The everything rally

Markets have been ebullient in recent weeks. It begins with the bond market, where the benchmark U.S. 10-year yield has rallied all the way from an intraday high of 5.00% on October 19 to 4.26% now – a 74 basis point decline in less than two months (see next chart).

U.S. yields soared, but now reversing as rate cut speculation mounts

As of 12/08/2023. Shaded area represents recession. Sources: U.S. Treasury, Macrobond, RBC GAM

Granted, the latest move needs to be put in the context of the massive selloff that preceded it, and yields are far higher than they were a year ago, let alone three years ago.

Still, the point is that bond yields are now falling. Why?

A main answer is that inflation cooperated nicely in October after a few months of misbehaviour. This has greatly increased the market’s comfort with the declining inflation narrative. We think the November numbers should also help the cause.

I continue to flag the chance that the rate cutting is even more substantial than the market prices in. In our view, a recession is likely and inflation will undershoot expectations, leading to faster, deeper rate cuts.

Markets have accordingly begun to price in additional rate cuts. Much of this move just represents an unwinding of more hawkish expectations that had formed in the late summer when inflation was temporarily not cooperating. But ultimately the move has extended even further downward than that. The market is now pricing a full 25bps rate cut by May for the U.S. and implying that the fed funds rate will eventually settle just under 3.5% in a few years – well below earlier assumptions (see next chart).

Market expects interest rate cuts to start in 2024

As of 12/08/2023. Sources: Bloomberg, RBC GAM

Markets are hoping for a soft landing as inflation settles. The stock market has rallied on the expectation that earnings continue to grow, further boosted by a lower discount rate (see next chart). Since October 27, the S&P 500 has appreciated by a remarkable 12%. The credit market is also happy.

Equities rebound as yields fall

As of 12/08/2023. Sources: S&P Global, Macrobond, RBC GAM

With admittedly less conviction than a few months ago, given how much bonds have rallied in the interim, I continue to flag the chance that the rate cutting is even more substantial than the market prices in. In our view, a recession is likely and inflation will undershoot expectations, leading to faster, deeper rate cuts.

In past recessions, the Fed has typically cut its policy rate by around 3 to 4 percentage points over the span of a year. A move of that magnitude is unlikely this time given the delicate inflation situation. The relevant observation, however, is that central banks usually take the elevator rather than the escalator down.

Of course, our anticipated hard landing wouldn’t be good for the stock market or other risk assets.

Strong data, weakening trendline

Some recent prominent economic indicators have been solid in the U.S., including consumer confidence and November payrolls. The latter managed a decent 199,000 new jobs. The number wasn’t quite as good as it looked, however, given that:

• 30,000 of those jobs were merely striking auto workers returning to their jobs.

• There were -35,000 in revisions to prior months.

• Private sector hiring alighted slightly short of expectations.

That said, and pointing in the opposite direction, the unemployment rate fell from 3.9% to 3.7%. That’s a large drop, with the result that the unemployment rate is now some distance from breaching the Sahm’s Rule recession threshold. The Sahm Rule identifies the start of a recession when the three-month moving average of the national unemployment rate rises by 0.50 percentage points or more relative to its low during the previous 12 months.

This latest unemployment decline needs some explanation. It is surprising in the context of the unremarkable number of new jobs created according to the latest payroll survey (Current Employment Statistics Survey).

The answer lies in the fact that the unemployment rate is calculated using a different labour market survey, the household survey (Current Population Survey). It claims that the U.S. added an incredible 747,000 new workers in November, pushing unemployment down. But this survey is known to be more volatile than the payroll survey.

The true (unobserved) unemployment rate probably didn’t actually fall by much this month. But it probably also didn’t rise by much the month before, either.

As an example, the prior month’s household survey claimed the loss of 348,000 jobs (see next chart). Averaging it with this month’s numbers produces a more believable +199,500 new jobs per month. This is roughly in line with the payroll survey.

In turn, the true (unobserved) unemployment rate probably didn’t actually fall by much this month. But it probably also didn’t rise by much the month before, either.

U.S. employment may be weaker than it looks

As of November 2023. Sources: U.S. Bureau of Labor Statistics (BLS), Macrobond, RBC GAM

We still see evidence of insidious deterioration in other labour market indicators. For example:

• The closely watched ADP survey managed just 103,000 new jobs for the same month.

