The Unbundling of the Index

The Problem with the ETF

The ETF is a flawed product in taxable accounts, and most investors do not realize it. The flaw is not in what ETFs deliver. It is in what they structurally cannot.

Index investing changed everything. Before ETFs, most retail investors had no practical way to own a diversified basket of stocks without paying expensive mutual fund managers. ETFs changed that. With a single trade, an investor could own the whole S&P 500 at near-zero cost, with institutional grade efficiency. By the 2010s they had become the default holding for index investors across every account type.

But ETFs are fundamentally a one-size-fits-all product, and that simplicity comes with structural compromises. For tax-sheltered accounts like RRSPs and TFSAs these tradeoffs are largely irrelevant. In a taxable account, they are significant.

The core limitation is that an ETF is a pooled fund. When you buy an ETF, gains and losses are trapped inside the vehicle. If the market drops 15% and your ETF falls with it, you can only realize a loss by selling the entire fund, giving up future upside and potentially disrupting your overall investment plan. You cannot surgically harvest losses on individual holdings because you don't own them. The ETF does.

The second limitation is personalization. ETFs offer no flexibility. If you want broad market exposure but want to exclude your employer's stock, avoid tobacco companies, or overweight technology, you are largely stuck. Niche thematic ETFs exist, but they require you to accept an entire predefined basket, often at higher fees, and frequently with methodology that only partially reflects your intentions.

For the average investor, using broad ETFs is a trade-off: you get simplicity at the cost of tax control and personalization. ETFs remain an excellent gross-return product, but for non-registered accounts they are fundamentally flawed when it comes to preference-based alignment and tax efficiency.

What Direct Indexing Actually Is

Direct indexing is a simple idea with significant implications. Instead of buying a fund that tracks an index, you buy the individual stocks of the index directly into your own brokerage account. Software keeps the portfolio aligned with the index, handles rebalancing, and monitors continuously for tax-loss opportunities.

The end result looks almost identical to an ETF from a performance standpoint. But underneath the surface, the structure is fundamentally different, and that difference matters enormously in a taxable account.

The natural objection here is complexity. Owning 200+ individual stocks sounds intimidating compared to holding a single ETF ticker. But the investor never interacts with those individual positions. The platform handles construction, rebalancing, and harvesting automatically. The experience from the investor's side is comparable to owning an ETF. The difference is entirely structural, not experiential.

Tax Alpha

The biggest advantage of direct indexing in taxable accounts is systematic tax-loss harvesting. Throughout the year, some stocks in any index will rise and some will fall. When a stock drops below a threshold (say 5-10% from its purchase price), the platform automatically sells it, realizes the capital loss, and immediately reinvests the proceeds in a correlated substitute. The overall portfolio continues tracking the index, but the investor has crystallized a loss that can offset capital gains realized elsewhere.

Capital losses can offset gains in the same year, and if there are more losses than gains they can typically be carried back to prior years or carried forward indefinitely. Over many years, turning volatility into a systematic stream of tax losses creates what practitioners call tax alpha. Academic studies and industry back-tests have shown this can add anywhere from 0.5-2.0% per year in post-tax returns for investors in taxable accounts. The exact benefit depends on volatility, tax bracket, and how frequently the investor realizes gains, but the directional case is clear.

To make this concrete, consider an Ontario investor contributing $50,000 per year to a non-registered account tracking a broad Canadian equity index. Assume a combined marginal tax rate of roughly 48% on income and an inclusion rate of 50% on capital gains. Over 10 years, that is $500,000 in total contributions. In a typical year, even a rising market will have 20-30% of its constituent stocks trading below their purchase price at some point. If the platform harvests losses equal to 2% of the portfolio value annually, a conservative estimate given normal stock-level volatility, the cumulative harvested losses over a decade could exceed $40,000. Those losses offset gains realized elsewhere, whether from selling an investment property, exercising stock options, or rebalancing other holdings. At Ontario's effective capital gains tax rate, that translates to roughly $10,000-$15,000 in actual tax savings over the period. An ETF holding the same index over the same period would generate zero harvested losses because the gains and losses remain trapped inside the fund. The investor's pre-tax returns would be nearly identical, but the after-tax outcome would be meaningfully worse.

This is not a benefit exclusive to the ultra-wealthy. Any ordinary investor who contributes regularly to a taxable account and occasionally realizes gains will benefit meaningfully. An ETF cannot provide this at the same level. Gains and losses are trapped inside the pooled fund structure, with no ability to selectively harvest at the individual stock level.

A fair question is whether tax-loss harvesting merely defers taxes rather than eliminating them. It does reduce cost basis over time, which means larger gains when positions are eventually sold. But the time value of money matters. Paying less tax today and more tax in 20 years is still a net benefit, especially if those deferred dollars remain invested and compounding in the interim. The math favors harvesting in virtually every realistic scenario.

