Private markets and AI create a scaling problem asset managers haven't solved
Canadian pension funds manage roughly $2.4 trillion in assets, and the operational costs of technology infrastructure for evaluating private credit deals are substantial, though precise per-professional figures vary widely across institutions. That labour cost doesn't include the deal itself. It's the cost of knowing whether the deal is worth doing.
The expense explains why alternative investments remained the domain of institutions for decades. Due diligence on a private loan to a mid-market manufacturer in Quebec requires bespoke analysis, credit models built from scratch, site visits, conversations with management, legal review of covenants that differ deal to deal. Public equities, by contrast, are evaluated once and traded by thousands. The operational model doesn't translate.
In 2026, asset managers are trying to make it translate anyway. Institutional capital continues shifting toward private credit and real estate as a hedge against public market volatility, and retail products marketed as "liquid alternatives" now allow TFSA and RRSP holders access to formerly restricted asset classes. The Ontario Securities Commission has issued multiple bulletins reminding firms that "liquid" in this context means the fund structure includes redemption windows, typically quarterly, subject to manager discretion.
The growth opportunity is real, but the operational bottleneck is becoming obvious. A mid-sized Canadian asset manager with $12 billion in AUM might handle 400 private deals in a given year. Each one requires analyst time, compliance sign-off, and custom valuation work. Scaling that process to serve retail demand, where average account sizes drop from $5 million to $50,000, means the cost per dollar managed rises unless something in the workflow changes.
Where firms are betting on AI
Asset managers are splitting their AI budgets into two streams. Generative AI tools are being deployed to automate the middle-office work: summarizing due diligence reports, drafting compliance memos, extracting key terms from 200-page credit agreements. The goal is to compress the time from deal identification to allocation decision by 30 to 40 percent without adding headcount.
Predictive AI, the other stream, is being tested for alpha generation, specifically, finding patterns in private market data that human analysts miss. A credit model trained on five years of Canadian private lending outcomes can flag risk concentrations in certain sectors or geographies before they become visible in default rates. The challenge is that private market data remains sparse and non-standardized. A model trained on public bond spreads and applied to private credit deals will misfire, because the two instruments don't share the same liquidity, covenant, or enforcement dynamics.
The deeper risk is cultural. Asset management has historically hired for judgment and relationship skills. The shift to AI-heavy workflows requires data scientists who understand market microstructure, not just Python. Firms are running dual hiring tracks: traditional portfolio managers for client-facing roles, and quantitative researchers for the operational layer. The integration between the two groups is uneven. At some shops, the quants report to technology. At others, they sit inside investment teams. Nobody has settled the org chart.
The margin problem that isn't going away
Management fees on private market funds remain in the 1 to 2 percent range, with carried interest on top. That fee level made sense when clients were institutions writing $50 million cheques. It becomes harder to defend when the same strategy is being packaged into a $10,000 minimum retail fund, where the operational cost per account is higher and the margin per investor is thinner.
The industry's answer has been to push toward greater efficiency through technology, but efficiency only solves half the equation. The other half is the illiquidity premium itself. As more retail capital flows into private markets, the premium, the extra return investors demand for locking up their money, may compress. If private credit starts pricing like high-yield bonds, the rationale for the fee structure weakens.
Scale, in other words, isn't just about managing more capital. It's about managing it profitably while the structural advantage that justified the fees starts to erode.
Canadian pension funds manage roughly $2.4 trillion in assets, and the operational costs of technology infrastructure for evaluating private credit deals are substantial, though precise per-professional figures vary widely across institutions. That labour cost doesn't include the deal itself. It's the cost of knowing whether the deal is worth doing.
The expense explains why alternative investments remained the domain of institutions for decades. Due diligence on a private loan to a mid-market manufacturer in Quebec requires bespoke analysis, credit models built from scratch, site visits, conversations with management, legal review of covenants that differ deal to deal. Public equities, by contrast, are evaluated once and traded by thousands. The operational model doesn't translate.
In 2026, asset managers are trying to make it translate anyway. Institutional capital continues shifting toward private credit and real estate as a hedge against public market volatility, and retail products marketed as "liquid alternatives" now allow TFSA and RRSP holders access to formerly restricted asset classes. The Ontario Securities Commission has issued multiple bulletins reminding firms that "liquid" in this context means the fund structure includes redemption windows, typically quarterly, subject to manager discretion.
The growth opportunity is real, but the operational bottleneck is becoming obvious. A mid-sized Canadian asset manager with $12 billion in AUM might handle 400 private deals in a given year. Each one requires analyst time, compliance sign-off, and custom valuation work. Scaling that process to serve retail demand, where average account sizes drop from $5 million to $50,000, means the cost per dollar managed rises unless something in the workflow changes.
Where firms are betting on AI
Asset managers are splitting their AI budgets into two streams. Generative AI tools are being deployed to automate the middle-office work: summarizing due diligence reports, drafting compliance memos, extracting key terms from 200-page credit agreements. The goal is to compress the time from deal identification to allocation decision by 30 to 40 percent without adding headcount.
Predictive AI, the other stream, is being tested for alpha generation, specifically, finding patterns in private market data that human analysts miss. A credit model trained on five years of Canadian private lending outcomes can flag risk concentrations in certain sectors or geographies before they become visible in default rates. The challenge is that private market data remains sparse and non-standardized. A model trained on public bond spreads and applied to private credit deals will misfire, because the two instruments don't share the same liquidity, covenant, or enforcement dynamics.
The deeper risk is cultural. Asset management has historically hired for judgment and relationship skills. The shift to AI-heavy workflows requires data scientists who understand market microstructure, not just Python. Firms are running dual hiring tracks: traditional portfolio managers for client-facing roles, and quantitative researchers for the operational layer. The integration between the two groups is uneven. At some shops, the quants report to technology. At others, they sit inside investment teams. Nobody has settled the org chart.
The margin problem that isn't going away
Management fees on private market funds remain in the 1 to 2 percent range, with carried interest on top. That fee level made sense when clients were institutions writing $50 million cheques. It becomes harder to defend when the same strategy is being packaged into a $10,000 minimum retail fund, where the operational cost per account is higher and the margin per investor is thinner.
The industry's answer has been to push toward greater efficiency through technology, but efficiency only solves half the equation. The other half is the illiquidity premium itself. As more retail capital flows into private markets, the premium, the extra return investors demand for locking up their money, may compress. If private credit starts pricing like high-yield bonds, the rationale for the fee structure weakens.
Scale, in other words, isn't just about managing more capital. It's about managing it profitably while the structural advantage that justified the fees starts to erode.
Sources
Read Next
Asset managers cut product portfolios to fund AI and outsourcing overhauls
ETFs now hold 42% of Canadian fund assets as OSC tightens crypto and liquidity rules
One in Five Canadian Parents Still Pays Bills for Kids in Their Late Thirties
Joint mortgages surge in Ontario and B.C. as first-time buyers face rising delinquency pressure