Questions? We've Got Answers 

Q1. Are energy markets reaching a choke point with the U.S.-Iran conflict?

Not yet, but the risks are rising. A combination of rerouted oil supply, drawdowns from commercial and strategic stockpiles, reduced Chinese imports, and rising production outside the Gulf region has so far prevented regional disruptions from spilling over into a broader global energy crisis (Chart 1). 

The risks ahead are twofold. The first is the disruption of both major shipping routes out of the Gulf region. The Strait of Hormuz and Bab el-Mandeb are critical arteries for global trade in crude oil, refined products, and liquefied natural gas. Markets can typically manage disruptions in one location, but concurrent and persistent disruptions across multiple routes would be far more difficult to absorb. 

The second is the erosion of buffers that have thus far stabilized markets. Inventory drawdown has limits, alternative trade routes are more costly, and spare production capacity is finite. Importantly, the first signs of strain often emerge not in crude oil itself, but in refined products. Distillates, such as diesel and jet fuel, are particularly sensitive to transportation bottlenecks and supply chain disruptions. As a result, prolonged trade disruptions could tighten fuel markets and push up costs even if global oil production remains largely intact. Any future shock would therefore become increasingly visible in fuel prices, inflation, and economic growth.

Our base case remains that the U.S. and Iran ultimately return to diplomatic negotiations and that current disruptions gradually ease over the coming months, albeit with some volatility. Under that scenario, most of the lost Gulf supply returns to market by next year, and oil prices moderate as geopolitical risk premiums fade. We forecast WTI oil prices to average roughly $84/bbl in Q3, and through year-end amid ongoing shipping risks and tight market conditions. Prices are expected to ease into the low-to-mid $70s/bbl range next year as supply conditions improve.

Energy markets may not be at a breaking point, but they are becoming more vulnerable to further shocks. The longer tensions persist, the more strain is placed on the inventories, spare capacity, and alternative trade routes that have stabilized markets so far. If negotiations fail and disruptions remain elevated into the fourth quarter, the global energy system would become increasingly exposed to a more lasting supply shock. 

Q2. Is the AI investment cycle creating financial imbalances?

The AI investment cycle has created pockets of financial vulnerability, but it has not produced the leverage or funding stresses typically associated with a systemic financial crisis. The buildout is supporting U.S. growth, but also concentrating market valuations and financing needs around a small group of firms. On the surface, strong operating cash flow and relatively low leverage limit the risk of a systemic event. However, the growing use of corporate debt, private credit and off-balance-sheet financing could amplify the ripple effects of a market correction if earnings fail to keep pace with the scale of investment.

The AI capex cycle remains in its expansion phase, with capital commitments and financing needs continuing to accelerate¹. Hyperscaler technology firms are expected to increase spending to roughly $780 billion this year and $1 trillion in 2027²,³. The current AI capex cycle is on pace to surpass every major infrastructure investment cycle since the canal boom of the 1830s⁴.  Large investment alone does not imply a financial imbalance, but the economic and financial consequences become larger if revenues fail to justify the scale of spending.

What makes this cycle different from previous infrastructure booms is the nature of the underlying assets. AI hardware depreciates within three to five years, compared with the multi-decade lives of railways or fiber-optic networks. Rapid technological change shortens the period over which investors can earn a return and increases residual-value risk. This leaves both borrowers and lenders more exposed to slower-than-expected revenue growth or excess capacity.

The AI investment boom has also increased equity-market concentration. The five largest AI related firms now account for roughly 24% of S&P 500 market capitalization, more than double their average share prior to the AI era⁵.   As a result, a reassessment of AI-related earnings expectations could have an outsized impact on broader markets, financial conditions, and investment.

Financing trends warrant attention as well. Strong operating cash flow has allowed the hyperscalers to finance much of the buildout without relying heavily on debt. More recently, however, firms have supplemented internal funding with corporate debt and private lending⁶. Private lending to the AI sector grew from near-zero to roughly $200 billion in 2025 and is generally less transparent than public equity or debt markets⁷.   Off-balance-sheet financing and interconnected lending relationships may also create indirect exposure for banks and other financial institutions.