• Temporary employment continued to fall.

• Youth unemployment remains notably higher than a few quarters ago (see next chart).

• Job openings tumbled in the latest month.

• Continuing jobless claims are now significantly higher than they had been in September.

U.S. youth unemployment has gone up and returned to pre-pandemic levels

As of November 2023. Shaded area represents recession. Sources: BLS, Macrobond, RBC GAM

Elsewhere, purchasing manager indices actually held together fairly well in November.

• The U.S. Institute for Supply Management (ISM) Manufacturing Index was unchanged at a soft 46.8. The ISM Service Index rose from 51.8 to a modest 52.7.

• Eurozone PMIs rose slightly, albeit from quite depressed levels.

• The U.K. PMIs rose substantially, though not to levels consistent with outright strength.

So is everything going swimmingly on the economic front? We would argue “no.” Despite the recent solid numbers, the bulk of the economic data remains fairly soft and is deteriorating. The Citigroup Economic Surprise Index for the U.S. has plummeted from extremely positive readings to only modestly positive ones (see next chart).

U.S. economic data have softened

As of 12/11/2023. Sources: Citigroup, Bloomberg, RBC GAM

The Citigroup Data Change Index for the U.S. rose almost without ceasing for the better part of two years, before beginning to stumble in November and December (see next chart).

U.S. economic data starting to waver

As of 12/11/2023. Sources: Citigroup, Bloomberg, RBC GAM

Beige Book

The U.S. Federal Reserve publishes the Beige Book eight times a year. The report usually provides deep qualitative insight into the state of the U.S. economy. As with the prior Beige Book, the main sentiment in the latest edition is gloom. Of the 12 Federal Reserve Districts, just four described an environment of “modest” growth, versus two that characterized the economy as flat to slightly down. A whopping six that said their economies are declining slightly. That means eight out of 12 U.S. Districts are not growing.

Past experience with this survey argues that it would be greatly premature to say that the U.S. is therefore in a recession already. But there is definitely weakness present.

This theme of weakness then repeatedly resurfaces in the more granular discussion: 

• Housing weakened slightly.

• The sales of discretionary items and durable goods like furniture and appliances declined.

• Manufacturing was mixed, and the manufacturing outlook weakened.

• The demand for labor continued to ease.

• The economic outlook over the next six to 12 months also diminished over the reporting period.

Black Friday

It is sometimes useful to examine the Black Friday and Cyber Monday sales figures to gain insight into how the consumer is doing. Estimates varied substantially, with overall sales thought to have increased by between 2% and 8% over last year.

The midpoint of this (+5%) sounds decent. But it is actually fairly pedestrian after adjusting for inflation and population growth. In addition, discounts were deeper this year and thus more attractive – pulling shoppers into purchasing.

The real X-factor, however, is the extent to which consumers have become more price sensitive as they exhaust their pandemic savings. They may have spent a disproportionate share of their holiday budget on discounted items over the Thanksgiving holiday, leaving less money for purchases later in December.

Canada

Turning to Canada, the country’s November job creation looked quite good on the surface, but it had a soft underbelly. The good news was that the country created 25,000 new jobs over the month. That’s a solid number by historical standards, handily beating the consensus forecast for 14,000 new positions. The hiring was all full time, and all private sector.

Despite all of that, the final interpretation was considerably soggier. Recall that at a time of rapid population growth, Canada actually needs 40,000-plus new jobs each month. That’s just to keep pace with the number of newly available workers. The unemployment rate accordingly rose, from 5.7% to 5.8% -- now fully 0.8 percentage points higher than the cycle low.

The economy is already essentially stagnating. Our composite of Canadian business expectations remains quite weak.

Finally, and perhaps most importantly (if slightly surprising give the aforementioned full-time hiring), total hours worked in the country fell by 0.7% in November versus October. The economy was using less labour for the second month out of the past three (see next chart).

Canadian labour market has softened lately

As of November 2023. Sources: Statistics Canada, Macrobond, RBC GAM

Canadian economic weakness was also visible in the release of the third-quarter gross domestic product (GDP) numbers. Canada suffered a 1.1% annualized decline in output. That’s materially worse than expected. The interpretation was somewhat blurred by the upward revision to the prior quarter, such that the ignominy of two consecutive quarters of declining GDP was avoided.

