Global Investment Outlook

Our quarterly Global Investment Outlook is now available online. The economics article, entitled “Less inflation plus slower growth equals rate cuts”, can be found on page 15. Note that our thinking continues to evolve since this was published, including an update to the recession risk discussed in the next section of this #MacroMemo.

Recession risk drops

We hereby downgrade the recession risk for the U.S. over the next 12 months from 40% to 30%. Several developments support this shift:

• The U.S. Federal Reserve (the Fed) cut its policy rate by 50 basis points – more than anticipated. This simultaneously supports the economy and signals that the Fed could continue to move quickly if the economy were to stumble.

• Economic data has improved over the past six weeks. The ISM (Institute for Supply Management) Services index rebounded back above 50. The unemployment rate edged lower. Jobless claims are falling steadily. Retail sales reported an unexpectedly chipper outcome. Third-quarter GDP (gross domestic product) is tracking a strong +2.9% annualized.

• There is little stress visible in credit markets. Credit spreads are narrow, mortgage rates are falling and bank lending standards are easing.

• Lower oil prices provide a further support for economic growth, and also reduce the risk of inflation reigniting.

All told, the risk of a U.S. recession has fallen substantially, though it remains higher than normal. A normal recession risk might be 10-15% for the average year. Some recession signals continue to blink red (refer to the yield-curve discussion later). In addition, the pain of high interest rates will linger for a while longer.

Recession risk elsewhere

The risk of recession remains incrementally higher elsewhere, such as in Canada and the Eurozone, where economic growth has sputtered to a greater extent.

In Canada’s case, the unemployment rate has now risen by a large 1.8 percentage points. The U.S.-oriented Sahm Rule (which decrees that a 0.5 percentage point increase in the 3-month moving average of the unemployment rate has always resulted in a recession) has not historically worked in Canada, where fewer than half of such events have culminated in a recession. But every time Canada’s unemployment rate has increased by more than 1.5 percentage points in the last 50 years, a recession has resulted. This particular threshold has already been breached now (though the unemployment rate did rise by 3.7 percentage points in the late 1960s through early 1970s without a recession).

That said, no one rule is perfect. Canada’s unemployment rate has been pushed up significantly by surging immigration rather than job losses, with less sinister implications. Optimists can further argue that Canada already experienced something like a recession given its sharp decline in GDP per capita, with rapid population growth minimizing the aggregate damage (see next chart). As GDP growth stabilizes and GDP per capita bottoms out, perhaps the worst is over.

Canadian growth has slowed markedly

As of Q2 2024. Sources: Statistics Canada, Macrobond, RBC GAM

In the context of the country’s recent economic performance and dovish comments from Governor Macklem, and with the U.S. Federal Reserve having broken the seal on 50-basis-point rate cuts for this cycle, there is a solid chance that the Bank of Canada opts to unveil a 50-basis-point move of its own within the next few months.

Inflation trend improves

The U.S. Consumer Price Index (CPI) improved somewhat in August. The headline print descended from +2.9% to +2.5% year-over-year (YoY). Core inflation was less impressive at +3.2% YoY, rising by a brisk 0.3% month-over-month (MoM). Overall, though, most inflation forces continue to abate. Indeed, of the 11 major inflation sub-components, only housing contributed in any material way to the increase in monthly prices in August (see next chart).

What has contributed to the latest U.S. monthly inflation rate

As of August 2024. Sources: U.S. Bureau of Labor Statistics, Macrobond, RBC GAM

A wide range of inflation metrics are now flashing a welcome shade of green (see next chart). This was not at all the case in the spring. The exception remains shelter costs. This shouldn’t be downplayed, but the lags associated with shelter inflation do continue to argue that the component can ease somewhat further from here, if not become altogether subdued.