These financing trends echo the telecom boom of the late 1990s, when supplier credit helped sustain demand until financing conditions tightened, at which point both equipment orders and credit performance deteriorated. Today's AI ecosystem is financially stronger. However, growing financing links among chipmakers, cloud providers, and AI developers make it harder to distinguish between end-user demand and investment-driven demand.

Most AI related firms remain resilient, but pockets of vulnerability are emerging among firms with high investment intensity and dependence on external financing (Chart 2). These exposures do not yet threaten financial stability, but a shift in market sentiment could trigger a sharper market correction and a broader tightening in financial conditions.

Q3. Why have global bond yields risen?

The recent rise in global bond yields reflects a more challenging policy backdrop. Financial markets have shifted from expecting interest rate cuts to pricing some risk of hikes, while also demanding greater compensation for uncertainty and duration risk. Inflation expectations pushed yields higher earlier in the year, but they have played a smaller role more recently (Chart 3). 

Long-term yields can be broken into three components: expectations for short-term interest rates, expected inflation, and the extra return investors require to hold a longer-term bond (rather than a series of short-term securities), known as the term premium.  The recent rise in yields has been driven primarily by changes in expected policy rates and the term premium.

Since June, the biggest shift has been in policy rate expectations. Markets have become less confident that central banks, particularly the Federal Reserve, will be able to deliver as much easing as previously expected. In the U.S., shorter-term Treasury yields and overnight interest rate swaps have risen more than longer-term yields. The pattern is consistent with cyclical policy repricing rather than a reassessment of where policy rates are likely to settle over the longer run.

The term premium has also moved higher. We estimate it has increased by roughly 20 basis points since U.S. Treasury yields reached a low in February and now stands around 90 basis points, above its more typical 60 to 70 basis point range. The increase has occurred alongside shifting expectations for Fed policy, oil prices, and economic growth. The speed of the move, combined with the absence of a major change in fiscal policy, suggest investors are demanding greater compensation for uncertainty when holding longer-dated bonds.

Importantly, this is not just a U.S. phenomenon. Term premia have risen across most advanced economies, including Canada. Similar dynamics have been evident in Europe, where long-term yields have risen alongside changing fiscal, growth and policy expectations. The result has been a broad-based increase in long-term borrowing costs across developed markets.

Taken together, the recent rise in global bond yields reflects a repricing of the expected path for monetary policy, led by the Federal Reserve, and a higher term premium across advanced economies. While expectations for further rate hikes should diminish as inflation moves closer to target, the rise in term premia may prove more persistent. Investors are demanding greater compensation for holding long-dated bonds amid growing uncertainty around the policy and geopolitical landscape, fiscal sustainability, and the institutional arrangements that have long anchored the global economy.

Q4. Are markets wrong to expect rate hikes from the Federal Reserve and Bank of Canada?

For the U.S., not necessarily. Markets are currently pricing in a 35% probability of a rate hike in September and are fully pricing in a hike by year-end. We view this as reasonable given lingering inflation concerns and uncertainty around how quickly the price impacts from past supply shocks will fade in the second half of this year. At the same time, the AI demand push has emerged as a new source of inflationary pressure, further clouding the outlook. 

Fortunately, recent inflation data has been more encouraging. Policymakers will see one more CPI report before the September meeting, and provided August inflation also comes in on the milder side, the Fed is likely to remain on the sidelines while maintaining a hawkish stance. Officials are likely to reiterate that each meeting is "live" and data dependent. Given Chair Warsh's leaner communication approach, that could mean greater volatility in the months ahead, as market participants closely parse each data point for clues on the Fed's next move.

Even so, the threshold for additional tightening has fallen. A growing number of Fed officials have expressed little willingness to continue looking through past price shocks, particularly with inflation having remained above the Fed's 2% target for over five years. Unless there are meaningful signs of progress in the months ahead, policymakers may feel compelled to act to preserve credibility. This is not our base case, but it is admittedly a close call given the Fed's recent hawkish tilt (Chart 4). 

