The Receipt

In March 2026, the Governor of the Bank of Canada called leveraged hedge fund trading in sovereign debt markets a systemic risk that could cause severe dislocations. Hedge funds purchase up to 50% of Government of Canada bonds sold at auction. They finance those positions largely through short-term repurchase agreements, much of it maturing within a week. The Bank announced it would join central clearing infrastructure for its repo operations, a structural change to reduce the risk of cascading failures. That infrastructure is not scheduled to arrive until 2027. [4] [15]

Since March, the Bank of Canada’s Financial Stability Report confirmed that hedge funds have further increased their repo borrowing to finance government bond purchases, while major international institutions have continued to identify leveraged sovereign-debt trading and repo dependence as growing financial-stability vulnerabilities. [5] [6] [7] [8] [9] [10]

On August 19, the U.S. Treasury announced it would at least double the size of its bond buyback operations. By August 24, officials said the nearly $1 trillion Treasury General Account could potentially be tapped to fund further purchases. [1] [3] The U.S. government is actively intervening in its own bond market to contain long-end yields. That is a live demonstration of the kind of sovereign-market stress that the Bank of Canada has warned could activate the repo vulnerability in Canada’s own bond market.

A complication has emerged since March: government bonds have not behaved as reliably as safe-haven assets during some recent stress episodes. The Bank of Canada has flagged this directly. [5] If the next stress event triggers selling of sovereign bonds rather than buying, the funding disruption and the collateral repricing could compound rather than offset each other.

Five days ago, U.S. Treasury Secretary Scott Bessent announced that the Treasury would at least double the maximum size of its bond buyback operations, from $2 billion to at least $4 billion per operation, targeting the 10- to 30-year portion of the market. [1] Long-end Treasury yields had reached their highest levels since 2007. [2] Bessent said the operations could be even larger. [1] By August 24, two senior Treasury officials confirmed the department was considering tapping the Treasury General Account, a cash reserve of roughly $950 billion, to potentially fund further purchases. [3]

The U.S. government is intervening in its own bond market to contain long-end yields. The question for Canada is what happens if that stress activates the mechanism the Bank of Canada has warned about: a rapid unwind of leveraged hedge fund positions in sovereign debt, transmitted through short-term repo funding that connects both markets.

This publication documented that structural exposure five months ago in The Repo Problem. The vulnerability it described has not diminished. The conditions that could activate it are closer.

What Has Changed Since March

On March 4, 2026, Bank of Canada Governor Tiff Macklem told the Global Risk Institute that hedge funds purchase up to 50% of Government of Canada bonds sold at auction, that their positions are usually highly leveraged through short-term repo funding, and that a rapid unwind could cause severe dislocations in sovereign debt markets. He described the scale and speed of leveraged hedge fund trading in sovereign debt as posing a systemic risk. [4]

Since that speech, major central banks and international institutions have continued to identify leveraged sovereign-debt trading and repo dependence as growing financial-stability vulnerabilities.

The Bank of Canada’s Financial Stability Report, published May 28, stated that hedge funds have further increased their repo borrowing to finance government bond purchases and that vulnerabilities related to non-bank financial intermediaries continue to grow. [5] At the same time, the report noted that hedge funds’ share of government bond auction purchases was roughly unchanged from the previous year and that their share of dealer-client secondary-market trading had slightly decreased. Some asset managers also reduced risk exposures. [5] The picture is not uniformly worse. But the dimension that matters most to the repo vulnerability — the amount of short-term leverage — has increased.

Internationally, the evidence of a structural trend is consistent. The Bank for International Settlements reported in June that U.S. sovereign debt exposures of hedge funds relative to GDP have more than doubled since 2022, with similar patterns observed in Canada and the United Kingdom. [6] The IMF’s Global Financial Stability Report estimated the global basis trade at approximately $1 trillion and warned that disorderly unwinding could amplify Treasury yield volatility. [7] The U.S. Federal Reserve published a detailed decomposition of hedge fund Treasury exposures in June confirming that basis trade positions have significantly increased since 2022, though the underlying data extend only through September 2025 and cannot demonstrate further deterioration since March 2026. [8] The European Central Bank warned in May that leveraged non-bank positions can amplify repricing and fire sales, and that global sovereign shocks can spill across jurisdictions. [9] The OECD noted that correlated deleveraging and fast outflows can translate volatility into higher funding costs for sovereign borrowers. [10]

No evidence has emerged since March that the conditions identified in The Repo Problem have materially improved. No public data show a shift from short-term to longer-term repo funding for government bond positions. No published research shows a decrease in the concentration of trading activity among the largest hedge fund firms. And no stress event has tested Canadian repo markets without central bank intervention. The closest analogue is the U.S. Treasury market itself, where yields rose high enough for the Treasury Secretary to intervene directly. [1] That is the opposite of smooth unassisted functioning.

