~With contributions from Vivien Lee, Aaron Ma, Sheena Khan

December economic webcast

Our latest monthly economic webcast for December is now available:  Improving growth outlook.

Economic data trickling out

With U.S. government employees returning to work, statistics agencies are now compiling shutdown-delayed economic data. But it’s an achingly slow process. Aside from a jobs report that was more or less ready to go pre-shutdown, many key indicators won’t be released until the second half of December. Some have been canceled entirely.

September’s jobs data was a mixed bag. The 119,000-increase in payroll employment – the biggest gain in five months (see chart) – was well above expectations and the likely “breakeven” pace of job growth needed to keep the unemployment rate steady over time. But timelier alternative indicators suggest some of that gain has since been retraced. ADP’s new weekly jobs report argues the economy shed 54,000 private sector positions in the four weeks to November 7.

U.S. payroll growth is generally slowing

As of September 2025. Sources: U.S. Bureau of Labor Statistics (BLS), Macrobond, RBC GAM

Despite the solid payroll gain in September, the jobless rate increased to a cycle high of 4.4% as more Americans searched for work. We’ve noted that a low-churn labour market – with an unusual combination of low hiring and firing rates – has made conditions particularly challenging for new entrants. About two-thirds of the net increase in the unemployment rate so far this year has come from new and re-entrants into the labour force (see next chart).

New and re-entrants account for 2/3 of the increase in unemployment YTD

As of September 2025. Sources: BLS, RBC GAM

While the September jobs report is somewhat stale, it’s the only top-tier (official) employment data the Fed will get ahead of its December 10 rate decision. October’s payrolls will be combined with November’s and released on December 16. November’s household survey, which estimates the unemployment rate, will also be reported then, but October’s release has been canceled as the data can’t be collected retroactively.

That report would have shown a jump in the unemployment rate as furloughed government employees are counted as unemployed, but November’s report shouldn’t be impacted as workers returned to their jobs during the survey week. Furloughed employees will be counted as employed in the October payroll survey.

October’s Consumer Price Index (CPI) release has also been canceled. The U.S. Bureau of Labor Statistics (BLS) recalled furloughed workers to compile September’s inflation data, which was needed to index social security benefit payments. But it didn’t do so for October, and data collectors can’t retroactively track many prices. November’s CPI data has been rescheduled for December 18, and we’ll simply have to interpolate increases for October. The BLS says selective October price data will be released alongside November CPI.

Q3 gross domestic product (GDP) data won’t be released until December 23. The Federal Reserve Bank of Atlanta’s latest tracking is for a 3.9% annualized increase, although that’s based on incomplete information. It could be subject to revision as more GDP source data becomes available.

All told, the Fed doesn’t have much data to digest ahead of its December meeting. Chair Powell said this could make the committee cautious about further moves – likening the Fed’s approach to slowing down when driving in the fog. News of data delays caused the futures market to price out a December rate cut.

However, subsequent dovish comments from the Federal Open Market Committee’s (FOMC’s) vice chair caused pricing to swing back toward a rate cut. The comments were interpreted as an effort by Fed leadership to steer the market. In the absence of official data, alternative indicators that point to a softening labour market could be used to justify another “risk management” cut in December.

Our holistic sense is that the U.S. economy has likely cooled somewhat over the past few months. Still, some Fed voters are likely to dissent. Powell could strike a hawkish tone in his press conference to placate those who would have preferred a hold.

-JN

Corporate management moving on from tariffs

With more than 90% of S&P 500 companies having reported for Q3, it’s time to revisit key topics mentioned on earnings calls and how management discussions are evolving. The chart below shows the change in mentions of various keywords on a quarter-over-quarter and year-over-year basis.

As of 11/20/2025. Partial data for latest quarter. Quarters prorated by number of earnings calls. Sources: Bloomberg, RBC GAM

As a reminder, the Q2 earnings season saw business leaders start to move on from the tariff talk that dominated Q1. That trend continued in Q3 with tariffs seeing the largest quarter-over-quarter decline in mentions. But the topic still recorded the biggest increase in mentions relative to a year ago. That mirrors trade policy uncertainty (see next chart), which peaked in April and is now at its lowest level since last year’s election but remains historically elevated.