The situation in Canada is different. Core inflation measures are mostly running below 2% year-on-year and, despite some much-welcome momentum coming in the second quarter, the unemployment rate remains elevated at 6.4%, suggesting continued slack in the economy. 

That said, after a soft patch in the winter months, the economy is finding its feet. Second quarter growth is likely to register north of 3%, while the unemployment rate fell to a two-year low in July. The direction is encouraging, but the economy remains in recovery mode. U.S. tariffs and the associated hit to confidence have left output roughly 1% below where it otherwise would have been, and the second-quarter bounce-back could fade quickly should trade disruptions re-emerge (see Question 6). With core inflation remaining well behaved despite volatility in energy prices, we expect the Bank of Canada to stay on hold (Chart 4). 

Q5. What could undermine the U.S. exceptionalism story and bring it back toward global peers?

The American economy is on track to outpace its peers for a fourth consecutive year, and by a healthy margin. U.S. growth is expected to reach 2.2% this year, while other G7 economies are all tracking sub-1% (Chart 5). While U.S. exceptionalism is not a new phenomenon, it has become more pronounced in recent years. Stronger productivity growth, AI-led capital spending, and more expansionary fiscal policy have all contributed. Earlier in the cycle, faster population and labor force growth also provided support, but that tailwind has faded amid tighter immigration policies. 

Perhaps the biggest downside risk to the exceptionalism narrative is the possibility that AI investments fail to deliver on expectations. Markets are assuming that the enormous AI data center buildout will translate into meaningful productivity gains. If adoption proves slower or efficiency gains fall short, business investment could weaken and equity valuations could come under pressure. The effects would spill over to households through slower income growth and weaker wealth effects, ultimately weighing on consumption. As addressed in Question 2, financial conditions represent another important risk. 

Beyond AI, the U.S. economy has benefited from unusually large fiscal support in the post-pandemic period, with annual government deficits far exceeding those of its peers. This has resulted in a significant rise in debt-servicing costs and a less sustainable fiscal trajectory. So far, investors have downplayed these risks, in part reflecting the potential for AI to meaningfully lift growth and improve the long-term fiscal outlook. However, a sustained rise in term premia could push longer-term interest rates meaningfully higher, crowding out investment, slowing AI adoption, and potentially forcing painful fiscal consolidation (see paper). 

While these risks warrant attention, America’s sustained productivity advantage remains its core strength and primary reason it continues to outperform its global peers. There is little to suggest this dynamic will reverse anytime soon. If anything, the balance of risks suggests that the AI buildout could further widen the gap.

Q6. How do the latest U.S. tariffs impact the economic outlook?

When the U.S. administration's temporary 10% global tariff imposed under Section 122 of the Trade Act of 1974 expired on July 24, it was replaced by a new tariff regime established under Section 301 of the same legislation. The administration justified the move on the grounds that trading partners had failed to adequately block imports produced with forced labor. The new tariffs range from 10% to 12.5% and apply to 60 trading partners. However, the administration has maintained broad product-specific exemptions for goods such as oil and gas, fertilizers, and other products covered by prior trade agreements. 

The changes preserve the administration's "broad tariff wall" and provide greater durability than the IEEPA tariffs, as they are more difficult to overturn. However, they do not fundamentally change the overall tariff rate. Estimates from the Yale Budget Lab suggest the effective tariff rate following implementation of Section 301 measures is roughly 11%, largely unchanged from its pre-July 24 level. From a macroeconomic perspective, little has changed. Most of the drag from tariffs, stemming from the uncertainty created by the stop-and-start implementation approach, is already baked in. While much of the inflationary impact has faded, firms continue to cite tariffs as a reason for raising prices, suggesting tariff pass-through will remain a modest source of inflation pressure through the second half of 2026.

For Canada, the Section 301 tariffs largely preserve the status quo. Under the new rules, a 10% tariff rate applies to goods that are not USMCA-compliant goods and are not covered by existing Section 232 tariffs. This is lower than the 35% rate previously applied under the IEEPA statute, and, on its own, does not materially change the economic outlook. 

