Q2 Update: Past the Shock, Not Past the Test
Corrado Tiralongo - Jun 30, 2026
Markets have moved quickly from crisis pricing to relief pricing.
That is understandable. The immediate risk of a highly adverse energy shock has faded as a fragile U.S.-Iran agreement reduces the risk that energy flows through the Strait of Hormuz remain severely disrupted. That matters. A prolonged disruption to one of the world’s most important energy chokepoints was never a narrow geopolitical issue. It was a direct threat to inflation, growth, consumer spending, corporate margins and central bank policy.
Relief is justified.
Complacency is not.
The market can reprice in a day. Energy systems do not repair themselves that quickly. Tankers need to be repositioned, insurance markets need to normalize, inventories need to be rebuilt, and production systems need time to return to capacity. Natural gas is even more complicated given the damage to Qatari production. The acute shock has faded, but the after-effects are still working their way through the global economy.
That is the right starting point for the quarter ahead. The outlook is better than it was a few weeks ago, but it is not clean.
Inflation is changing, not disappearing
Lower oil prices should help headline inflation. Consumers will feel some relief at the pump, and central banks will have less immediate pressure to respond to energy-driven inflation with higher interest rates.
That is the good news.
The more difficult point is that inflation does not move through the economy all at once. Energy prices show up quickly in gasoline. They show up later in airfares, electricity, food and other parts of the consumer basket. Fertilizer prices have already been affected by the conflict, and food inflation tends to arrive with a lag. Even if oil prices settle lower, some of the inflationary impulse from the shock is still in the pipeline.
This is why markets and central banks can look at the same data and draw different conclusions. Markets see falling oil prices and want to declare the inflation problem over. Central banks see a shock that may be easing but still needs to be monitored for pass-through.
The distinction matters.
A temporary energy shock can be looked through. A shock that spreads into wages, expectations and core inflation cannot. So far, there is limited evidence of broad second-round effects in most major economies. That gives central banks room to be patient. It does not give them room to be careless.
Canada sits directly inside that tension. The economy is weak; growth has disappointed and there is still more capacity in the economy than demand to use it. At the same time, productivity is poor, wage growth has been resilient and unit labour costs remain uncomfortable. The Bank of Canada is boxed in. Cutting rates risks sending the wrong signal on inflation. Raising rates risks tightening into weakness.
Our base case remains that the Bank of Canada stays on hold. That is not a strong policy tailwind. It is a pause forced by conflicting risks.
The U.S. is different. The economy has held up better, the labour market has firmed, and fiscal policy remains highly supportive. The market’s repeated hope for easier policy keeps running into a simple problem: the U.S. economy has not earned rate cuts. If anything, the risk is that the Federal Reserve (Fed) has more work to do.
That creates a difficult backdrop for risk assets. Inflation is no longer the acute threat it was during the energy spike, but it is not gone. Growth is not collapsing, but it is uneven. Central banks are not all tightening, but they are not delivering the rate relief investors wanted earlier this year.
That leaves earnings.
And earnings leave us with AI.
The AI story is real, but earnings quality deserves more scrutiny
AI remains the dominant force in markets.
The buildout is real. It is supporting U.S. business investment, capital spending, demand for semiconductors, power infrastructure, data centres, software and hardware. It is also reshaping global trade. AI-related goods are becoming a larger share of export growth, particularly across parts of Asia.
For now, AI is showing up more clearly as a demand shock than a productivity shock. The buildout is boosting capital spending, hardware demand and trade, while the broader productivity payoff remains more gradual.
This is not the late 1990s in every respect. The current rally has more earnings support than the dot-com bubble did. The AI infrastructure cycle is not imaginary. Companies are spending real money, generating real demand and creating real revenue across the supply chain.
But that does not mean the earnings story is as clean as the headline numbers suggest.
One of the more important issues in this cycle is that reported earnings can be affected by unrealized gains on equity holdings. In plain English, some companies own stakes in other public or private companies. When the value of those stakes rises, accounting rules can allow those gains to flow through reported earnings, even if the gains have not been realized in cash.