Update on key inflation metrics

As of August 2024 for CPI and Producer Price Index (PPI) measures, July 2024 for Personal Consumption Expenditures (PCE) measures. Sources: U.S. Bureau of Economic Analysis (BEA), BLS, Federal Reserve Bank of Cleveland, Federal Reserve Bank of Dallas, Macrobond, RBC GAM

Excluding shelter costs, U.S. CPI is already well below the country’s 2.0% inflation target (see next chart).

U.S. inflation excluding shelter has returned to 2%

As of August 2024. Shaded area represents recession. Sources: BLS, Federal Reserve Bank of Cleveland, Macrobond, RBC AM

Turning away from the latest inflation results to key inflation drivers, the news is mostly good. The fraction of businesses planning price hikes is again in retreat and not too far from normal after a brief flare-up (see next chart).

Fraction of U.S. businesses planning to raise prices is approaching pre-pandemic levels

As of August 2024. Shaded area represents recession. Sources: National Federation of Independent Business, Macrobond, RBC GAM

Medium-term inflation expectations – both over the next five years and over years six through 10 – have decreased meaningfully in recent months, confirming a rising confidence that inflation is indeed enduringly settling (see next chart).

U.S. medium-term inflation expectations declined

As of 09/12/2024. Sources: Bloomberg, RBC GAM

While it is important to appreciate that economy-wide prices are unlikely to meaningfully decline despite their large earlier run-up, some individual components are giving back a portion of their earlier gains. A great example is U.S. wholesale used car prices, which continue to trend downward after exploding higher during the first few years of the pandemic (see next chart).

Wholesale used car prices have retreated from peak but remain elevated

As of August 2024. Shaded area represents recession. Sources: Manheim Consulting, Macrobond, RBC GAM

Commodity prices are lower than a few years ago, and also down relative to a few months ago (see next chart).

Commodity prices have declined recently

As of 09/11/2024. Shaded area represents U.S. recession. Sources: S&P Global, Macrobond, RBC GAM

Shipping costs, which had risen concerningly on Red Sea disruptions and a Panama Canal water shortage, are now again falling. This has reduced the threat of serious supply chain problems interfering with the economy or reigniting inflation.

Shipping costs are now decreasing as peak season was pulled forward

As of the week ending 09/12/2024. Sources: Drewry Shipping Consultants Ltd., Macrobond, RBC GAM

Overall, inflation appears likely to continue edging lower over the coming quarters, on its journey back toward something resembling pre-pandemic normality (perhaps plus a few tenths of a percentage point given new forces such as pandemic scarring, de-globalization and climate change).

The U.S. Fed goes 50

The U.S. Fed eased by a chunky 50 basis points on September 18, opting for the greater of the two rate-cut magnitudes debated by markets (the other was 25 basis points, which we had favoured). In so doing, the Fed has significantly made up for its late start, now out-easing the Bank of England in its cumulative action, and catching up to the European Central Bank’s two 25bps rate cuts.

A risk in cutting by 50 basis points is that the market would interpret the move as smacking of panic in the context of a decelerating economy. But that isn’t how the market took it, and to the Fed’s credit it skillfully communicated the big rate cut as representing a good start, and reflecting confidence that inflation is coming down toward 2%.

Looking forward, financial markets expect a further 75 basis points of easing by the end of the year, meaning one 50bps rate cut and one 25bps rate cut spread over the next two decisions. For its part, the Fed’s dot plots indicate just 50 basis points of easing over the next two meetings. This is the debate, with the flow of economic data likely to determine just how quickly the Fed continues to move. For the moment, we would guess at two 25 basis point rate cuts to finish the year.

For 2025, markets price the Fed cutting fairly quickly down to a fed funds rate below 3% by the end of the year (see next chart). This is entirely possible, though it runs ahead of the Fed’s own forecast that the policy rate ends 2025 at 3.4%. Again, we think the truth may lie between the two, though much depends on whether significant slack opens up in the economy, in which case the Fed would be entirely justified in cutting into outright stimulative territory. For the moment, we note that the market has had a tendency to overshoot in one direction and then the other when it comes to Fed rate cuts. In that context, it might be pricing in a bit too much right now.