The more significant risk is the proposed 50% tariff rate under Section 338 of the Smoot-Hawley Tariff Act of 1930, affecting roughly 5% of Canadian exports, including those covered by the USMCA. Canadian and U.S. negotiators have made progress in recent days to avert their implementation, pushing the deadline for a deal back to August 22 (from August 19th). 

Should these measures come into force, they would impose a drag of roughly 0.3 percentage points on Canadian economic growth over the next year (Chart 6). Importantly, this estimate assumes no Canadian retaliation, no further escalation in trade tensions, and no renewed deterioration in business confidence. Any of these developments would amplify the downside risks. 

At the same time, there remains scope for a more constructive outcome. A détente before the August 22 deadline that brings relief on the Section 232 tariffs affecting metal products and other goods, or a clearer path toward USMCA renewal would represent upside risks to growth. For the time being, as the parameters of any new agreement are unknown, we assume no change to the tariff regime and expect Canadian growth to register 0.9% in 2026, and 1.8% in 2027.  

Q7. How is the job market faring amid these cross currents?

The U.S. labor market remains in a low hire, low fire equilibrium. While hiring activity picked up earlier in the year, it likely reflected catch-up hiring following a very weak pace of job creation through most of 2025. More recently, job growth has slowed, with the three- and six-month averages through July running at just 20k and 44k, respectively. This is broadly consistent with estimates of the breakeven pace of job growth. 

At the same time, the unemployment rate has drifted lower and now sits at a thirteen-month low of 4.1%. On the surface, this is encouraging. However, the decline has been driven by a sharp drop in labor force participation, largely among prime-age (25-54) workers. Reasons for the pullback likely include softer labor demand, rising worker discouragement, and weaker immigration. Rising household financial cushions may also be reducing the need for a second income, though that effect is harder to measure and likely concentrated among higher-income households, which are a relatively small share of the workforce. Older cohorts have also experienced a pullback in labor force participation, reflecting aging demographics. 

At this point, we don't view the cooling in labor supply as a worrying signal, particularly if the core age participation rate stabilizes in the months ahead. If anything, softening labor supply is helping to keep the labor market in balance as hiring demand moderates. 

In Canada, the story is a bit different. After a slow start to the year, the labour market appears to have regained momentum (Chart 7). Employment has climbed by 200k positions since February, largely supported by the services sector. Even Canada's beleaguered manufacturing sector added jobs over this period but is still 2-3% below levels before the U.S./Canada trade conflict began. The details are equally encouraging, with hiring largely concentrated in full-time, private sector positions. Meanwhile, hours worked are up and the unemployment rate is down 0.5 ppts from its near-term April peak.

The rise in hiring is particularly notable against the backdrop of a declining population, which is putting downward pressure on the pool of available workers. It also aligns with recent Bank of Canada surveys suggesting businesses on both sides of the border are adapting to a more volatile trade environment. Indeed, fewer U.S. firms appear to be delaying orders, while Canadian businesses are diversifying into new markets. What's more, activity tied to the U.S. AI boom is boosting demand for Canadian goods and services.      

Despite these favourable trends, some caution is warranted. At 6.4%, Canada's unemployment rate remains elevated, suggesting the economy is operating with some slack. The trade backdrop also remains uncertain, with trade negotiations ongoing. All told, we remain comfortable with our call for fewer job gains in the months ahead, resulting in only modest downward pressure on the unemployment rate (Chart 8).      

Q8. Should we be worried about household saving rates in Canada and the U.S. falling to multi-year lows?

The short answer is no. A low household saving rate only becomes concerning if it persists because spending consistently outpaces income, and households increasingly rely on borrowing or asset drawdowns. While saving rates have declined in both countries, the underlying drivers differ. In the U.S., resilient consumer spending has outpaced income growth. In Canada, the decline mainly reflects the unwinding of elevated precautionary savings accumulated during the last tightening cycle.