That matters in an AI-driven market because several large companies own stakes in private AI-related businesses. As those private companies raise capital at higher valuations, the value of those stakes can rise materially. The result is that reported earnings may look stronger, not only because the operating business is producing more profit, but also because asset values embedded on the balance sheet have been marked higher.
The scale may not be trivial. One recent estimate found that unrealized, non-operating mark-to-market gains at just three companies, Alphabet, Amazon and Nvidia, added approximately $69 billion to reported S&P 500 earnings in the first quarter. On that estimate, these gains represented roughly 12% of total quarterly S&P 500 earnings and helped lift reported earnings growth from approximately 16% to 28%. The same estimate suggested that these non-operating gains represented approximately 60% of Alphabet’s reported net income, 51% of Amazon’s and 27% of Nvidia’s.[1]
The underlying businesses remain strong. That is not the point. The point is that not every dollar of reported earnings growth should be treated the same way.
Operating earnings and valuation-driven accounting gains are different things. One speaks to the repeatable economics of the business. The other may reflect a funding round, an IPO, or a change in market valuation. That information is useful, but it can mislead investors if it is treated as recurring profitability.
This is where the market risks becoming circular. Public equity valuations rise on AI enthusiasm. Higher public valuations help justify higher private valuations. Higher private valuations boost the reported earnings of public companies that own stakes in those businesses. Stronger reported earnings then make the public market look less expensive, which can further support the original valuation.
We have seen this movie before in different forms. Feedback loops are a common feature of market booms. The details change. The psychology does not.
The risk is not only that these gains reverse. The more basic risk is that they do not recur. If a portion of earnings growth came from valuation marks rather than operating profit, future earnings growth will need stronger operating performance to keep the same pace. That raises the bar at a time when expectations are already high.
This does not invalidate the AI investment case. It does make the quality of earnings more important. We need to distinguish between durable operating earnings, one-time accounting gains, financial engineering and valuation-driven markups. That distinction will matter more as AI companies come public, private valuations become observable and investors get a clearer look at the economics behind the buildout.
The better question is no longer whether AI matters. It does. The better question is how much of the good news is already reflected in prices, and how much of the recent earnings strength is repeatable.
A narrower margin for error
The market has become increasingly dependent on a narrow set of assumptions.
AI capital spending must remain strong. Semiconductor demand must continue to surprise positively. Hyperscaler earnings must validate the investment cycle. Power and energy constraints must remain manageable. Public market investors must absorb a wave of new AI-related equity issuance without damaging sentiment.
That is a lot to ask.
Recent market behaviour supports a more disciplined view. Large technology and semiconductor stocks have become more volatile. Some of the companies tied most closely to AI infrastructure are moving sharply on relatively narrow pieces of news. That is what crowded positioning looks like. It does not take a collapse in the AI thesis to create a correction. It only takes a small disappointment.
The coming wave of AI-related IPOs adds another warning light. It is not a precise timing signal. IPO cycles rarely are. But large equity issuance often appears late in market booms, when private owners recognize that public investors are willing to pay high prices for future growth.
These listings will also force more transparency. Private valuations that were previously observed mainly through funding rounds will become part of the public market conversation. That will matter not only for the newly listed companies, but also for the companies that own stakes in them. If rising private valuations have helped reported earnings on the way up, falling valuations can work in reverse. Even if valuations do not fall, a slowdown in new markups would remove a source of earnings growth that investors may have treated as more durable than it really was.
Markets have been willing to fund the AI buildout on faith, enthusiasm and extrapolation. Public market scrutiny will force more discipline. That is healthy over the long run. It may be uncomfortable over the short run.
For now, neither higher shipping costs nor the expected shift in U.S. tariff authority appears large enough to derail world trade. The AI supply-chain impulse remains the more important trade story.
The quarterly outlook
The quarter ahead is likely to be defined by a tug of war between relief and validation.
Relief comes from the fading energy shock, lower oil prices and reduced pressure on central banks outside the U.S. Validation has to come from earnings, especially in the parts of the market that have already priced in a large share of the AI opportunity.