How the Stress Reaches Canada

The mechanism the Bank of Canada has warned about operates through a bridge that both countries share. Globally active hedge funds hold positions in sovereign debt across multiple jurisdictions, financing those positions through short-term repo borrowing. Bank of Canada researchers estimated that transactions associated with basis trading rose to just under 8% of Government of Canada bond trading volume by April 2024, based on exchange-for-physical proxies and simplifying assumptions. [16] A separate Bank study found that the top five hedge fund firms account for approximately 70% of trading activity in key government bond strategies such as macro and micro curve trades. [17]

If U.S. bond market volatility triggers margin calls or funding disruptions for those firms, they do not sell only U.S. positions. They sell across their portfolio to raise cash, including sovereign bonds of countries where they hold positions. Canada, where hedge funds buy up to half the bonds sold at auction, [4] would be in the path of that selling.

But hedge funds are not merely a vulnerability. Macklem made the positive case explicitly in March: their growing role is part of a healthy market-based system. They help distribute sovereign issuance to end investors, improve price discovery and liquidity, and make markets work more efficiently in normal conditions. [4] The vulnerability and the benefit share a root. Canada increasingly relies on an investor class that improves market capacity in normal times precisely because it can operate with leverage. The same structure can become destabilizing under stress.

Whether it has already done so is a separate question. U.S. long-end Treasury yields reached their highest levels since 2007 in August, and Canadian 10-year government bond yields moved into a range not seen in roughly two years around the same period. [2] [12] But Canadian yields during August were also being influenced by domestic factors, including jobs and GDP data, oil prices, and shifting expectations around a Canada-U.S. trade agreement. Multiple forces act on the Canadian curve simultaneously. Co-movement between U.S. and Canadian yields does not by itself establish that the hedge fund repo channel was the cause of any particular move.

A peer-reviewed study published in June examined exactly this question. The authors found high integration between U.S. and Canadian sovereign bond markets and persistent co-movement under normal conditions. But their stress-case analysis found that short-horizon U.S.-to-Canada spillover risk actually declined, that extreme U.S. outcomes became less informative about extreme Canadian outcomes, and that correlation-based analysis can substantially overstate cross-border spillover risk during stress. [11] That finding does not eliminate the possibility of repo-transmitted contagion. It means the evidence does not yet confirm it.

The relevant question is not whether August proves the mechanism has activated. It is whether the conditions for activation are closer. By the Bank of Canada’s own measures, they are. Hedge fund repo leverage has increased. [5] The U.S. bond market is under enough stress for the Treasury Secretary to intervene. [1] And the infrastructure intended to strengthen the Canadian repo market against precisely this kind of event is not yet operational. [15]

The strongest case for the system’s capacity to absorb this stress is specific and recent. The Bank of Canada’s Financial Stability Report noted that during 2025 Middle East-related volatility, some globally active hedge funds rapidly reduced leverage and sold government bonds, liquidity deteriorated temporarily, and markets nevertheless normalized quickly. Dealer repo balance sheets remained stable. [5] Canada’s large banks reported an average common equity Tier 1 capital ratio of 13.7% in the first quarter of 2026, well above regulatory minimums. [5] At the report’s release, Senior Deputy Governor Carolyn Rogers said the Canadian financial system remains well positioned to weather shocks. [14] That is evidence that existing liquidity management, dealer capacity, and central bank backstops can contain the mechanism described here.

It is not evidence they can contain it at every scale. Capital protects banks against losses. It does not by itself prevent market liquidity from deteriorating. The episodes the Bank cites as evidence of resilience involved moderate, localized stress. The scenario the Bank itself warns about — a broad, rapid, leveraged unwind of sovereign positions by firms operating across multiple jurisdictions — has not yet tested the system. The question is whether it will, and whether the infrastructure will be ready when it does.

The Bank of Canada’s policy rate has been held at 2.25% for six consecutive decisions through July 2026. [13] Fixed mortgage rates are priced primarily off comparable-term market funding and government bond yields, while the policy rate more directly anchors short-term and variable borrowing costs. That means the Bank of Canada can hold its rate steady and fixed-rate mortgage renewals can still become more expensive if longer-term yields rise, regardless of the cause.