‘Tariff’ mentions are down alongside trade policy uncertainty

As of 11/23/2025. Sources: Bloomberg, RBC GAM

Q3 also saw a sizeable drop in mentions of margins, reversing the jump seen in Q2. Domestic businesses have absorbed a surprising share of tariff costs thus far – as much as half according to some estimates. Margins were under pressure in Q2, particularly in consumer discretionary and materials sectors. But that margin compression appears to have abated in Q3 (see next chart), consistent with less discussion on earnings calls.

Margin pressure generally eased in Q3

As of 11/24/2025. Sources: Bloomberg, RBC GAM

A few other trends stand out:

• There was more talk of returning capital to shareholders with mentions of dividends and buybacks rising.

• CapEx mentions also increased. This was likely helped by accelerated depreciation measures included in the One Big Beautiful Bill Act (OBBBA) tax bill passed at the start of the quarter.

• AI mentions were little changed from the previous quarter and still up significantly relative to a year ago. AI is second only to tariffs in that regard.

• Crypto discussions quieted somewhat but mentions remained elevated relative to last year.

• There was a slight increase in mentions of headwinds relative to tailwinds. Talk of an economic slowdown and job cuts picked up slightly in Q3.

Overall, less discussion of tariffs and margins (and less margin compression reported in Q3) suggests a key headwind for corporate America is beginning to ease. Yet tariff costs continue to rise. October’s government tariff revenue was up 11% from the Q3 average.

Someone has to foot the bill, and we have flagged the risk that pass-through to consumers increases over time as companies adjust to persistently higher tariff rates. That’s good for businesses but bad for households. We think some tariff-driven consumer price inflation and modest economic damage is still in the pipeline in the coming months.

-JN

Tallying economic tailwinds for 2026

The new year rapidly approaches and, from an economic perspective, an underwhelming 2025 draws to a close. As we gaze forward into 2026, there is the clear potential for a year of somewhat quicker GDP growth as a variety of new tailwinds join forces (see next chart) and a few prior headwinds fade.

Growth tailwinds for 2026

As of 12/01/2025. Source: RBC GAM

Monetary policy

Developed-world central banks have broadly been cutting rates over the past two years. North American central banks (and the Bank of England) resumed their easing cycles this fall and look to have the capacity for further cuts into 2026. Loosely, 100 basis points (bps) of rate cuts equals an extra percentage point of annual economic output within a few years, and a smaller 0.5% or so of extra growth in the short run.

Fiscal policy

The American One Big Beautiful Bill Act unlocked some additional economic growth in 2025, but the biggest boost should occur in 2026. Substantial tax refunds will be paid between February and April, elevating growth. It is important not to overstate this since a significant fraction of the purported boost is just the avoidance of the economic drag that would have resulted if expiring tax provisions had been allowed to sunset. But there is also a genuine addition to growth. We figure U.S. GDP growth could be on the order of 0.5% faster in 2026 as a result.

There is upside risk to this view given recent White House musings about $2,000 stimulus cheques for Americans.

Looking beyond the U.S., Canada has just announced a fairly stimulative federal budget. Germany is now undertaking its own fiscal expansion. Japan’s new Prime Minister is also stimulus minded. The UK and France are admittedly travelling in a more austerity-minded direction, but they have had to scale back such aspirations in the face of public discontent.

Stock market wealth effect

The U.S. stock market has managed extraordinary gains for several years, and international markets joined in over the past year. As a result, stock market investors have accumulated a great deal of additional wealth.

Only a small fraction of this marginal wealth is normally converted into spending in the short run. But even a relatively paltry 3% of trillions of dollars of gains constitute a significant sum of money. Some simple analysis would argue U.S. GDP growth could be up to 0.75% faster than otherwise in the near term.

Lower oil prices

Oil prices have continued to fall over the past year and are now at fairly tame levels by recent standards. This lowers global inflation and by extension increases purchasing power – an economic stimulant.

Weaker U.S. dollar

The weak U.S. dollar doesn’t help other countries, but it does help the U.S. via improved competitiveness. The currency has fallen by approximately 8% on a trade-weighted basis since the start of the year. That may overstate the economic boost since the currency was unusually strong at the start of the year. At a minimum, however, one can observe that relative to a counterfactual scenario in which the dollar had remained that strong across 2025, the economy should theoretically be about 0.25% to 0.75% larger by the end of 2026.

AI tailwind

Artificial intelligence provides a theoretical tailwind as well: something like a 0.25% boost to GDP growth in 2026 due to rising AI capital expenditures, and then an uncertain further amount from AI-driven productivity gains. Major tech companies seem to think they can grow their output without expanding their workforce (and in several cases shrinking it), suggesting such productivity advances may be imminent.