But the trend is certainly not a friendly one. The economy is already essentially stagnating. Our composite of Canadian business expectations remains quite weak (see next chart).

Business leading indicators paint a gloomy outlook for Canadian economy

As of November 2023. Composite constructed using four leading indicators from surveys on Canadian businesses. Shaded area represents recession. Sources: CFIB, Haver Analytics, Macrobond, RBC GAM

Happy inflation trend

The inflation trend remains favourable. Late-arriving data for October continues to arrive below expectations, and considerably lower than the prior month:

• Canadian CPI (Consumer Price Index) declined from 3.8% to 3.1% year-over-year (YoY).

• The U.S. PCE (Personal Consumption Expenditures) deflator for October was similarly reduced, from 3.4% to 3.0% YoY.

The November numbers are now starting to arrive, and continuing the happy trend. Eurozone CPI has plummeted to just 2.4% YoY. However, it should be noted that shelter costs figure less centrally in the European basket, and this has been one of the components preventing a faster decline elsewhere. Note also that the sky-high European energy prices of a year ago are now coming off, providing a temporary depressant for Eurozone inflation numbers that other regions cannot hope to replicate.

At the time this is written, U.S. CPI for November will be released imminently. The consensus forecast is for a flat price level in November versus October. In fact, our analysis of real-time metrics among other inputs argues that the risks may even lie to the downside of this estimate (see next chart).

That admittedly won’t allow the year-over-year metric to improve all that much further – it might drop by a tenth of a percentage point or two, landing at or just above 3% YoY. Core inflation is expected to log another 0.3% monthly increase, leaving the annual number unchanged at 4.0% YoY.

U.S. Daily PriceStats Inflation Index shows positive trend

PriceStats Inflation Index as of 12/08/2023. CPI as of October 2023. Sources: State Street Global Markets Research, RBC GAM

Inflation in December could then be slightly hotter again, based on our preliminary readings. But the broader point is that inflation has been pleasantly surprising for most of the past year, and we continue to believe it may fall a little faster than the markets imagine over the year ahead.

Revisiting the risk of structurally high inflation

It is worth occasionally checking in on other inflation scenarios. The worst outcome would be a 1970s-style stagflation outcome, with inflation remaining persistently high and this, in turn, corroding economic growth.

This risk cannot be ruled out. After all, inflation is still running at 3%-plus, almost regardless of the metric chosen. But the risk is surely lower than it was 18 months ago:

• The four big original drivers of inflation have all turned lower and continue to point in a more favourable direction.

• The commodity shock has faded.

• Supply chains have healed.

• Monetary policy has gone from stimulus to restraint.

• Even fiscal stimulus is far less generous than it was a few years ago.

Accordingly, the money supply is shrinking and the breadth of high inflation has narrowed massively. Just 1% of the U.S. price basket is now rising at a double-digit annual rate, down from more than a third in the worst month.

A more likely scenario in our view is that inflation falls further and ends up being only a little higher than normal.

To be fair, shelter costs are proving tricky in some countries. If home prices and rents don’t settle down further, inflation may also struggle to settle. Also, wage growth is decelerating but not fully normalized. Large union settlements will keep a portion of wages rising quickly for several years to come.

Finally, if a recession is avoided, it isn’t entirely clear that the labour market and corporate pricing power will cool to the extent necessary to achieve fully normal inflation. This cocktail of factors could congeal into a persistently high inflation rate.

Despite this, a more likely scenario in our view is that inflation falls further and ends up being only a little higher than normal due to a bit of persistent scarring, and as central banks make a small unstated tradeoff between the economy and prices.

Thankfully, there are some important differences between the 1970s and today:

• Inflation expectations are more anchored today, and at a lower level.

• The level of unionization is lower today. There are fewer multi-year wage settlements that keep costs rising for an extended period of time.

• The demographic environment is also very different. The 1970s had a population that skewed young whereas developed nations are much older today. While the theoretical implication of this for inflation is much debated, the empirical reality is that older countries have less inflation.