Fed jump-starts the easing cycle

As of 09/19/2024. Sources: Bloomberg, RBC GAM

There was one dissent in the vote at the Federal Open Market Committee (FOMC). While this itself is not especially unusual, and all the more so given that the debate wasn’t over action versus inaction but instead merely over the size of the rate cut, what was unusual was that the dissent came from a Fed governor rather than a Fed district president. District presidents are geographically removed from one another, have regional economic considerations to weigh, and have their own research teams that generate different forecasts and thus optimal monetary policy. In contrast, the Fed governors are all located in Washington DC and have historically spoken with something closer to a single voice. This was the first dissent of a Fed governor in 19 years.

It is tempting to read something political into the decision given that it was a Trump appointee and longtime Republican who preferred a smaller rate cut just before the presidential election, but that isn’t fair as the majority of economic professionals also favoured a 25-basis-point move and one could just as easily (and just as wrongly) argue the other Fed voters – all with their own political inclinations – were playing politics by making a surprisingly large move just before the election.

Yield curve un-inverts

A key part of the U.S. yield curve just un-inverted. The 2-year bond yield has just fallen back below the 10-year bond yield after a modern-day record 566 days of inversion.

Given that an inverted yield curve is a classic recession signal, did the recession signal just go away?

No, for two reasons. First, two other parts of the yield curve that have historically presaged recession remain inverted (see next chart). So the inversion story isn’t completely over.

Yield curve indicators diverge

As of 09/20/2024. Near-term forward spread measured as forward rate of 3-month Treasury bill six quarters from now minus spot 3-month Treasury yield. Shaded area represents recession. Sources: Engstrom and Sharpe (2018), FEDS Notes, Washington Board of Governors of the Federal Reserve System, Bloomberg, Haver Analytics, RBC GAM

Second, a yield curve that un-inverts is actually still predicting a recession under certain circumstances. Those circumstances are currently being met. An inverted yield curve is a good predictor that a recession will arrive within a year or two, not tomorrow. It isn’t a near-term signal. The near-term signal is that the yield curve then un-inverts right before the recession as part of a bull steepener (meaning that bond yields are falling and short-term rates are falling by more than long-term rates). The recent bond market action has indeed been a bull steepener. That means the 2-10 curve is still moving along a traditional recession trajectory.

So why don’t we think a recession is likely? There are a few reasons.

1. There are plenty of other recession signals that are not presently triggered, including the fact that lending standards are easing and global trade is rising. No one signal is infallible, even if it this particular one hasn’t made any mistakes yet. The sample size is small for the U.S., and other countries like the U.K. have spent significant chunks of time with inverted yield curves that led to nothing.

2. The yield curve normally inverts (and then un-inverts) before a recession because rates start at a roughly normal level and then a pessimistic outlook pulls down the long end before urgent rate cuts designed to address the recession then pull down the short end by even more. Those aren’t the circumstances now. Today, interest rates are substantially restrictive. The recent curve inversion has less to do with a pessimistic outlook expressed by the long end, and more to do with an expectation that rates will eventually settle back down to a more normal level. Today, as central banks cut rates, they aren’t doing it out of desperation as a recession sets in, but instead because they see the battle with inflation being won. Put a different way, rate cuts are usually trouble because they are delivered in response to an undesired economic downturn. These rate cuts may not be trouble as they are being delivered mainly in response to a much-desired inflation downturn.

3. The term premium remains unusually compressed. In this environment, bond investors don’t demand significant additional compensation for longer term bonds (controlling for expected inflation and expected monetary policy). As a result, it is much easier for the yield curve to invert. It used to take a giant rally in the long end (and implicitly a really pessimistic medium-term economic outlook) to invert the curve. Without much of a term premium, you need much less of a move to get the curve inverted, and so the level of economic pessimism is not as extreme today.