Other considerations

The U.S. and Canada may also enjoy a mild tailwind from World Cup-related tourism, and the U.S. may benefit from its 250^th anniversary celebrations.

At the same time, some headwinds that dominated headlines in 2025 are fading. The U.S. government shutdown has now ended (though there remains a risk of another one on February 1). Tariff damage is increasingly being absorbed such that it should exert less of a drag on 2026 growth.

To be sure, there is still some policy uncertainty. Lower income households are struggling. North American population growth is set to remain paltry. It is not a perfect picture. But the odds are good of better growth in 2026 than 2025.

The relative strength of each tailwind varies by country. But, on the whole, a large swath of the developed world should benefit:

• We accordingly forecast faster growth in 2026.

• We maintain above-consensus GDP forecasts for each of the U.S., Eurozone and Japan.

• U.S. Q4/Q4 GDP growth should accelerate from 1.7% in 2025 to 2.3% in 2026.

• Growth should accelerate from 0.3% to 1.7% for Canada, from 1.1% to 1.6% for the Eurozone, and from 0.6% to 0.9% for Japan.

• The UK is an exception, with a steady growth rate of 1.2% anticipated for both years.

-EL

Non-recessionary rate cutting cycles

The Fed looks set to deliver a third consecutive 25 bps rate cut in December. That would bring total easing for the current cycle, which began in September 2024, to 175 bps. It’s been 40 years since the Fed lowered rates by that much outside of a recession and three decades since the last soft-landing easing cycle.

Non-recessionary rate cutting cycles are somewhat rare. Over the past 70 years we count just six (only a few of which can really be called soft landings, in our view). Compare that with 10 recessionary easing cycles (see next chart).

Non-recessionary easing cycles are relatively rare

As at 11/27/2025. Effective fed funds rate shown prior to 1971. Sources: Bloomberg, RBC GAM

We expect the current cycle will add to the tally of soft landings. The U.S. economy has proven surprisingly resilient to tariffs, and the consensus odds of a U.S. recession over the next year have eased to 30% from 40% six months ago.

Our own subjective probability of a downturn within the next year is slightly lower. As noted above, we see a number of tailwinds supporting stronger growth in 2026. As such, it’s worth revisiting past non-recessionary episodes for hints – and warnings – as to how the current cycle might evolve.

1966-67: The risk of cutting too soon

The Fed cut its policy rate by 200 bps in late 1966 and early 1967 to address a credit crunch brought on by earlier tightening and changes in banking regulation. Falling headline inflation helped make the case for unwinding past hikes, but core inflation remained sticky amid stimulative fiscal policy. The Fed was back to raising rates by late 1967 but inflation continued to rise until a combination of fiscal and monetary policy tightening led to the 1969-70 recession.

1971: Sticky inflation and political pressure

Following the 1969-70 recession, the Fed was dealing with a challenging combination of sticky inflation and still-elevated unemployment. It raised its policy rate by 200 bps in the first half of 1971 but unwound those hikes later in the year amid intense political pressure from the Nixon administration (as discussed in a previous #MacroMemo). The economy accelerated and the Fed reversed course again with rate hikes in 1972, but unemployment continued to fall. The removal of wage and price controls and higher oil prices worsened inflation, which eventually reached double digits. So did the Fed’s policy rate, and a recession followed in 1973-75.

1984-86: Hard-earned credibility pays off

A strong recovery following back-to-back recessions in the early-1980s necessitated yet another foray into double digit policy rates. A manufacturing slowdown followed but the economy avoided recession. The Fed was able to cut rates roughly in half (albeit to a still high 6%) between 1984-86 while inflation and unemployment stabilized. The central bank’s newly established inflation-fighting credibility, limited leverage buildup heading into the rate cutting cycle, and a productivity recovery helped stick the soft landing.

1995-96: Fine tuning for a soft landing

The Fed lifted its policy rate to 6% in early-1995 from 3% a year earlier to keep the economy from overheating. That held inflation in check but caused growth to slow and unemployment started to tick higher. The Fed recalibrated monetary policy with 75 bps of cuts in 1995-96. That easing, combined with tech-driven productivity gains, helped reinvigorate the economy. Inflation remained contained thanks in part to a strengthening U.S. dollar and fiscal consolidation.