Finally, some lessons were learned from the errors of the 1970s. Notably these include preventing central banks from being politicized and encouraging them act quickly when inflation first accelerates. Modern central banks have some egg on their face given that it took them more than six months to snap into rate hikes. But this is nothing compared to the errors of the 1960s and 1970s. Before the 1970s even began, U.S. inflation had already run above a 3% annual rate for four straight years. That simply wasn’t the case today. 

Central bank rate cuts?

Federal Reserve

The next U.S. Federal Reserve decision is immediately ahead as these words are written. The Fed funds rate will almost certainly be left unchanged at 5.50%, but there is much debate over whether the Fed could signal that rates are at a peak and whether it could even begin signaling cuts before too long.

We do believe rate cuts are entirely conceivable once 2024 has begun – predicated on a recession and inflation continuing to cooperate. But this is unlikely to be the meeting that officially points in that direction. Instead, the Fed will probably push back against dovish expectations, and convey only a modest decline in the dot plots.

The economy has not been soft enough over the past six weeks to warrant a major change of direction, and inflation – while improved – is still clearly too high.

The most recent Fed Minutes continued to discuss the possibility of additional policy tightening. We are not far removed from Chair Powell saying that the battle against inflation “has a long way to go.” The economy has not been soft enough over the past six weeks to warrant a major change of direction, and inflation – while improved – is still clearly too high. Financial conditions have eased quite a lot since the last meeting, reducing the need for rate cuts.

Practically, if the Fed were to signal near-term rate cuts, bond yields would almost certainly fall further, undoing the monetary restraint that the Fed presumably believes is appropriate until a later date.

Bank of Canada

The Bank of Canada’s most recent decision on December 6 unveiled some useful insights. The central bank explicitly acknowledged that higher interest rates were restraining spending and that Canadian GDP had stalled across the middle of 2023. The U.S. economy was described as having been stronger than expected so far. But it is forecast to weaken in the months ahead.

Despite this, the Canadian inflation picture is tricky because shelter inflation picked up. Canadian CPI picks up mortgage interest costs (which have soared) and rent payments (which are still rising robustly). Falling home prices help, but do not provide a full offset.

Perhaps providing a preview of the U.S. approach at its next policy decision, the Bank of Canada maintained its hawkish bias, saying that it “remains prepared to raise the policy rate further if needed.”

Global trend turning

Even though developed world central banks are still some distance away from rate cuts, a broader pivot at the global level is already visible (see next chart). The fraction of central banks raising rates has plummeted from nearly 80% to barely any. Meanwhile, the fraction of central banks cutting rates have increased from nothing to a modest number of mostly emerging market central banks. Should the trend continue, more of the world’s central banks will be cutting rates than raising rates within the next few months.

That sounds reasonable, with emerging market central banks again leading the way, just as they did on the way up.

Central banks are almost done raising rates

As of 11/27/2023. Based on policy rates for 30 countries. Sources: Haver Analytics, RBC GAM

No more negative yields

Incidentally, it is amazing when one reflects back on the wild ride for bonds over the past nine year (see next chart). In 2014, virtually all of the world’s bonds traded at a positive yield. Then, due disproportionately to declining Japanese and European policy rates, a significant fraction of the world’s bond started to trade with a negative yield.

Share of bonds with negative yields is now close to zero

As of 12/08/2023. Percentage of bonds in Bloomberg Global Aggregate Bond Index trading at negative yields. Sources: Bloomberg, RBC GAM

This was, of course, absurd on the surface: who would want to buy something that promised to pay less money back at a later date? But the reality was that several factors sweetened the story, including:

• forecasts for deflation (a scenario in which a negative nominal yield could still generate a positive real return)

• expectations that yields would fall further (allowing for a capital gain in bonds that might outweigh the negative yield)

• the need for a safe liquid investment.

Indeed, at the worst moment in early 2020, around 30% of the world’s bonds traded with a negative yield.

Today, only a tiny handful of short-dated Japanese bonds trade at a negative yield. It is much preferable to be a bond investor today!

As central banks have increased interest rates in recent years, this fraction has mercifully plummeted. Today, only a tiny handful of short-dated Japanese bonds trade at a negative yield. It is much preferable to be a bond investor today!

Busy election year ahead

The past year was one of the slowest for national elections in decades, and without a single G7 election. In sharp contrast, 2024 is set to be the busiest year for national elections on record. For the first time, more than half of the world’s population will vote in national elections in a single year.

