FTSE Russell's September 2026 report finds investors navigating a barbell market adding AI-linked growth assets while loading up on high-quality fixed income
Global investors are simultaneously adding AI-driven equity exposure and loading up on high-quality fixed income, according to FTSE Russell's September 2026 Asset Allocation Insights report.
It’s a barbell dynamic that carries direct implications for how Canadian advisors are positioning client portfolios heading into the fourth quarter.
Sovereign bond yields have hit post-global financial crisis highs across the US, UK, and Europe, with the move driven primarily by rising real yields rather than inflation expectations. For advisors, that distinction matters: it signals a structural repricing of capital costs, not a short-term inflation spike that central banks are likely to reverse quickly.
AI capex is reshaping investment-grade credit
Hyperscaler AI capital expenditure is now estimated to surpass $1 trillion (USD) in 2027, with Nvidia guiding to 70 per cent revenue growth for fiscal year 2028. The scale of that spending is creating downstream effects in credit markets that fixed income allocations may not yet fully reflect.
The five largest US hyperscalers (Amazon, Alphabet, Meta, Microsoft, and Oracle) accounted for 12.9 per cent of USD investment-grade bond issuance year-to-date in 2026, up sharply from 1.7 per cent in 2024. Their share within the industrial segment of the IG market has climbed to 23.6 per cent. Utilities are also issuing more debt to fund the power infrastructure feeding AI data centres, compounding the supply pressure.
As Wealth Professional has reported, the scale of AI capex debt issuance is already reshaping credit markets for advisors</a>, with concentration risk emerging as a longer-term concern even as near-term demand from investors remains firm. Total USD IG issuance is running 27 per cent higher year-over-year in 2026, versus 11.7 per cent growth for high yield. Free-cash-flow margins have turned negative for four of the five major hyperscalers as capex commitments outpace near-term cash generation.
What structurally higher yields mean for duration
The 30-year US Treasury yield hit a post-GFC high of 5.3 per cent in August 2026 after the US Treasury Department announced it would double its long-dated debt buyback programme; a move the market read as an attempt to suppress long-end yields rather than standard liquidity management.
According to the OECD Global Debt Report, interest expense across OECD countries reached 3.3 per cent of GDP in 2025, with rising yields and expanded issuance expected to add a further 2.5 percentage points in 2026. The structural driver is governments borrowing more at higher yields — and that pressure is not fading quickly.
As Wealth Professional has covered, elevated sovereign yields are strengthening the investment case for longer-duration bonds, with higher starting yields historically translating into improved forward returns for fixed income investors. FTSE Russell's own research found correlation coefficients as high as 0.95 between starting Treasury yields and two-year forward returns on short-duration indexes suggesting the entry point for government bonds is historically attractive, even if the near-term path carries rate risk.
With August nonfarm payrolls in the US coming in at 162,000 and headline CPI at 3.4 per cent, market-implied probability of a September Federal Reserve rate hike has risen to approximately 94 per cent, according to FTSE Russell data as of September 14, 2026.
Barbell flows and the concentration problem
Fund flow data across both three-month and six-month windows shows a barbell taking shape: capital is moving simultaneously into US and global growth equity, particularly technology and AI-adjacent names, and into high-quality fixed income. It is not a clean risk-on or risk-off rotation. Investors are seeking yield and growth at the same time, a difficult balance to sustain if either leg of the trade comes under pressure.
Equity market concentration is at or near historic highs. As of August 2026, the US represents 61.7 per cent of FTSE All-World market capitalisation, while technology alone accounts for 44.1 per cent of the FTSE USA index, now greater than the combined weight of nine other industries, including financials, healthcare, energy, and real estate.
As Wealth Professional has reported, the risks investors face as AI dominates an ever-larger share of market capitalisation are becoming harder to manage through standard index exposure alone. FTSE Russell flags broader regional diversification, equally weighted indexes, and RIC-capped structures as potential tools to reduce unintended factor concentration in client portfolios.
Credit spreads offer thin compensation
US high-yield credit spreads tightened to a 10-year low of 260 basis points in August 2026, while world IG spreads sit at approximately the fifth percentile of their 10-year range. That is historically thin protection.
For Canadian advisors using credit allocations to generate yield pickup over government bonds, those spread levels warrant scrutiny, particularly given that sovereign bonds are now offering starting yields not seen since before the GFC.
The FTSE Russell report frames the current asset class hierarchy clearly: government bonds are attractive relative to their own history, equities are doing heavy lifting on earnings expectations, and credit is priced for near-perfection. That ordering represents a meaningful departure from the post-2008 playbook most advisors have built their fixed income strategy around.
The September 2026 picture presents Canadian advisors with a genuine portfolio construction challenge: government bonds at historically compelling entry points, equity valuations dependent on sustained earnings upgrades, and a credit market absorbing a surge of AI-driven supply at the tightest spreads in a decade.
The barbell that global capital flows are tracing may reflect less a confident strategy than a hedge against uncertainty and navigating it will require advisors to take an explicit view on duration, sector concentration, and how long AI spending can continue to justify current equity multiples.