1998: Insurance cuts amid an economic boom

A productivity boom continued to power the U.S. economy in the second half of the 1990s, while a strong U.S. dollar and lower oil prices drove inflation to multi-decade lows. That gave the Fed scope to deliver 75 bps of insurance cuts in the second half of 1998 to buffer against a series of financial crises abroad (Asia in 1997 and Russia in 1998) and the collapse of Long-Term Capital Management at home. The Fed returned to tightening mode less than a year later, hiking by 175 bps by mid-2000. But the tech bubble was in full swing by then and its eventual collapse caused the 2001 recession.

2019: Mid-cycle adjustment upended by Covid

The Fed delivered yet another 75 bps of mid-cycle rate cuts in 2019 amid low inflation, slowing global growth, and a domestic manufacturing slowdown exacerbated by the U.S.-China trade war. The unemployment rate remained at multi-decade lows and growth started to pick up, but the COVID pandemic upended what might have been another soft landing.

Putting the current cycle in context

Half of those non-recessionary easing cycles were quite limited in scope with the Fed lowering its policy rate by just 75 bps. It has already cut by twice as much in the current cycle. While all six occurred outside of a recession, four were followed by a downturn within three years of the first rate cut.

Easing cycles in the mid-80s and mid-90s were arguably the most successful with the economy enjoying several years of robust growth afterwards. In both cases, leverage cycles were in their early stages and rising productivity helped sustain the expansion (see chart). There are encouraging similarities today with relatively low private sector debt (notwithstanding concerns in the private equity space) and what could be the early stages of an AI-driven productivity resurgence.

Low leverage, strong productivity growth supported soft landings in the mid-80s and mid-90s

As of Q1 2025. Credit to GDP gap measures deviation from trend. Sources: Bank for International Settlements (BIS), BLS, Macrobond, RBC GAM

Those cycles also saw the Fed take advantage of inflation-fighting credibility established in earlier tightening campaigns. Whether the Fed has as much credibility today is debatable. It took aggressive (but in hindsight belated) action to rein in inflation post-pandemic, but never fully returned inflation to its 2% target. While inflation expectations were trending lower heading into the mid-80s and mid-90s easing cycles, expectations remain elevated today amid still-high prices and tariff-driven goods inflation (see next chart).

Inflation expectations were low or declining heading into mid-80s and mid-90s easing cycles

As of November 2025. Sources: University of Michigan, Macrobond, RBC GAM

The 1967-68 and 1971 rate cutting cycles demonstrate the perils of easing monetary policy before inflation has been brought to heel. In both cases, rate cuts were unwound shortly after, but inflation continued to rise. Aggressive policy tightening ensued, and recessions eventually followed. Stimulative fiscal policy in 1967-68 and political pressure in 1971 have concerning parallels today.

The current AI CapEx boom and rich equity market valuations draw comparisons with the 1990s dotcom bubble, but it’s hard to say whether the current situation is closer to 1995 or 1998. How long the AI cycle lasts may very well determine whether today’s non-recessionary rate cuts are followed by a sustained soft landing, as in the mid-80s and mid-90s, or simply buy a year or two before the next downturn.

Recent equity market wobbles amid shifting rate cut expectations illustrate how important Fed policy is for maintaining investor enthusiasm. If sticky inflation forces the Fed to significantly under-deliver on rate cuts or pivot back toward hikes, equity markets and the soft landing would be at risk. At this stage, we think the 3-4 rate cuts priced in over the next year look fairly reasonable, with less restrictive monetary policy helping to keep the expansion going in 2026.

So far, the S&P 500’s 22% gain since the first rate cut of this cycle is in line with the median performance of past non-recessionary easing cycles (see next chart). Non-recessionary easing cycles have seen equities continue to perform well two years after the first cut – the median gain is around 37%. Unsurprisingly, this is far better than during recessionary rate cutting cycles, when equities have tended to be roughly flat at this point in the process.

S&P 500 before and after the first Fed rate cut

As of 11/28/2025. Source: RBC GAM

-JN

U.S. Social Security fears

Amid broad concerns about the U.S. fiscal position, there are particularly acute worries about the sustainability of entitlement programs such as Social Security – the primary government income supplement for retired people. Earlier this year, the Social Security Board of Trustees estimated that the Social Security Trust Fund will be exhausted in 2033 (see next chart).