• The U.S. election will be a high-stakes affair, potentially pitting incumbent Biden against former President Trump and with polls showing a potentially close race.

• The U.K. election is very likely to entail a switch from Conservative to Labour Party rule.

• Polling for the upcoming Indian election shows that President Modi’s coalition remains in the lead, though the margin of 2—4 percentage points in polls is not insurmountable.

Indonesia, with an enormous population of 274 million people, is set to be the fourth largest driver of economic growth over the next five years, behind only China, India and the U.S. What is more, it has a presidential election early in the New Year. Incumbent Widodo has already served two terms, meaning that someone new must take charge. However, the change may prove to be minimal as the leading presidential candidate – Defense Minister Prabowo – is not only from the same party, but recently selected current President Widodo’s son as his running mate.

Prominently, South Korea, Russia, Taiwan, South Africa and Mexico also have elections in 2024.

Global minimum tax

In 2021, nearly 140 countries agreed on a global minimum corporate tax rate of 15%. They pledged to charge at least 15% domestically and to apply a special top-up to companies that lowered their effective tax rate below the minimum through the use of non-compliant tax havens.

This has been slow to move from international pledge to national-level implementation. Finally, 2024 will see a significant fraction of the global economy bring these rules into force. This included the European Union, the U.K., Japan and South Korea. A number of low-tax jurisdictions are also in the process of raising their corporate tax rates to 15% to avoid being ensnared by the rules.

It remains to be seen just how effective the new rules will be . . . [I]t may be a time when corporate profit margins do not rise as easily as in the past.

The U.S. and China remain prominent holdouts. The U.S. actually passed a minimum corporate tax rate of 15% domestically, embedded within the Inflation Reduction Act. But this is different than the internationally agreed-upon version, and there is sufficient political opposition that it isn’t clear that it will pass.

A significant number of emerging market countries have lately expressed reservations about the program. It limits their ability to use tax holidays and incentives to nurture nascent industries, if such subsidies would reduce the effective tax rate too far.

It thus remains to be seen just how effective the new rules will be. But when combined with greater anti-trust efforts and additional scrutiny against corporations, it may be a time when corporate profit margins do not rise as easily as in the past.

De-globalization under the microscope

After several decades of globalization – an economic tailwind that saw countries trading with one another ever more -- the pendulum is beginning to swing back in the opposite direction. In recent years, global trade has risen less quickly than GDP (see next chart). That is the definition of de-globalization.

Globalization now in reverse

Ratio of 5-year growth of real export of goods and services to that of real GDP. Ratios for 2023 and later based on Organisation for Economic Cooperation and Development (OECD) forecast. Shaded area represents U.S. recession. Sources: OECD Economic Outlook November 2023, Haver Analytics, RBC GAM

Globalization had been decelerating for years, for a range of reasons:

• The world had already integrated to a remarkable degree.

• There was little room left for already-low tariffs to decline further.

• Wage differentials have shrunk to an extent that reduces the scope for wage arbitrage.

• Rising automation reduces the importance of labour cost differentials.

• The world is simply becoming more homogenous – resulting in fewer gains from trade.

What has prompted the latest shift to outright de-globalization?

The pandemic didn’t help global trade in the early going. However, there have been a few years of unencumbered and only minimally distorted demand and supply patterns since, without a full revival. So it likely wasn’t just the pandemic.

A little bit of recent de-globalization may relate to a cyclical economic deceleration that affects highly volatile sectors like trade more than the broader economy (see next chart).

Global trade is contracting

As of September 2023. Shaded area represents U.S. recession. Sources: CPB Netherlands Bureau for Economic Policy Analysis, Macrobond, RBC GAM

But there are also structural reasons for de-globalization. Tariffs have risen in recent years for a few different reasons, including:

• U.S.-China frictions

• embargo against Russia after the invasion of Ukraine

• countries simply tilting in more protectionist directions.

Global trade restrictions of all types have boomed in recent years (see next chart).

Global trade restrictions ballooned in recent years

As of 2022. Sources: Global Trade Alert, International Monetary Fund (IMF), fDi Intelligence, RBC GAM

Industrial policies that subsidize critical domestic industries are on the rise. While not strictly tariffs, they do impede foreign competition and trade. The same goes for “buy local” legislation.