Confused? The main points are a) the recent un-inversion of the 2-10 curve doesn’t negate the recession signal, instead advancing it one step closer to its theoretical fulfilment; but b) there is reason to be more skeptical than usual about the economic signal coming from yield curves right now.

U.S. election math remains close

The U.S. presidential election is now a mere six weeks away. Since we last reported on the subject, Trump suffered a second assassination attempt and Harris won the final debate. Betting markets continue to argue that Harris’ small advantage is growing. PredictIt data indicates that Harris had a 52% chance of winning the election before the latest debate, a 54% chance immediately after the debate, and a 57% chance today (see next chart). Conversely, this means Trump has a 43% chance of winning.

Harris now leads in the presidential race

As of 09/23/2024. Based on prediction markets data and RBC GAM calculations. Sources: PredictIt, Macrobond, RBC GAM

But the amount of daylight suggested by PredictIt may be slightly exaggerated.  Many betting markets do not offer the ability to track the probabilities over time (hence the use of PredictIt for the longitudinal chart, above). Our synthesis of this data puts the Harris advantage at just 53.3% to 46.7% (see next chart).

To be clear, these are probabilities, not polling numbers – which are much closer. Furthermore, the election will ultimately come down to a mere tens of thousands of votes in a handful of swing states, rendering the outcome highly volatile. The point is that Harris is now the favourite, but either candidate could win without it being a major upset.

2024 U.S. election outlook continues to shift

As of 09/23/2024.Probabilities for presidential election measured as the median probability of winning from oddschecker. PolyMarket, PredictIt and RealClearPolitics (RCP). Probabilities for Senate and House are crowd forecast from Good Judgment. Sources: Good Judgment, oddschecker, Polymarket, PredictIt, RCP, Macrobond RBC GAM

From the standpoint of economic implications, the main considerations are tariffs, immigration and fiscal policy. Let us review each.

Tariffs

In an earlier #MacroMemo we laid out our thinking on the economic effect of proposed tariffs (see next table). In brief, the full Trump tariffs would have a significant and fairly large detrimental effect on economic growth. More likely, however, is that the tariff plan is only partially implemented, with the implication that the economic damage is more modest, to the tune of a few tenths of a percentage point subtracted from the level of output in the event of a Trump win. In contrast, we assume no material tariff changes under Harris, and thus a neutral effect.

Trump tariffs would have a fairly large detrimental effect

As at 08/05/2024. Deviation (in percentage) in level of GDP and CPI from normal trend after two years. Sources: Oxford Economics, RBC GAM

Immigration

U.S. immigration is hard to nail down to the extent that the Census Bureau has failed to fully capture the extent of illegal immigration into the country in recent years. A report from the Congressional Budget Office gives some sense as to the approximate numbers, but this is also only an approximation and the data is becoming stale.

With those caveats flagged, the rate of U.S. immigration appears to be substantially slowing already, even before either candidate is elected. Immigration may have been about 3.2 million people in 2023, is tracking around 2.4 million people in 2024, and we assume would slow to around 1.5 million people in 2025 under Harris versus 1.25 million under Trump. If the difference seems small, keep in mind that both parties are attempting to stem the flow of illegal immigration and that legal rulings may constrain some of the more aggressive strategies for constricting its flow. Factoring in a lower level of productivity among recently arrived undocumented immigrants, one might argue that immigration policy would subtract around 0.2% from annual economic growth in the event of a Harris win and subtract 0.3% in a Trump victory.

Fiscal policy

That leaves the remainder of fiscal policy – taxes, government spending and the like. This is where things get especially wooly. Estimates vary widely, with some researchers concluding that Harris would be more economically supportive than Trump in the short run, and others saying the opposite. Simultaneously, some models argue that both candidates will be economically stimulative in their policies, while others argue that the policies will be on the net economically negative.