Social Security Trust Fund will be depleted by the first quarter of 2033

Source: The 2025 Old-Age, Survivors, and Disability Insurance (OASDI) Trustees Report, Social Security Administration (SSA) .gov, RBC GAM

Let’s dig into the details.

By way of background, Social Security is an important program. It paid benefits of US$1.47 trillion in calendar year 2024, serving a remarkable 68 million beneficiaries – nearly 20% of the population (and indirectly benefiting other household members). The average annual payment is US$22,880 per year, providing about 40% of the retirement income for the average retired beneficiary. For lower income recipients, Social Security payments constitute 70-90% of total income, and for many in the bottom quartile it represents the entirety of their income.

Unlike many pensions, Social Security is not fully funded. It is primarily a pay-as-you-go system in which current workers pay for current retirees.

This arrangement is becoming problematic because fertility rates are declining and longevity is increasing. As the working-age population declines on a relative basis, the U.S. dependency ratio goes up (see next chart). Put simply, there are now too many retirees counting on too few workers to pay the 12.4% Federal Insurance Contributions Act (FICA) payroll tax (the burden is shared equally between employers and employees).

Economic burden on workers is rising

Working population to total population defined as population ages 20-64 to total population. Dependency ratio is the ratio of population ages 0-14 and over 65 to working-age population (ages 15-65). Sources: World Population Prospects 2024, RBC GAM

What about the trust fund we mentioned earlier? Admittedly, the Social Security program isn’t completely pay-as-you-go. From the mid-1980s until the late 2000s, the population was relatively young and so the payroll tax was collecting more money than was paid out to retirees. That allowed the aforementioned trust fund to form and grow. But the situation is now reversed. The fund has been shrinking since 2010 and is now being rapidly depleted.

What happens when the trust fund runs out in 2033? Current law dictates that Social Security payments can only be made from payroll tax revenues plus the fund’s reserves. Without the latter, Social Security payments would have to be cut to just 77% of the normal level. By the end of the century, demographic forecasts indicate the payment would decline to just 70% of normal.

What options does the government have?

• The government could legislate a payroll tax increase. Raising it from the current 12.4% to 14.8% would probably stabilize Social Security.

• The retirement age could be increased. The government has done this before, minimizing backlash by adjusting the age in small increments over a lengthy period of time.

• The government could loosen immigration policy, increasing the number of working-age adults relative to retirees (though this is something of a treadmill, as the same trick must be applied repeatedly to keep the worker-to-retiree ratio elevated).

• The government could simply change the law and allow general government revenue to fund Social Security. Of course, the deficit would rise even further if this were adopted.

• The government could just allow Social Security entitlements to shrink. This could be as simple as an across-the-board haircut as per the 77% number above, or something more nuanced such as reducing benefits for those with the highest incomes (at the risk of eroding nearly universal public support for the program) or making the mechanics of inflation and wage indexation less generous (which would be less visible to recipients).

Given the current political climate, the first three options seem less likely than the latter two. But the political climate could yet change over the next eight years, meaning all five are genuine options. We do not presently expect a sharp cut to Social Security payments; instead, we imagine that relatively subtle benefit adjustments may be made alongside a decision to allow Social Security to tap general government funds.

-EL

Wage growth takes heighted importance

With very little hiring happening in the U.S. due to minimal population growth, an unusually large fraction of consumer spending growth is contingent on rising wages. So what is happening on that front?

Nominal wage growth remains unusually robust, up 3.8% over the past year (see next chart). But this exaggerates its strength, for three reasons:

1. The rate of growth is gradually slowing.

2. Measures of expected wage growth are declining (see subsequent chart)

3. Most importantly, inflation-adjusted wage growth is a much more sedate +0.8% year-over-year.

In other words, wage growth only looks strong because inflation is high. Worker purchasing power is rising considerably more slowly, though approximately in line with the norm of the past two decades.

U.S. wage growth is slowing

As of September 2025. Sources: BLS, Macrobond, RBC GAM

Wage pressure in U.S. has eased

Atlanta Fed Wage Growth Tracker as of August 2025, wage expectations as of November 2025. Wage Pressure Composite constructed using business intentions to raise wages. Shaded area represents recession. Sources: Macrobond, RBC GAM

It makes sense that real wage growth is just OK. The unemployment rate has been inching higher over the past few years and is no longer strikingly low (see next chart). That argues worker bargaining power isn’t exceptionally strong right now. We presume the unemployment rate moves slightly higher through early 2026 before stabilizing and then potentially beginning to reverse.

