November 18, 2025
Daniel Garant, Executive Vice President & Global Head of Capital Markets & Credit Investments, recently participated in CFA Montreal’s webinar “Should Investors Worry About Market Concentration?”
When investors ask whether they should worry about market concentration, they may be missing a bigger challenge: how do you build portfolios when a handful of companies dominate public equity benchmarks?
From left to right: Stephen Hui (Moderator, Pembroke); Owen Lamont (Acadian); Daniel Garant (BCI).
Today’s market concentration landscape
The S&P 500’s top 10 holdings now represent roughly 40% of the index, up from nearly 19% in 2010.1 Technology giants drive this concentration — the so-called Magnificent Seven: Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla.
This could create real portfolio construction challenges for some investors. When benchmark allocations tilt heavily toward similar names, in this case technology, active stock selection becomes difficult and passive investors find themselves with significant exposure to a handful of companies.
Some investors draw parallels to the 1990’s/2000’s dot-com era, yet the fundamentals tell a different story. At the time, the S&P 500 index included several overvalued tech companies with speculative business models. These companies traded at high valuations with minimal earnings, valuing potential rather than actual profitability. Pets.com, the poster child of dot-com excess, raised US$82.5 million in its February 2000 IPO but filed for bankruptcy by November of the same year.
Today’s reality differs. The Magnificent Seven generated over $1 trillion2 in combined annual revenue in 2024, representing established businesses that make dot-com comparisons inappropriate. Tech stock market leadership only becomes problematic when valuations disconnect from underlying performance.
AI infrastructure deployment: The timing challenge
Raising a bigger question for the entire technology sector: when will AI investments deliver expected returns and how long will it take? Companies across the industry are investing heavily in AI infrastructure but not everyone can be a winner.
Daniel Garant: “My biggest worry is it might take longer and it might cost more. Some of these assumptions about translating revenue into profit are optimistic.”
Garant’s timing concerns become clearer when examining actual deployment challenges. Physical constraints create unexpected bottlenecks. McKinsey estimates AI demand requires $5.2 trillion in capital expenditure through 2030, translating to adding 156 gigawatts (GW) of AI-related data center demand.3 The challenge then extends beyond building data centers — powering them creates the real timing gap.
McKinsey identifies a mismatch between data center builds, which can be done in 18 to 24 months, and power infrastructure development, which can take anywhere from three to ten years to complete.4 This mismatch means today’s AI infrastructure investments may not be operational until the late 2020s.
Northern Virginia Reality Check: Data center operators seeking space in Northern Virginia – the United States’ largest data center market – face six to eight years wait times for power generation, revealing capacity constraints even in the most established markets.
Despite massive capital commitments, the pace of AI adoption and actual economic impacts remain uncertain. Garant expects the winners will be the companies that can execute, rather than just announce investments.
How BCI addresses concentration risks through portfolio construction
For BCI, the concentration discussion underscored a challenge with public equity benchmarks. Garant explained how the quality of public equity benchmarks has deteriorated. Over the past 15 years, private equity firms have increasingly acquired troubled public companies. Improved growth of these delisted companies occurs outside of public markets, leading to benchmark degradation across multiple sectors, not just technology. This has changed the risk-return profile available to public equity investors.
Our approach emphasizes robust portfolio construction across asset classes, geographies, and risk factors rather than attempting to solve concentration through individual stock selection alone.
With strong funded ratios, clients have reduced overall public equity allocations from approximately 50% five years ago to 25% today. To align with this shift in client allocation, BCI has identified opportunities in areas such as private debt.
Building a winning strategy: BCI’s private debt program launched in 2018 and has become a portfolio cornerstone for our clients. Dive deeper into Garant’s insights on the strategy and asset class evolution with Top1000Funds here.
Key takeaways
The AI transformation is real and permanent. Companies are investing heavily in AI to improve productivity, decision-making, and operational efficiency. Yet not all AI investments will deliver proportional returns.
The next five years will test which institutional investors can adapt fastest. AI deployment realities will separate winners from losers, while private markets will likely continue to absorb growth opportunities. Success will depend less on predicting market concentration and more on positioning portfolios where value creation can occur.
While market concentration dominates the investment landscape, BCI is focused on:
- Multi-factor analysis: BCI analyzes various factors, such as data center capacity gaps to factor into decision making
- Staying ahead of market changes: BCI’s total portfolio approach adapts to market changes demonstrated by our move into other strategies such as private debt and investment grade private debt.
- Investing with purpose: BCI invests where returns make sense for clients. Our mandate is simple: generate returns for our clients.
References
1 Source: Bloomberg. S&P 500 Index data as of November 13, 2010, and November 13, 2025.
2 Source: Bloomberg. USD sales revenue data as of December 31, 2024.
4 https://www.mckinsey.com/featured-insights/mckinsey-explainers/what-is-a-data-center
