You will probably never IPO. Prepare like you might.

Is the IPO window open?

It is open at the top and shut at the bottom, and the first-half data says so precisely. Global listings fell 7.1% to 509 while proceeds rose 212.9% to $194bn; in the Americas the count fell 27.4% to 85 while proceeds rose 664.7% to $130bn (EY Global IPO Trends Q2 2026, published 7 July 2026, Dealogic data, operating companies only, as of 30 June 2026).

The window got taller rather than wider. The United States took 66% of global IPO proceeds in the first half, up from 25% across full-year 2025, yet ranked only fourth by count behind India, Hong Kong and mainland China on the same EY basis. The issuer who dropped out is the middle-market one.

On a separate count, Renaissance Capital recorded $114.7bn across 82 US IPOs in the first half, of which one company, SpaceX, raised $75bn in June. Strip that single deal and the half was $39.7bn, still close to the whole of 2025 at $44.0bn across 202 deals (Renaissance Capital, 2Q 2026 US IPO Market Review, 1 July 2026). The capital re-rating is real without the outlier. The access problem is a separate question.

One thing no source supplies is a number for next year. No named forecaster has published a quantified 2027 IPO count or proceeds figure that could be verified. EY calls the near-term environment episodic and advises issuers to stay flexible on timing. Every 2027 view available, including this one, is directional.

What is a listing actually worth to an issuer now?

Less on day one and more on day thirty than it was a year ago. The Americas class of the first half of 2026 was up 14% at offer plus one day and 11% at offer plus thirty days, against 33% and 20% for the class of 2025 (EY, 7 July 2026). A halved first-day premium is a better outcome for the seller and a thinner demand cushion for the deal.

Underwrite the median, not the mean. Across all 2026 US listings with price data the mean return is 7.6% and the median is negative 0.2%, with 54.4% trading below offer. Narrow it to operating companies and the mean is 16.4%, the median 2.3%, with 46.3% below offer (derived from the Stock Analysis 2026 US IPO table, 219 rows, as of 6 August 2026). The mean is carried by a handful of biotech and infrastructure outcomes that nobody should plan around.

The encouraging part is the direction of travel: each successive 2026 cohort has priced better than the one before it, and the twelve operating companies that listed between 1 July and 5 August are up a median 18.2% with only a quarter below offer. Seasoned listings are being sold while fresh paper is bid, which is the constructive version of a weak tape. Books are being built at conservative prices that then trade up.

For a sponsor the sharpest read-across is narrower and less comfortable. July's two sponsor-backed consumer listings priced and then went nowhere: Jersey Mike's, backed by Blackstone, closed 0.6% above its $23 offer and Reformation 7.5% above $15, while therapeutics and software names in the same month ran 29% to 66%. A sponsor exit by listing in the second half of 2026 is achievable. It should be underwritten to zero pop.

What has to be true before a listing is available at all?

The floor moved twice this year and a third change is unresolved. Since roughly 17 January 2026 the minimum market value of unrestricted publicly held shares is $15m on both the Nasdaq Capital Market net income standard and the Nasdaq Global Market income standard, raised from $5m and $8m respectively (SEC Release 34-104450, SR-NASDAQ-2025-068). Fifteen million dollars of freely tradable stock sold in the offering is the hard floor for a small-cap US listing into 2027.

Since roughly 13 June 2026, a China-based issuer must raise at least $25m gross in a firm-commitment US IPO, is barred from direct listing on the Nasdaq Global Market, and faces a $25m public float test on transfers and business combinations (SEC Release 34-105494, SR-NASDAQ-2025-069). Nasdaq's own analysis found that 143 of 151 China-headquartered or incorporated Nasdaq IPOs between April 2022 and April 2025, or 94.7%, raised under $25m and would not have qualified. That channel is closed.

The rule most likely to be reported wrongly is the one that is not in force. A $5m minimum market value of listed securities for continued listing, with immediate suspension and no cure period, was approved on 22 July 2026 and then stayed on 29 July 2026 after two parties filed notices of intention to petition for Commission review. It remained stayed as of 5 August 2026. Anyone told today that the $5m rule binds has been told something untrue.

Separately, Nasdaq may now decline an initial listing even where the applicant satisfies every published standard (SEC Release 34-104464, filed 12 December 2025, immediately effective). Meeting the criteria is no longer sufficient to guarantee a 2027 listing, and the SPAC route narrowed in April 2026 as well, with a $100m market value of listed securities test on the Global Market and a 400 public shareholder test on the Capital Market.