Points of particular uncertainty include whether to consider every possible policy a candidate has ever mentioned, or to instead focus on those that they themselves give repeated and significant attention to. The details are often vague. Even if these can be ascertained, will the policy be implemented in full, or will it instead be implemented in part due to matters of practicality, political constraints or even legal constraints? There are also plain old disagreements over how much different policies could cost, what their economic effect would be, and how that economic effect could be distributed over the years ahead.

Given all of this, we stick to the broad economic contours rather than specific numbers. Trump definitely starts in a hole given his more economically damaging plans for tariffs and immigration policy. But he also proposes tax cuts where Harris proposes tax hikes. On the other hand, Harris proposes more new spending than Trump does.

With the acknowledgement that tweaks to the assumptions can yield quite a range of conclusions, we tentatively find that even after factoring in the tariff and immigration implications, a Trump presidency could be marginally more economically supportive in the short run than a Harris presidency (see next graphic). In other words, the economic burst generated by tax cuts and perhaps bolstered by a measure of deregulation more than offsets the Harris platform’s combination of higher taxes (on corporations and the wealthy) and more spending (on such items as a child tax credit, earned income tax credit and housing support), and what might be a reduction in inequality. This is not to say that the net effect is likely to be especially stimulative for either of them.

U.S. election macro and market thoughts

As of 09/19.2024. Sources: RBC GAM. For information purposes only.

This squares with the so-called Trump trade, which is that a Trump win is incrementally positive for equities (if less so for mega-cap stocks), positive for bond yields and (at least theoretically) positive for the U.S. dollar. Businesses certainly appear to have a preference, with 91% preferring the Republican candidate in a Wolfe Research survey that asks, “which presidential candidate is better for your industry?” Of course, businesses stand to benefit disproportionately from lower corporate taxes and deregulation. That, in turn, could plausibly stir the sort of animal spirits that can help an economy grow.

Of course, that’s the short-term economic story. Over the medium run, the tide could then reverse. The bulk of any additional Trump economic support in the short run comes from a larger deficit. That debt must be serviced later, and eventually even paid down. Furthermore, the prospect of a slower rate of immigration subtracts not just once but from growth each and every year that the policy is maintained. And the damage from tariffs could accrue more significantly over a period of multiple years. Thus, it could be that, several years later, a Trump win begins to drag on growth in a way that a Harris presidency would not.

Finally, let us step even further back and think for a moment. There is a fair chance that all of this analysis exaggerates the (non-tariff, non-immigration) economic differences of the two candidates. Neither has expressed much thought about balanced budgets or concern about the large deficit and public debt. Both would presumably spend (either in the form of tax cuts or government spending) as much as Congress will let them get away with.

Ultimately, the central determination of whether fiscal policy is net stimulative or net restrictive over the coming four years may be whether Washington is united or split across both parties. If united, the scope for fiscal stimulus exists on both sides. If divided, the scope for this is greatly diminished. At present, a divided Congress is more likely (refer back to the earlier bar chart), meaning we should not count on a big fiscal boost in the years ahead. That might be a good thing to the extent that the U.S. fiscal position is already challenging. This also means there is a risk that the Trump economic implications discussed above may be somewhat flattered relative to Harris’ more fiscally modest proposals.

Oil price weakness continues

Oil prices are fairly low at present, both in the context of the last year and relative to the norm of the past several years (see next chart). At just US$72 per West Texas barrel, this is one of the weaker points for oil since the worst of the pandemic lockdowns ended (when oil prices were temporarily negative, amazingly!).

Crude oil prices fall on demand concerns, despite geopolitical tensions

As of 09/19/2024. Sources: Macrobond, RBC GAM

The main story is one of fairly soft demand for oil. The combination of decelerating U.S. and Chinese economic growth plus a declining oil intensity in the economy are expected to keep demand roughly flat in 2025 (see next chart).