Customization as Everyday Risk Management

Beyond tax efficiency, direct indexing provides personalization that functions as a practical risk management tool for ordinary investors.

The most common use case is employer stock exclusion. Many investors already hold significant exposure to their employer through stock compensation or employee share purchase plans. If their employer is also included in a broad index, they are compounding that concentration risk without realizing it. With direct indexing, they can replicate an index and simply exclude that one stock, reducing the risk that both their income and their portfolio decline at the same time.

A second use case is values-based exclusion. An investor who is comfortable with broad market exposure but wants to avoid tobacco, weapons, or gambling companies no longer needs to hunt for a niche thematic ETF with narrow methodology and higher fees. They can track the index while systematically excluding specific companies or sectors.

Perhaps the most interesting use case is sector tilting. Direct indexing allows investors to act on their own knowledge without becoming stock pickers. If an investor has high conviction in technology and low conviction in green energy, they can tilt their index, overweighting tech by 10% and underweighting green energy by 10%, while still tracking an overall broad market benchmark. This captures their informed view without the concentration risk of individual stock selection.

These forms of customization help investors stick to their long-term plans. A portfolio that reflects your actual values, risk tolerance, and existing exposures is one you are less likely to abandon in a downturn.

The obvious concern with customization is tracking error. Every exclusion or tilt pushes the portfolio further from the benchmark, and over time those deviations can compound into meaningful performance differences, both positive and negative. This is a real tradeoff. But most investors are not actually trying to replicate an index perfectly. They are trying to build wealth efficiently over decades. A portfolio that excludes your employer and avoids sectors you find objectionable will underperform or outperform the benchmark by small amounts in any given year, but it is a portfolio you will actually hold through volatility instead of panic-selling at the bottom. Behavioral persistence matters more than basis-point precision.

Falling Minimums, Fractional Shares, and Automation

Direct indexing is not a new idea. For most of its history it was limited to the ultra-wealthy because it was operationally complex and expensive. Advisers managing hundreds of positions faced high commissions on every trade, and the burden of tracking tax-lot data across a large portfolio made small accounts completely uneconomical.

Three technical shifts have changed this. Zero-commission trading made it viable to rebalance and harvest frequently without transaction costs eating into returns. Fractional shares enabled investors with a few thousand dollars to hold proportional slices of hundreds of stocks. Previously you needed hundreds of thousands of dollars to hold full share increments of every name in a broad index. And automated portfolio engines can now handle the algorithmic heavy lifting: constructing portfolios, monitoring tax opportunities, enforcing tax rules, and tracking index error, all continuously and at scale.

From an investor's perspective, the experience is simple. Open a taxable account, select a strategy like Canadian Equity Index with some basic exclusions, and the platform handles the rest. The complexity is abstracted away. It feels like buying an ETF, but the underlying structure is fundamentally more tax-efficient.

There is a legitimate platform risk to acknowledge here. When you own an ETF, your holdings are standardized and portable. If your brokerage shuts down, you transfer your ETF units to another broker. With direct indexing, your portfolio is a bespoke construction of hundreds of individual positions with specific tax lots, cost bases, and substitution histories. Moving that to another platform is significantly more complex, and if your provider changes its pricing model or discontinues the product, unwinding the portfolio could trigger the very tax events you were trying to avoid. This is not a reason to avoid direct indexing, but it is a reason to choose your platform carefully and to understand that you are accepting some degree of vendor lock-in.

Canada's Inflection Point

Canada is entering an inflection point where direct indexing is transitioning from a theoretical concept for the ultra-wealthy to a live product for retail clients. While US markets have had these tools for several years, 2025 was the year the infrastructure arrived in Canada.

Wealthsimple launched Direct Indexing in fall 2025 as part of its managed portfolios, initially through an invite-only beta. The $1,000 CAD minimum signals explicitly that they are targeting ordinary investors, not high-net-worth clients. At a 0.15% portfolio fee, pricing is comparable to broad market ETF products, but with automated tax-loss harvesting included. Wealthsimple frames it not as a complex trading strategy but as the smarter default for taxable accounts.

Questrade moved in a different direction. At their product showcase in November 2025, they announced a native custom indexing and portfolio rebalancing engine built entirely in-house, a major departure from their previous approach of using third-party tools. This signals that direct indexing infrastructure is no longer an optional add-on but a core feature of the modern brokerage stack. Unlike Wealthsimple's managed, set-and-forget approach, Questrade's system is designed for more active investors, allowing manual exclusions and sector overweights with a single click, while the rebalancing engine executes the underlying bulk trades automatically.

There is also a uniquely Canadian regulatory layer that makes automation essential. The CRA's superficial loss rule denies a capital loss if the investor, or an affiliated person, purchases the same or identical property within 30 days before or after the sale. To harvest losses without triggering this rule, the platform must sell the losing stock and purchase a correlated-but-non-identical substitute. Tracking 61-day blackout windows across hundreds of positions is nearly impossible for a human and trivial for software. This is exactly why direct indexing is only now arriving in Canada, and why it arrives through platforms rather than as a DIY strategy.

