3 min read
What the Anthropic IPO Means for Client Portfolios

Anthropic’s IPO has been one of the most anticipated events in the equity markets this year, and for good reason. The company is expected to debut at a total market value of up to $1.4 trillion, which would make it larger than Berkshire Hathaway on day one, even at the low end of that range. A listing this size comes with a flood of media coverage and very little price history, and clients are bound to ask what it means for them.
That’s why we brought together experts from across Morningstar and PitchBook for our recent webinar, “Inside the Anthropic IPO: Valuation Risks and Fund Impact.” The live Q&A covered what advisors want to know most: what the IPO price does and doesn’t tell you, how index rules shape fund exposure, and which risks we’re watching. Here are the questions that came up and how we answered them.
Q: Does Anthropic's IPO Dilute Existing Shareholder Value?
A: No. Whether shares sit in the public float or remain held privately does not change the ownership stake tied to each outstanding share. Alphabet GOOGL and Amazon AMZN are both strategic holders of Anthropic, and their shares simply have not entered the public float yet.
As Alex Bryan, Director of Product Management, explained, "Float does not change the total number of shares outstanding. It just changes what's available to a non-related party who wants to invest in the company." In other words, a low initial float affects how much of the company the public can buy, not how much of the company each shareholder actually owns.
Q: How Are Rising Data Center And Compute Costs Likely to Affect Anthropic?
A: Anthropic currently leases most of its compute capacity from Google, Amazon, and several other providers rather than owning its own data centers, which means the company absorbs the full compute cost as an expense rather than the smaller depreciation charge that an owner would record. We expect compute costs to rise further in 2027 and 2028, which will likely pressure margins in the near term.
Malik Ahmed Khan, Senior Equity Analyst, noted that the key question for long-term investors is whether that spending is generating durable value: "The way we think about it is, the company's post-IPO cash balance and the company's continuing revenue growth should be enough to cover the compute costs." He also pointed to Anthropic's move to work with Broadcom AVGO on owning its own chips as a sign the company is working to improve its cost structure over time.
Q: How Should Advisors Think About the Risk That Cheaper AI Models Could Undercut Anthropic?
A: We don't expect lower-cost, open-weight models, including those from Chinese developers, to displace premium models like Anthropic's Claude in most US enterprise use cases, largely because of trust, security, and total cost of task completion rather than price per token alone. Enterprises weigh accuracy and reliability heavily, and cheaper models often require more attempts to produce usable output, which erodes their apparent cost advantage.
Khan put it directly: "What enterprises actually care about is not really the cost per token. It's the cost per completed task, and the cost for completed task delta is much smaller than the headline price would indicate." Harrison Rolfes, Senior Research Analyst, Late-Stage Company Research at PitchBook, added that open-weight models also tend to lack the guardrails and safety track record that give enterprises confidence in high-stakes workflows, reinforcing why premium providers can sustain their pricing.
Q: Why Might Index Funds Show Different Weightings for Anthropic's Stock?
A: Weightings will vary meaningfully across index providers because methodologies differ in how they calculate market value and when they add new listings. Most major indexes weight holdings by free-float-adjusted market cap, so Anthropic's relatively small initial float will keep its weighting modest at first across most benchmarks.
Welch highlighted one notable exception: "The NASDAQ actually takes a different approach to its weighting methodology than most other indexes. It looks at things by total market cap." That means ETFs tracking the NASDAQ Composite and NASDAQ 100 are likely to assign Anthropic a comparatively larger weighting than funds tracking other major benchmarks, at least until Anthropic's float expands.
Q: How Does Morningstar Decide When a Holding Offers an Acceptable Margin of Safety for a Managed Portfolio?
A: There is no single formula. The right margin of safety depends on the nature of the underlying investment, including its volatility and stage of maturity, as well as the objective of the specific strategy or portfolio under consideration.
Dominic Papalardo, Chief Multi-Asset Strategist, explained the trade-off this way: "If the strategy is designed to be higher volatility, higher growth, those investors should expect a wider range of outcomes, so the margin of safety may be smaller to pull the trigger and add exposure in that type of strategy. Whereas something that's more conservative or income-based, the portfolio managers may require a wider or larger margin of safety." Ultimately, it comes down to a judgment call by the investment team, weighed against each portfolio's goals.
Q: Could Consolidation Among AI Model Providers Affect Anthropic's Revenue?
A: Many enterprises and consumers are currently testing multiple AI providers at once to determine which model performs best for their specific workflows, and we expect that behavior to consolidate over time as the market matures. That consolidation could reduce raw usage of any single model, including Claude, even as it increases the value enterprises place on the providers they keep.
Harrison Rolfes, Senior Analyst, Late-Stage Company Research, described the likely pattern: "What I'm starting to see is just a delegation or routing toward these premium models and then companies building their own internal models... It will ultimately affect the revenues, but the quality will remain the same, and it will force users to choose the model that works best for what they're doing." For Anthropic specifically, this suggests revenue growth may become more concentrated among high-value enterprise accounts rather than broad-based usage.