Oil demand growth to slow considerably in China

U.S. Energy Information Association (EIA) forecast from 2024 onward. Sources: Short-Term Energy Outlook, September 2024, EIA, Macrobond, RBC GAM

While supply has also been roughly steady, OPEC (Organization for Petroleum Exporting Countries) has been criticized for not cutting its output to revive prices. But the dynamic of the oil market is changing in ways that render such a decision more difficult to make today. A key factor is the rising clout of the U.S. shale oil sector, which now outproduces Saudi Arabia and Russia (see next chart). Its famous nimbleness means that any time oil prices rise, it is in the best position to rapidly capitalize on the increase. That means a reduction in OPEC production could simply transfer market share to the U.S., defeating the purpose.

U.S. is now the largest oil producer

As of May 2024. Sources: EIA, Macrobond, RBC GAM

A second – more speculative – structural development in the oil market is that as electric cars attempt to compete with traditional internal combustion engines, any period of time with higher oil prices could trigger an outsized and permanent shift away from fossil fuels. Although uptake is still slow in much of the developed world, it is skyrocketing in China. This is obviously something that OPEC would like to discourage, and so it is perhaps grudgingly satisfied with prices in the US$70 to US$90 range rather than wishing to see them back to triple digits.

Contrary to what one might imagine given low oil prices, oil demand is not actually running well below supply (see next chart), though that could transpire over the second half of next year. OECD (Organisation for Economic Co-operation and Development) crude oil inventories are also not especially distorted (see next chart). While inventories are a bit high by the norm of the past several decades, they are actually a bit low by the standard of the past decade.

Global oil deficit to switch to surplus in the near term

World production and consumption of petroleum and other liquid fuels. Sources: Short-Term Energy Outlook, Sept 2024, EIA, Macrobond, RBC GAM

Global crude inventory has been running above historical average

As of August 2024. Sources: EIA, Macrobond, RBC GAM

For all of this, we continue to flag the upside risk to oil prices that extends from the geopolitical realm. The conflict in the Middle East continues to escalate, and the war in Ukraine remains fraught with risk, particularly as Ukraine now operates on Russian soil and Russia has warned that it will view any use of U.S. arms in Russia as an attack by NATO. Suffice it to say, there are scenarios in which the availability of oil becomes more limited, though this is not our base-case outlook.

In the meantime, oil prices are lower. This is, on the net, good for global growth and especially for households. Thus, the global economy enjoys the joint benefit of lower interest rates and lower oil prices, which both point promisingly to continued economic growth over the coming year. The implications are more nuanced if not outright negative for major oil-producing countries, including Canada. The U.S. no longer enjoys the large benefit it once did, but we assume it is at worst a flat effect for overall GDP, and positive for the consumer.

Canadian wage mystery grows

Canadian hourly wage growth was a strong +5.0% year-over-year in August according to the Labour Force Survey (LFS). This is quite high relative to the U.S., where the equivalent wage metric is rising at just +3.8% YoY pace despite a stronger economy. The Canadian wage gains have also been unusually steady over the past two and a half years even as inflation has retreated and the unemployment rate has increased.

Why is this? One might speculate about Canada’s higher rate of unionization and the delayed impact of the higher cost of living on wages as contracts expire with a lag. But this is unlikely the whole story. To be honest, much of the wage strength is not just mysterious but counter-intuitive given that Canada has lately welcomed many unskilled temporary workers who one might imagine would earn low wages.

We don’t ultimately manage to crack the case, but can instead do something else nearly as useful, which is to point out that, for whatever reason, the LFS wage measure is at the extreme high end of Canadian wage estimates. Other wage figures are lower (see next chart). For instance, the Survey of Employment, Payroll and Hours (SEPH) says that wage growth is a more moderate +3.2-4.2% YoY, depending on whether one controls for the composition of workers or not. Other wage metrics are mostly in the 3-4% range, with the Bank of Canada’s wage-common metric currently sitting at just +3.0% YoY.