U.S. unemployment rate is moving higher

As of September 2025. Sources: BLS, Macrobond, RBC GAM

There are also some interesting things going on beneath the surface. We investigate three different aspects of this: the extent to which the cost of employee benefits are outpacing wages, how wage growth is varying by income level, and whether a decline in undocumented workers is boosting pay in affected sectors.

Employee benefits

The conventional wisdom is that the growth in the cost of employee benefits for employers has greatly exceeded the growth in wages. That was true in the 2000s and over the first half of the 2010s. But, in actual fact, wage growth has slightly outpaced benefits growth over the past decade (see next chart). This doesn’t make the cost of funding health insurance and myriad other programs any cheaper, but it hasn’t been a disproportionate driver of labour costs.

U.S. wage growth has slightly outpaced benefits growth

As of Q2 2025. Sources: BLS, Macrobond, RBC GAM

K-shaped wages

Further to our earlier work on the K-shaped economy, wages are also showing something of a K-shape. This is to say, the wages of top-quartile workers are now outgrowing those of bottom-quartile workers (see next chart). That was not actually the norm over the past two decades – lower-income workers experienced faster wage growth over the bulk of that period, and massively outperformed the top quartile around the pandemic.

But the pattern has flipped over the past year. That adds to the suffering of lower income households, who were already grappling with worse outcomes from tariffs and the OBBBA, who are more exposed to slow hiring, and who do not benefit as much as wealthier cohorts from a rising stock market.

Lower-income wages are growing more slowly than higher-income

As of 09/11/2025. Sources: Federal Reserve Bank of Atlanta, RBC GAM

Undocumented workers

The number of illegal immigrants entering the U.S. appears to have declined sharply over the past two years. Increased efforts to deport undocumented residents in 2025 may also be discouraging some within the U.S. from seeking employment, as we articulated in a September 23 MacroMemo.

In theory, one would expect wages to accelerate in sectors where these undocumented residents have traditionally been most concentrated, since there should suddenly be a worker shortage. Curiously, that is not yet visible in the data (see next two charts).

Wage growth in accommodation and food services has decelerated in 2025 and is currently running more slowly than overall wages. Wage growth in construction is running faster than the economy-wide average and has picked up slightly since the spring, but is still notably weaker than the norm of the past few years.

Wage growth remains weak in undocumented worker-intensive industries

As of September 2025. Sources: BLS, Macrobond, RBC GAM

Similarly, overall wage growth in cities with the highest concentration of undocumented residents is not substantially faster than in cities with the lowest such concentration.

Wage growth similar in cities with many or few undocumented residents

As of August 2025. Sources: BLS, Macrobond, RBC GAM

Perhaps the effect of fewer undocumented workers will become more visible over time, but it is hard to find in the wage data so far.

-EL

Capital still flowing into U.S.

There has been a great deal of talk about declining U.S. exceptionalism, concerns about the U.S. fiscal position, rising American polarization, worries about the politicization of the Fed, the weaponization of the dollar, declining international trust in the country, and concerns about U.S. equity valuations. But the big question is whether foreign economic actors are putting their money where their mouth is and actually removing significant funds from the country.

The surprising answer is a succinct “no”, at least at the aggregate level.

Admittedly, much of this is a mechanical identity. The U.S. continues to run a large current account deficit (see next chart), which means that the country spends more than it earns. In turn, the U.S. is definitionally a net borrower against the rest of the world. By extension, the rest of the world is steadily accumulating an ownership stake in U.S. assets so long as this situation remains in place.

Sure, foreign investors have free will, but the market is clearing at current interest rates and equity valuations.

U.S. continues to run a large current account balance

As of Q2 2025. Sources: U.S. Bureau of Economic Analysis (BEA), Macrobond, RBC GAM

Let’s take a look at a few aspects of this.

1. Foreigners continue to be net buyers of U.S. securities and have even significantly increased their buying clip over the past year and a half (see next chart).

Foreign buying of U.S. securities remains high

As of September 2025. Sources: U.S. Department of Treasury, Macrobond, RBC GAM

1. While foreigners have been fairly selective about U.S. stocks over the years, they are currently the most enthusiastic buyers of the asset class in the history of the series (see next chart).

