New stocks can pop. But are IPOs right for you?

Cerebras Systems jumped 68% in its Nasdaq debut after pricing its IPO at a level that gave the AI chipmaker a US$60-billion valuation, and the offering raised US$5.55-billion, the largest U.S. tech IPO since Uber in 2019. The article frames this as part of a broader 2026 IPO surge driven by AI enthusiasm, with potential listings from SpaceX, OpenAI, and Anthropic carrying implied valuations as high as US$1-trillion, US$852-billion, and US$900-billion, respectively. It also warns that IPOs can be volatile and that retail investors may struggle to access or profit from the initial pop.

Analysis

The immediate trade is not the IPO itself but the monetization of the IPO pipeline. If AI private-market marks keep getting repriced into public-market premiums, the underwriting ecosystem — exchanges, listing venues, trading platforms, and allocation intermediaries — gets a multi-quarter volume tailwind even if individual deals are volatile. That favors NDAQ as a toll collector more than any single issuer, because the asset-light economics of more listings, more trading days, and more index rebalancing events compound without requiring perfect fundamental outcomes from the new names.

The second-order winner is incumbent AI infrastructure adjacent to the IPO boom: every marquee listing resets comp tables for compute, networking, and data-center beneficiaries, which can temporarily lift the whole “AI picks-and-shovels” basket. But that also raises the risk of capital intensity disappointment later; once public-market scrutiny hits, these companies will likely spend harder to defend growth, compressing margins across the supply chain and increasing competition for scarce HBM, packaging, and cloud capacity. In other words, the near-term signal is risk appetite; the 6-12 month signal could be margin pressure and a higher bar for post-IPO retention of first-day gains.

TD’s retail access angle is modestly positive but not enough to move the needle unless IPO activity broadens well beyond the largest U.S. deals. The real behavioral inflection is that retail investors, frustrated by allocation scarcity, will increasingly substitute ETFs and secondary-market momentum trades for true IPO access. That supports broad market cap-weighted exposure to the theme while making direct new-issue participation a poor risk-adjusted bet unless one has favorable allocation terms.

The contrarian miss is that a hot IPO window often masks weak forward returns: the more heavily oversubscribed the deal, the more likely the best economics were captured before listing. A frothy launch can actually be bearish for later entrants if it front-loads valuation and invites supply from insiders the first time lockups clear. The market may be underpricing how quickly “AI scarcity premium” turns into a float-overhang story once multiple mega-IPOs hit in sequence.

Key Decisions for Investors

• Long NDAQ on a 3-6 month horizon; use pullbacks after weak broader tech sessions to build. Risk/reward is favorable because IPO and secondary issuance activity can expand even if underlying tech beta cools.
• Avoid chasing the first-day print on future AI IPOs; wait 30-60 trading days for lockup and guidance visibility before initiating. Most of the easy upside is typically captured by underwriters and anchor accounts.
• Use a basket long in AI infrastructure beneficiaries versus short a basket of newly listed AI names after the initial pop (1-3 month trade). This isolates the pricing-power/real-demand gap and benefits from post-IPO multiple compression.
• For TD, treat this as a small positive optionality story rather than a core thesis; only add if IPO volume in North American retail channels accelerates materially. Otherwise the impact is too small to overcome bank-wide earnings drivers.
• Prefer an ETF wrapper over single-name IPO exposure for retail-style portfolios; the risk/reward is better because it diversifies deal-specific execution risk, especially into a late-cycle hype phase.