Canadian wage growth remains elevated

Indeed Wage Tracker and Labour Force Survey (LFS) as of August 2024. SEPH (Survey of Employment, Payroll and Hours) and wages & salaries as of June 2024. Wage-common Indicator and compensation per hour as of Q2 2024. Sources: Haver Analytics, RBC GAM

The point is that wage growth in Canada is probably not quite as fast as the most popular survey claims. The real number is probably at or below +4% rather than +5.0% YoY.

But this is not to say that Canada’s labour market therefore isn’t inflationary. Because of an atrocious productivity performance, unit labour costs in Canada are nevertheless rising fairly quickly, which means that companies are having to pay quite a lot more for labour relative to what workers are delivering (see next chart). A return to productivity growth would fix this in short order.

Canadian unit labour cost has been rising

As of Q2 2024. Shaded area represents recession. Sources: Statistics Canada, Haver Analytics, Macrobond, RBC GAM

China stockpiles commodities?

There have been rumours of China stockpiling commodities, with fears that this could reflect near-term Chinese military aspirations. But this is far from the only interpretation, and indeed the stockpiling is not especially extreme.

It is impossible to observe these stockpiles directly since they are a matter of national security. But we can see the country’s imports of various materials, and then make guesses as to whether the increase is in line with economic needs or beyond (see next chart).

Chinese imports of commodities have grown since the pandemic

As of August 2024. Sources: China General Administration of Customs (GAC), China Ministry of Finance, Macrobond, RBC GAM

While the various lines are trending upwards, their rise doesn’t seem especially extreme, and in fact all, other than copper, have actually gone sideways over the past year, presumably reflecting China’s economic weakness.

Oil imports have roughly doubled since a decade ago, but so has the size of the Chinese economy. There are reports of China adding up to a million barrels per day to its oil reserves, but the accumulation to date would still only cover 115 days of imports. Of course, oil doesn’t necessarily tell us much about China’s aspirations given its close relationship with Russia, which could provide ample oil in the event other sources were halted.

China definitely maintains large food stores. The U.S. Department of Agriculture forecasts that China will have 51% and 67% respectively of the world’s wheat and maize stocks by the end of this growing season. This is said to be enough to keep Chinese households supplied with these commodities for something in the realm of a year and a half. These are big numbers and would permit the country to survive a persistent outage. But this doesn’t suggest China is suddenly preparing for war: the country has long maintained a high level of food stockpiles. Its global share was even higher a few years ago. This is presumably in part due to its enormous population and in part to its history of famine.

China’s demand for metals is famously ravenous, with the country regularly importing half of the world’s base metals. Its copper demand in particular continues to rise incrementally, though iron ore demand has been roughly flat over the past several years despite having the most obvious connection to military uses. China’s cobalt purchases are expected to rise by a remarkable 87% in 2024 versus 2023, but the leading use of cobalt is in batteries and so it makes sense that a country leading the electric car charge would be ramping this up.

So what explains the stockpiling that is happening? While one can’t discount military aspirations, there are plenty of civilian justifications as well:

• Chinese stockpiles were considered low around 2018, so some of this stockpiling is just rebuilding to more normal levels.

• Commodity prices are presently fairly cheap by the standard of the last several years, making it a good time to load up on commodities that might be used later.

• China has good reason to fear that it could face significant additional tariffs after the U.S. election, so it makes sense to stockpile key materials in advance.

• The entire world learned some lessons over the past half decade about the importance of constructing more resilient supply chains. That includes having ample supplies of raw materials. China is no exception.

To be sure, one cannot rule out the risk that China does become more militarily assertive after the U.S. election, particularly in the event of a Trump presidency. Taiwan may receive less American support and China may feel less constrained in pursuing its strategic foreign-policy objectives if it is already being hammered by U.S. tariffs. Perhaps these additional commodity stockpiles have that possibility in mind. But the stockpiling has not really accelerated, it is not especially extreme in the context of China’s economic needs or its historical preferences. There are also solid economic reasons for increasing stockpiles at this juncture. As such, we don’t see anything especially concerning in the action.

-With contributions from Vivien Lee and Aaron Ma