The U.S. personal saving rate fell to 2.7% in June, its lowest level in four years and well below the five-year pre-pandemic average of 6.1%. The decline reflects resilient spending alongside softer real income growth. Inflation has picked up again this year, and income has not kept pace. Even so, consumer demand has remained resilient. Spending has continued to outpace income growth, supported by strong equity market gains, larger tax refunds, and a modest reacceleration in consumer credit growth. There is little evidence to suggest rising consumer stress, with delinquencies on credit card and auto loans remaining stable across FICO scores. 

With the boost from tax refunds now largely exhausted and household savings diminished, spending is likely to cool in the second half of this year. Fortunately, growth in real disposable income has improved in recent months, alongside the improvement in the labor market and some easing in inflation. Provided these gains are sustained, consumption would not need to slow dramatically to bring it in better alignment with income growth. 

Canada's saving rate has also moved lower over the past several quarters, falling to 3.5% in Q1 2026, down from 5.9% in Q3 2024⁸.  However, the decline appears to reflect normalization rather than deterioration. The earlier saving rate was unusually high, as households tightened their belts in response to rising interest rates and prepared for higher mortgage payments at renewal. As the renewal cycle has progressed, some of those accumulated savings have been drawn down, leaving the saving rate closer to more typical levels (Chart 9). 

Near-term movements in the savings rate are likely to remain uneven. We expect it to rise temporarily in the second quarter as disposable income growth accelerates, supported by federal income-support measures, while consumer spending remains relatively subdued as higher energy prices weigh on household budgets. Beyond Q2, spending should gradually strengthen alongside improving fundamentals and confidence, causing the saving rate to drift lower. 

Q9. Is the Loonie headed for 60 cents?

No. The phrase "60-cent loonie" has become shorthand for the most bearish view of the Canadian dollar: that Canada’s challenges have become so large, and the global backdrop so different, that a recovery is no longer realistic. But pessimism is not a forecast. Looking at the fundamentals that have historically mattered for the CAD/USD exchange rate, including Canadian and U.S. interest rates, relative growth, data surprises, commodity prices, and market volatility, our fair-value model points to a loonie closer to $US0.73 today (Chart 10)⁹.  

That is not to say that the loonie should trade at fair value every day. The CAD averaged roughly $US0.71 in July and sits near $US 0.715 in mid-August, about 3% below our estimated fair value of US$0.73. Gaps like this are common in currency markets. Exchange rates are asset prices and can move ahead of, lag, or temporarily detach from measured fundamentals as investors absorb new information, reassess risks and respond to factors that are difficult to observe directly, including liquidity conditions and swings in risk appetite. The relevant question is whether the gap is unusually large, persistent, and supported by a meaningful change in fundamentals. On that test, today’s undervaluation does not stand out. Larger negative deviations persisted in 2014-2015 on the other side, and the CAD was modestly overvalued according to our model through much of 2022-2024. For that reason, we do not treat every bout of volatility as evidence that fair value has changed. Market moves matter when they signal a lasting shift in the economic outlook. Otherwise, it is better to focus on whether the underlying drivers have changed.

Our scenario analysis points to further volatility, but not a collapse to 60 cents. In one scenario, today’s economic and market conditions remain broadly unchanged. In another, U.S.-Canada interest rate differentials move further against Canada due to two additional Federal Reserve rate hikes. Across these scenarios, the model points to a CAD/USD range of roughly US$0.69 to US$0.74 over the next 18 months, weaker than fair value today but still far from US$0.60. Our baseline forecast reaches US$0.75 by the end of 2028, conditional on our forecast that the U.S. Federal Reserve will begin to reduce interest rates next year. 

Getting to US$0.60 — a nearly 20% decline from today — would require more than a normal forecast miss or another period of market unease. It would take several adverse shocks at once: a sharp and sustained fall in commodity prices, a historically large widening in the Canada-U.S. growth gap, a much larger rise in bond-market volatility, and limited offset from a broader U.S. dollar move against other major currencies. Put differently, a 60-cent loonie is not simply outside our base case, it is a stress scenario that would require either a severe deterioration in Canada’s fundamentals or a breakdown in the historical relationships that have anchored the currency over time.