Foreign buying of U.S. equities reaches all-time highs

As of September 2025. Sources: U.S. Department of Treasury, Macrobond, RBC GAM

1. Foreign direct investment into the U.S. also continues to be a net positive (see next chart). This is perhaps less surprising given the incentives provided by tariffs plus political pressure for international businesses to expand their U.S. operations.

Foreign direct investment into the U.S. surges

As of Q2 2025. Sources: BEA, Macrobond, RBC GAM

1. Finally, foreign official entities (reserve managers and their ilk) are not actually dumping U.S. Treasuries but have instead simply stopped expanding their holdings over the past 15 years (see next chart). Meanwhile, foreign private investors continue to acquire the asset class, so there is still a considerable market for American public debt and that hasn’t obviously been interrupted in recent months despite mounting concern about the country’s fiscal trajectory.

Private foreign investors of U.S. Treasuries are buying while official entities pause

As of September 2025. Sources: U.S. Department of Treasury, Macrobond, RBC GAM

In conclusion, the “sell U.S.” trade remains more talk than action at this juncture.

-EL

Canadian economy weaker than it looks

Canada has been spared from the U.S. shutdown-related data fog. But economic indicators have been particularly noisy of late, making it equally difficult to determine the direction of travel.

Q3 GDP was significantly stronger than expected with the economy expanding at a 2.6% annualized pace – a far cry from the earlier 0.4% flash estimate thanks to upward revisions to monthly GDP figures. The Q3 gain more than retraced a 1.8% annualized decline in the prior quarter and means the economy avoided a technical recession (back-to-back quarterly declines).

But looking under the hood, the composition of activity was much poorer in Q3. Final domestic demand (GDP excluding swings in trade and inventory investment) was flat following a 3.5% annualized gain in the prior quarter. Consumer spending and business investment both declined, while housing activity picked up and defense spending increased. Somewhat bizarrely, it was a sizeable pullback in imports (which are subtracted from growth as they don’t reflect domestic production) that drove headline GDP higher in Q3.

Canadian GDP stronger than expected in Q3 but composition suggests weakness in the economy

As of 11/28/2025. Sources: Statistics Canada, RBC GAM

Looking through the quarterly noise, Canada’s economy has grown at a 1% average annualized pace year-to-date. That is below its potential growth rate or speed limit, indicating some increase in slack in 2025. Domestic demand grew at a similar pace, led by consumer spending which has benefited from positive wealth effects, tax cuts and lower interest rates. Meanwhile, business investment has been weighed down by tariffs and trade policy uncertainty. The story makes sense, but the quarterly pattern is questionable.

Growth likely remained subdued in Q4 with the October flash estimate pointing to a 0.3% monthly decline. While it’s early days – and revision-prone monthly readings need to be taken with a grain of salt – that suggests downside risk to the Bank of Canada’s forecast for a 1% annualized gain.

It’s not just GDP data that’s throwing analysts for a loop. As we noted in our previous #MacroMemo, Canadian jobs data showed a surprising pickup in hiring in September and October with more than 60,000 jobs added in each month. However, the separate payroll survey calls that into question with 58,000 jobs reportedly lost in September (the data is released with a lag, so no October reading is yet available). Year to date, the timelier household survey shows 165,000 jobs have been added, while payrolls point to a 37,000 net loss.

Separate Canadian jobs reports tell two different stories

As of October 2025. Sources: Statistics Canada, Macrobond, RBC GAM

The true answer is likely somewhere in between, although we lean toward the payroll survey’s more subdued figure. Population smoothing in the household survey is likely overstating job gains. A significant slowdown in immigration has reduced the breakeven rate for hiring – that is, the number of jobs that need to be added to keep unemployment steady. With the jobless rate having increased year-to-date and GDP growing below potential, it would be hard to believe the economy has added tens of thousands of jobs per month.

We don’t believe Canada’s economy has accelerated to the extent Q3 GDP and (household survey) jobs data would suggest. But recent stabilization in the unemployment rate is encouraging and it seems fair to say a recession has been avoided thus far. We think 2025 will go down as a year of economic underperformance, but prospects for 2026 look somewhat brighter.

Just as the Fed discussed moving cautiously amid shutdown-related data fog, noisy Canadian indicators give the Bank of Canada reason to remain on hold in the near-term, hoping the economy’s trend becomes clearer with fresh data. We still wouldn’t rule out another rate cut if recent upside surprises prove to be a head fake.

-JN