After suffering through supply chain disruptions during the pandemic, there is today a greater emphasis on supply chain resilience. It is no longer just about the cheapest production but also about ensuring that production doesn’t halt. Two factories in two countries are probably more expensive than one, but achieve greater resilience.

Companies are particularly sensitive to geopolitical risks in their calculus. Having factories in a country that might become an adversary is risky, and so a degree of “friend-shoring” and onshoring is now happening to minimize this risk.

But just how much de-globalization, friend-shoring and on-shoring is actually happening? Some, clearly . . . [b]ut the rate of decline in globalization is nothing like the pace of advance across the 1990s and 2000s.

Many companies are dipping their toes into the water by engaging in China Plus One strategies that involve continuing to manufacture in China, but experimenting with production in other countries. The idea is that if this production goes well, and as the infrastructure is upgraded in other countries, there will be scope to transfer more production later. Many multinational companies simply contract Chinese factories rather than own their own facilities, making this an even more attractive strategy.

Finally, as environmental concerns mount, some countries are beginning to charge carbon taxes – effectively, tariffs – on goods produced in environmentally unfriendly ways in foreign countries. This approach allows countries to properly capture the environmental externality while simultaneously allowing their domestic corporation champions to compete on a level playing field. This has the effect of increasing trade barriers.

But just how much de-globalization, friend-shoring and on-shoring is actually happening? Some, clearly, as evidenced by the first chart in this section. But the rate of decline in globalization is nothing like the pace of advance across the 1990s and 2000s. Our base-case scenario does not involve a full retreat back to the stone age of the 1980s.

One reason for this tempered view is that a significant amount of so-called “South-South” trade is now occurring: developing nations trading with fellow developing nations. This remains a potential source of trade growth and makes a great deal of sense as these economies capture an ever-greater share of the global economy.

What about friend-shoring? This is genuinely quite significant. We can see the exports for certain industries such as furniture and footwear pivoting from China to other countries like Vietnam – even if some of this is due to China moving up the value chain rather than purely on the basis of friend versus foe (see next chart).

Vietnam capturing rising share of global furniture exports

Sources: Gabriel Cortes, CNBC, MDS Transmodal

Direct trade between the U.S. and China has fallen markedly as a share of the total trade conducted by each nation over the past five years (see next chart). The U.S. share of Chinese exports is now at its lowest in decades, while the Chinese share of U.S. imports is in steep decline and the lowest in nearly 20 years.

U.S.-China trade has been declining

U.S. import data as of September 2023. China’s export data as of September 2023. 12-month moving average of trade data used in the calculations. Shaded area represents U.S. recession. Sources: China General Administration of Customs (GAC), U.S. Census Bureau, Macrobond, RBC GAM

As a result, China has just relinquished its decade-long crown as the largest exporter to the U.S. (see next chart). China has now fallen below the EU. If that’s cheating because the EU is a consortium of many countries, then it is notable that China has also just fallen below Mexico, and by all appearances is about to fall below Canada as well in the near term. The EU and Mexico are simultaneously on the upswing, expanding their share of U.S. trade.

U.S. trade is shifting away from China

As of September 2023. Sources: U.S. Census Bureau, Macrobond, RBC GAM

Mexico is an especially promising trade partner for the U.S, given its excellent geographic location, large population and low-cost labour (only one-third the cost of Chinese workers). There are certainly barriers in the form of inadequate infrastructure (including not just ports and highways but electricity), insufficient human capital and gang violence. But the proof is in the pudding: Mexican trade accelerating with the U.S. and new factory announcements are happening regularly.

At the same time, new foreign direct investment into Mexico is rising palpably. Investment in heavy machinery is skyrocketing. Industrial park vacancies have fallen to just 2.1%. Exports to the U.S. of Mexican automotive and electronics products are particularly large and growing quickly.

The U.S. is also increasingly trading with other developing Asian nations (see next chart). Alongside Mexico, these are perhaps the most important friend-shoring partners given their low labour costs, stability and improving infrastructure.

China is losing share of U.S. exports to the rest of developing Asia

As of Q2 2023. Developing Asia includes China, India, Vietnam, Thailand, Malaysia, Indonesia, Philippines, Bangladesh, Cambodia and Laos. Sources: IMF, Macrobond, RBC GAM

China is not just losing out on U.S. trade to developing Asian nations but is also losing ground on a global basis (see next chart).

