicon Valley became the world’s technology hub.

But it is not clear that the new regulations will survive anticipated legal challenges, as the Federal Trade Commission may have overstepped its bounds. Incumbent businesses do not like the change as it allows their own workers to more easily depart with valuable corporate knowledge. In short, this is a change that should incrementally increase overall productivity growth while incrementally hurting large existing firms.

Artificial Intelligence (AI) expected roll out by sector

We’ve repeatedly expressed our excitement about the prospect of generative AI helping to drive productivity growth over the next few decades. It should first show up in the form of additional research & development and capital expenditures, and then later in the form of faster productivity growth.

It is also worth reviewing the expected sequencing by sector in terms of whose profits go up when.

• One would expect profits to first rise among the companies facilitating an investment in the underpinnings of AI – computer chips and data centres, primarily.

• Then, the profits should rise among the companies producing the new AI technologies as they successfully monetize them.

• Finally – and this is the most important part that often gets overlooked – the majority of companies spanning all sectors should be in a position to implement AI solutions within their businesses, achieving productivity gains and higher profits.

Computers level up

Finally, it is amazing to see how computers and artificial intelligence have rapidly acquired skills previously the sole domain of humans (see next chart). A recent International Monetary Fund (IMF) analysis plots the abilities of AI versus humans in handwriting recognition, speech recognition, image recognition, language understanding and the Graduate Record Examination (GRE) math test.  In all cases, AI has caught up to and surpassed humans over the past decade.

AI performance on human tasks rising

Figure is based on a number of tests in which human and AI performance were evaluated in five different domains, from handwriting recognition to language understanding. For the GRE mathematics test, the human benchmark is set at the paper on GPTs. AI = artificial intelligence; GPT = generative pretrained transformer; GRE = Graduate Record Examination. Sources: Kiela and others 2021; OpenAI; IMF staff calculations, RBC GAM

The set of skills evaluated by the IMF is obviously not complete. AI still cannot drive a car flawlessly, let alone in adverse conditions or without extensive advanced mapping. AI is similarly still lacking in creativity, common sense, emotional intelligence and adaptability. But progress continues to be made, and generative artificial intelligence in particular is already making inroads into some of these previously elusive qualities.

-With contributions from Vivien Lee, Vanita Maharaj and Aaron Ma