That is the challenge.
The macro backdrop is not strong enough to carry equities by itself. The rate backdrop is not easy enough to bail out stretched valuations. The earnings backdrop is good, but increasingly narrow. AI is powerful, but the market is now testing the difference between a durable investment theme and an overextended trade.
Our view remains constructive, but more selective. This is not an environment to retreat from markets. It is an environment to be more deliberate about where risk is taken.
We continue to favour U.S. equities. The U.S. has stronger structural support than most major developed markets, including deeper capital markets, stronger technology leadership, better productivity potential and a more durable earnings base. That does not mean chasing the narrowest version of the AI trade. U.S. exposure should not be reduced to owning the most crowded stocks at any price. We prefer exposure that balances participation in the AI investment cycle with quality, earnings delivery and valuation discipline.
We remain underweight Canada. The Canadian market can benefit from higher commodity prices during periods of energy stress, but that is not the same as a broad structural advantage. The domestic economy is soft, productivity remains weak and the equity market is concentrated in sectors that do not provide the same exposure to the strongest global earnings themes. Canada still has a role in portfolios, but the case for a large overweight is not compelling.
We remain underweight developed international equities. Valuations are more attractive, but valuation alone is not a catalyst. Europe and Japan remain more exposed to energy and import cost shocks, earnings momentum is less compelling, and policy support is less straightforward. These markets may perform well in short bursts, especially when the U.S. dollar weakens or global cyclicality improves, but we do not see enough evidence to make them a preferred source of equity risk.
We remain overweight emerging markets, but selectively. Parts of Asia are direct beneficiaries of the AI supply chain, particularly through semiconductors, hardware and related exports. Emerging markets also offer more attractive valuations and exposure to parts of the global economy that are benefiting from supply chain realignment and commodity demand. The risk is that EM exposure becomes too narrow a proxy for semiconductors. We want the AI supply-chain exposure, but we do not want the entire emerging market thesis to depend on one crowded trade. Selectivity matters because some emerging markets remain more vulnerable to food inflation, currency pressure and policy tightening, even as others benefit from the AI supply chain.
Within fixed income, we continue to see a role for duration as a stabilizer. That does not require a near-term rate-cutting cycle. It requires recognition that growth risks have not disappeared and that high-quality bonds can still provide ballast if equity markets stumble. Credit requires more caution. Spreads are not pricing much stress, and a slower growth environment leaves less room for disappointment. We prefer quality overreaching for yield.
Alternatives remain important. The point is not that alternatives remove risk. They provide different risks. In this environment, that distinction matters. Liquid diversifiers, managed futures, managed volatility strategies and other diversifying return streams can create flexibility when equity markets become too dependent on a narrow set of outcomes.
Liquidity plus diversification gives portfolios the ability to monetize gains at the right time. That is especially important when the market is expensive, leadership is narrow and volatility can reappear without warning.
What needs to go right
The market can keep rising. That is not the issue.
The issue is what now has to go right.
The fragile U.S.-Iran agreement that has reduced the risk of disruption to energy flows through the Strait of Hormuz has to hold. Inflation has to cool without forcing central banks back into a tightening mindset. The Fed has to avoid becoming more hawkish than markets expect. AI earnings have to keep validating the capital spending cycle. Reported earnings need to be supported by durable operating profits, not flattered by valuation-driven accounting gains that may not recur. The wave of equity issuance has to be absorbed without weakening sentiment. Semiconductor leadership has to broaden or at least stop wobbling. Trade tensions need to remain manageable.
None of those conditions is impossible. Together, they leave less room for error.
That is why we enter the quarter with deliberate risk, not blind risk. We want participation in the parts of the market where earnings, innovation and capital spending remain strong. We do not want portfolios overly dependent on the assumption that one theme, one region or one narrow group of companies can carry the entire market indefinitely.
Markets have moved past the shock.
The test is whether they can move past the consequences.
Sincerely,
Corrado Tiralongo (he/him)
Vice President, Asset Allocation & Chief Investment Officer
Canada Life Investment Management Ltd.