China is gradually losing share of global exports to the rest of developing Asia

As of Q2 2023. Developing Asia includes China, India, Vietnam, Thailand, Malaysia, Indonesia, Philippines, Bangladesh, Cambodia and Laos. Sources: IMF, Macrobond, RBC GAM

A significant complication is that the West may be decoupling from China less than the trade numbers would suggest. Even though imports are increasingly coming from the likes of India and Vietnam, very often China is a key supplier of components to these countries. Frequently, the companies operating in the rest of developing Asia are also Chinese. The result is that China still benefits from the trade, China is integrating itself more deeply with those nations, and any serious conflict with China would still freeze production.

Some countries will win and others will lose. The most obvious loser is China. Prospective winners include India, Vietnam and Mexico most prominently.

What about on-shoring? This is also a real trend, though probably overstated. The recent rise in manufacturing construction is extremely focused in the computer, electronic and electrical sector, which has been subsidized by recent U.S. legislation (see next chart). Conversely, other sectors show little evidence of robust growth, though it should be conceded that most are trending mildly higher. Without tax inducements, companies are not actually racing back to their home market.

Recent rise in U.S. manufacturing construction driven by computer, electronic and electrical sector

As of June 2023. Private manufacturing construction spending deflated by Producer Price Index for Intermediate Demand, materials and Components for Construction. Shaded area represents recession. Sources: Census Bureau, BLS, Macrobond, RBC GAM

Thus, we can say that on-shoring is real but overstated, friend-shoring is very real, and de-globalization is also real but perhaps slightly overstated.

The central implication is incrementally less economic growth than otherwise as the world breaks into cliques. Inflation should run slightly higher than otherwise as production costs are de-prioritized. Some countries will win, and others will lose. The most obvious loser is China. Prospective winners include India, Vietnam and Mexico most prominently.

History shows that multi-polar eras like this one tend to last for an extended period of time, so a reversal is not especially likely in the coming years.

The promise of artificial intelligence

We are already on the record with the view that recent advances in artificial intelligence (AI) may constitute the world’s next general-purpose technology – something that doesn’t just make one company or sector richer, but that pervades all sectors and the lives of most people. Past examples of general-purpose technologies include the internet, the computer, the air conditioner, the internal combustion engine, the corporation, and many more. If we are correct, productivity growth can advance more quickly for a sustained period of time.

Whether AI meets this lofty expectation remains an open question, but recent developments have been promising, not just in language models, but also in quite different pursuits. A prominent and incredible example comes from a recent paper published in Nature, which used AI to identify 2.2 million new theoretical crystal structures, increasing the number known to scientists by a factor of 46 in one fell swoop.

If this sounds esoteric, such crystals may yet be used to design superior materials for human use, perhaps ones that are especially strong, flexible, durable and/or environmentally sustainable. One of the crystals could unlock superconductivity at room temperature, or allow for a leap forward in battery technology, affording sufficient range gains and cost savings as to make electric cars indisputably superior to their internal combustion competitors.

A secondary question is how long will it take before productivity starts to rise? Usually, it is a matter of many years as people and companies grapple with how to best scale up and deploy new technologies.

It still seems reasonable to expect a substantial lag, especially given the shortage and cost of computer chips necessary to train modern AI models. The OpenAI supercomputer reportedly required 10,000 graphics processing units. The market price for the top NVIDIA AI chip is presently US$10,000 retail and up to US$40,000 on the grey market.  This means a cost of US$100 million to US$400 million just for one AI company, and presumably future generations of the technology will require even more/better chips.

The current generative AI technology is also still prone to hallucinations and biases that are problematic for many potential uses – and so requires further improvement.

On the other hand, it is notable that it took just two months for ChatGPT to reach 100 million users, far faster than the technologies and celebrated applications that preceded it. Generative AI technologies are also inexpensive to use once the model has been built, the new technology has been widely publicized, it is entering a highly IT literate world, it is easily scalable, and the avenues to innovate using it appear virtually limitless. Thus, AI may boost the productivity numbers sooner than normal, though still not instantaneously.

-With contributions from Vivien Lee and Aaron Ma