The Safe-Haven Problem

The Repo Problem described a plumbing failure: what happens when overnight repo funding seizes up and leveraged investors are forced to sell government bonds into an already stressed market. That analysis assumed the bonds themselves retained their safe-asset status. If funding froze, the bonds would still be valuable collateral. The system would be disrupted, but the underlying asset would hold.

The Bank of Canada’s 2026 Financial Stability Report adds a complication. Government bonds have traditionally been seen as safe-haven assets, with demand rising during periods of market stress. The report notes that this pattern has not held as reliably in some recent episodes. [5] In the U.S. Treasury market in April 2025, in Japanese government bonds in January 2026, and in UK gilts in September 2022, sovereign bond yields spiked rather than falling during stress. The Bank warned that in the extreme, a sudden increase in government yields could trigger a liquidity spiral. [5]

The Bank also notes that recent markets displayed resilience and that liquidity returned quickly after bouts of volatility. [5] And the European Central Bank’s May 2026 review observed that U.S. Treasuries did serve as safe-haven assets during the Middle East conflict, while warning that fiscal concerns could change that perception. [9] The safe-haven relationship is not broken. It is state-dependent, and less reliable during episodes driven by inflation, fiscal pressure, or supply concerns.

What this means for the repo vulnerability: if the next stress event is one where sovereign bonds amplify rather than absorb the shock, the collateral backing overnight repos can lose value at the same moment the funding structure seizes. A hedge fund that cannot roll over its overnight borrowing must sell bonds to raise cash. If those bonds are simultaneously falling in price, the forced sale yields less cash, which means more bonds must be sold, which pushes prices lower. That feedback loop is what the Bank of Canada described. It does not require a permanent change in the safe-haven relationship. It requires one episode of sufficient severity occurring before the infrastructure is in place.

The Response Gap

On March 4, Macklem announced two infrastructure changes. The Bank of Canada would join the Canadian Collateral Management Service for its repo operations by early 2027. And the Bank intends to use the Canadian Derivatives Clearing Corporation for central clearing once TMX Group completes system upgrades. [15] In the same speech, Macklem noted that the U.S. SEC has mandated central clearing for most Treasury-collateralized repo transactions by June 2027, and that the European Central Bank began centrally clearing some operations in 2026. [4]

Central clearing is the right structural response. A central counterparty sitting between both sides of a repo transaction reduces the risk that one firm’s failure cascades through the system. It allows netting of positions, which reduces total collateral requirements. It imposes standardized margin requirements that remain stable through market cycles rather than spiking during stress. No serious participant disputes this. Work is progressing toward the announced timeline.

The question is whether the vulnerability grows faster than the infrastructure. This piece is published 24 weeks after The Repo Problem. The Bank of Canada’s own Financial Stability Report confirms that hedge fund repo leverage has increased over that period. [5] The U.S. bond market has come under enough stress for the Treasury Secretary to expand intervention. [1] No public announcement has indicated the Canadian infrastructure schedule has been accelerated.

Private market liquidity alone was insufficient to restore orderly functioning during March 2020, the UK gilt crisis of September 2022, and the U.S. Treasury selloff of April 2025. Each of those episodes required official intervention to prevent a funding disruption from becoming a broader financial crisis. The infrastructure being built is designed to reduce the need for that intervention. It is not yet in place.

What Would Change This Assessment

This assessment would need revision under any of the following conditions.

  • Hedge fund repo leverage declines. A future Bank of Canada Financial Stability Report finds that hedge fund short-term repo borrowing to finance government bond positions has materially declined from the May 2026 baseline.
  • Concentration decreases. Future Bank of Canada transaction-data research finds that top-five concentration in the relevant hedge fund strategies has fallen materially from the February 2026 baseline.
  • U.S. bond market stabilizes without extraordinary intervention. The Treasury reduces the expanded long-end buyback program back to its prior schedule while standard liquidity and auction metrics remain normal.
  • Canadian repo markets absorb a stress event without official intervention. During a future episode of elevated U.S. Treasury volatility, Canadian repo volumes, haircuts, dealer balance sheets, and Government of Canada bond market liquidity remain stable without Bank of Canada emergency operations.
  • Infrastructure becomes operational before the next stress test. The Bank of Canada or TMX Group reports materially expanded central clearing or CCMS adoption sufficient to reduce bilateral short-term repo exposure before the next Financial Stability Report.