What is competing for the same dollars?

A dated, disclosed supply calendar that runs to August 2027. SpaceX's staged lockup releases 7% of the 180-day cohort on each of 20 August, 9 September, 24 September, 9 October and 24 October 2026, a further tranche after third-quarter earnings, and the balance on 8 December 2026; the extended cohort releases in steps from 31 December 2026 through to the second-quarter 2027 earnings date, with founder shares locked until 12 June 2027 and no early release provision at all (SpaceX Form 424B4, filed 12 June 2026). Shares under extended lock plus founder shares exceed 63% of shares outstanding immediately before the offering.

That calendar is not theoretical. A conditional early-release tranche of a further 10% required the stock to close at or above $175.50 on at least five of the ten trading days ending on the first earnings date. It closed between $108.27 and $125.33 across that window, so zero of ten days qualified and the tranche stayed locked. Performance-contingent lockups are now a term sponsors should expect to negotiate in 2027.

Behind the calendar sits a backlog that is thinner than its headline. There are 441 pending US filings, but only 116 were filed in 2026; 152 date from 2025, 55 from 2024, 34 from 2023 and 59 from 2021 (derived from the Stock Analysis filings table, as of 5 August 2026). More than seventy percent of the pipeline is prior-year paper. Withdrawals are running at 40 year to date against 46 in all of 2025, and roughly two-thirds of them were sub-$10m micro-caps.

The genuinely large names are still private and still coming. Both Anthropic and OpenAI have filed confidentially (Renaissance Capital, 1 July 2026). When they move they will absorb the available book-building capacity, which is what EY means by episodic.

What are the alternatives worth?

Two of the three commonly cited alternatives are weaker than they look. The direct listing has become a micro-cap route: twelve US direct listings through 27 May 2026 returned a mean of negative 8.7% and a median of negative 5.0% on day one, with seven of twelve negative, and the researcher who maintains the series notes that direct listers since 2022 have generally been micro-cap stocks (Jay R. Ritter, University of Florida, Direct Listings in the U.S., 2018 to 2026, table updated 5 June 2026).

Blank-check vehicles are 62% of 2026 US listings by count against 41% in 2025, and the shells being filed now are the 2027 de-SPAC class. As trend context, average one-year post-merger returns by merger cohort ran negative 64.2%, negative 63.8%, negative 59.1%, negative 62.0% and negative 57.1% for the 2021 through 2025 vintages, and average redemption rates on 2025 de-SPACs by quarter were 94.7%, 97.6%, 78.8% and 68.3%, with medians above 91% in every quarter (Ritter, SPAC IPOs, tables 12 and 15c, updated through FY2025 and 10 July 2026). Those are 2025 figures and they describe the track record a 2027 pitch runs against. A de-SPAC delivers a listing. It should not be financed as though it delivers capital.

The third alternative is the one that is quietly winning. Private secondary markets now carry 325 tracked private companies at $5.3tn of aggregate valuation, which lets an owner obtain a mark and partial liquidity without listing at all. That is the structural reason listing counts keep falling in a record-proceeds year.

What has not arrived is refinancing relief. The federal funds target range has stood at 3.50% to 3.75% since 11 December 2025 with no change at any 2026 meeting, and in the retail private credit channel two of the largest platforms rationed liquidity to their own investors: Blue Owl Credit Income Corp accepted 22.8% of shares validly tendered in its first-quarter 2026 offer and 26.6% in the second, both against a 5% quarterly cap, and Apollo Debt Solutions BDC accepted 45.2% of shares tendered in the offer expiring 16 March 2026, paying in non-interest-bearing promissory notes. Equity markets, not credit markets, are carrying the exit environment into 2027.

So what should an owner plan for?

Plan for readiness rather than timing, because an episodic window cannot be timed. The listing is a credible route into 2027 with an energy, defence, health or infrastructure story and nine-figure scale, and a difficult one without both. Energy took 24% of Americas listings by count in the first half against 8% across 2025, while technology, media and telecoms fell from 19% to 10% (EY, 7 July 2026). A software business should not model a 2027 listing on 2021 comparables.

For everyone else the realistic set is a sale, a continuation vehicle or a secondary, with the listing held in reserve for whichever window happens to open. The preparation that makes a company listable is largely the same preparation that makes it saleable, which is the argument for doing it before either decision is forced.