It is also worth noting that most Canadians do not have large non-registered investment accounts. RRSP and TFSA contribution room absorbs the majority of savings for most households, and direct indexing provides no incremental tax benefit in those sheltered accounts. The addressable market for this product is real but narrower than the US equivalent, where taxable brokerage accounts are far more common. This does not invalidate the thesis, but it does mean adoption will be concentrated among higher-income earners, business owners, and investors who have maxed out their registered room.

The US as a Roadmap

The United States market is significantly ahead of Canada in direct indexing adoption, and its trajectory offers useful context for what is likely to happen here.

Before 2020, direct indexing was defined by specialist firms like Parametric and Aperio, accessible only to clients with minimums between $250,000 and $5 million. Between 2020 and 2024, the major asset managers moved aggressively to acquire these capabilities. Morgan Stanley bought Eaton Vance (Parametric), BlackRock bought Aperio, Vanguard acquired Just Invest, and Franklin Templeton acquired O'Shaughnessy Asset Management. By 2024, the top five providers controlled nearly all direct indexing assets. What followed was rapid democratization. Fidelity launched Managed FidFolios with a $5,000 minimum. Frec brought the barrier down to $20,000 for mass-market investors. Altrust enabled independent advisors to offer strategies with minimums as low as $2,000.

The interesting lesson from the US is not just the compression of timelines but what surprised people along the way. Adoption was slower than early advocates predicted because most retail investors did not immediately understand the tax benefit, and advisors needed time to integrate the tools into their workflows. The products that gained traction were the ones that made the value proposition obvious and the onboarding frictionless. Canada appears to be learning from this. Wealthsimple and Questrade are deploying infrastructure that took a decade to mature in the US and taking it directly to the mass-market level, skipping the high-minimum institutional phase entirely.

Platforms, Advisors, and Infrastructure

For direct indexing to become the default for taxable accounts, the experience has to be as simple as buying an ETF. That requires infrastructure most brokerages have not historically had.

On the direct-to-consumer side, Wealthsimple and Questrade are the primary channels for reaching Canadian retail investors. Bank-owned brokerages like RBC Direct Investing and TD Direct Investing have been slower to move, likely because they have high-margin proprietary ETF and mutual fund businesses to protect. As fintech competition intensifies, pressure will mount for banks to waterfall these capabilities down from their high-net-worth channels to their mass-market platforms.

The advisor channel represents a parallel opportunity. Independent advisors have historically struggled to compete with robo-advisors because they lacked the technology to offer automated tax-loss harvesting. Platforms like Envestnet, which launched a Canadian-specific direct indexing solution in March 2025 through a partnership with RBC Dominion Securities, are solving this by bundling these capabilities into advisors' existing tech stacks. Advisors can now offer differentiated, tax-optimized service without building the underlying trading infrastructure themselves.

The shift from retail investors to infrastructure and venture dynamics deserves a brief note on why both matter in the same conversation. Direct indexing is not just a consumer product. It is an emerging layer of financial infrastructure, and understanding the market structure helps explain why the consumer product is arriving now and why it will improve quickly. The platforms competing for retail clients are built on top of B2B infrastructure companies that have their own competitive dynamics and economics.

From a venture perspective, the most interesting opportunity is that B2B infrastructure layer. Companies building the APIs for tax-lot tracking, portfolio rebalancing, and CRA-compliant substitution logic operate with high switching costs and scalable recurring revenue that grows with assets on the platform. Tilt, a Canadian fintech, raised a $7.1 million seed round in September 2025 to offer exactly this capability via API to other fintechs and registered investment advisors. These companies are high-growth entry points into a large market, without taking on the customer acquisition costs of the direct-to-consumer model. The US has already validated the exit path: Morgan Stanley paid $7 billion for Eaton Vance, BlackRock paid $1 billion for Aperio. Large asset managers are willing to pay significant premiums to acquire the technology needed to retain taxable assets.

The Case for Inevitability

The structural argument is straightforward. In a taxable account, a product that systematically generates tax alpha through automated harvesting will mathematically outperform a pooled product that cannot, if their fees are similar. And fees are now similar.

ETFs remain excellent for tax-sheltered accounts and for very small balances where minimum investment thresholds matter. The simplicity, liquidity, and cost structure of ETFs is unmatched in those contexts. But in taxable non-registered accounts, the pooled structure is increasingly difficult to justify as direct indexing becomes accessible at mainstream minimums.

For a generation of investors, the ETF was the only practical way to buy the market. Fractional shares, zero-commission trading, and automated portfolio engines have removed that constraint. The question is no longer whether direct indexing is better in a taxable account. The math on that is settled. The question is how quickly the platforms, the advisors, and the investors themselves catch up to what the structure already makes possible.