Ordering a company valuation also supports the project’s development.
Help develop World Market Atlas — an independent project for individual investors. Your support helps us produce analyses and improve our tools.
| Year | 2025 |
|---|---|
| Global equity capitalisation (USD) | 151,940,000,000,000 |
| Global IPOs | 1,471 |
| Capitalisation growth vs 2024 | +18.5% |
| IPO proceeds growth vs 2024 | +42.7% |
Listed domestic companies · year-end, 1990–2025
1996: 8,090 → 2025: 3,908Decline from the 1996 peak: −51.7%
| Year | Companies |
|---|---|
| 1990 | 6,429 |
| 1991 | 6,513 |
| 1992 | 6,562 |
| 1993 | 6,912 |
| 1994 | 7,255 |
| 1995 | 7,487 |
| 1996 | 8,090 |
| 1997 | 7,905 |
| 1998 | 7,499 |
| 1999 | 7,229 |
| 2000 | 6,917 |
| 2001 | 6,177 |
| 2002 | 5,685 |
| 2003 | 5,295 |
| 2004 | 5,226 |
| 2005 | 5,145 |
| 2006 | 5,133 |
| 2007 | 5,109 |
| 2008 | 4,666 |
| 2009 | 4,401 |
| 2010 | 4,279 |
| 2011 | 4,171 |
| 2012 | 4,102 |
| 2013 | 4,180 |
| 2014 | 4,369 |
| 2015 | 4,381 |
| 2016 | 4,331 |
| 2017 | 4,336 |
| 2018 | 4,013 |
| 2019 | 3,910 |
| 2020 | 4,104 |
| 2021 | 4,774 |
| 2022 | 4,642 |
| 2023 | 4,317 |
| 2024 | 4,010 |
| 2025 | 3,908 |
Data source: World Bank WDI / WFE ↗ · API ↗ · Data source · CM.MKT.LDOM.NO ↗
Off-exchange share of trading volume · three monthly snapshots
| Month | Nasdaq-listed stocks | NYSE-listed stocks |
|---|---|---|
| Feb 2005 | 29.4% | 13.0% |
| Feb 2014 | 38.6% | 34.6% |
| Jan 2026 | 51.9% | 47.0% |
Market commentary
My perspective on markets, the economy and events that matter to investors.
Will stock exchanges survive? Equity markets beyond the trading floor
27 October 2026 · Research and official-data analysis. Charts: US 1990–2025, SEC monthly observations in 2005/2014/2026; WFE statistics for 2025. Scenarios are not numerical forecasts.
Stock exchanges will probably survive, but their infrastructure may look very different. More durable than a particular trading floor or operator are the needs of companies and investors: raising capital, discovering prices, providing liquidity and maintaining credible trading rules. This is a cautious inference from research, not scientific proof that exchanges will exist forever.
1. The market is still enormous
WFE reports global equity market capitalisation of USD 151.94 trillion at the end of 2025, 18.5% above the previous year. There were 1,471 IPOs, while capital raised through them increased 42.7%. These measure different things: capitalisation values outstanding shares; it is not new cash paid into companies. Prices and exchange rates can also drive its growth. WFE, 27 February 2026
2. Fewer listed companies does not mean the end of exchanges
The official World Bank series shows US listed domestic companies falling from 8,090 in 1996 to 3,908 in 2025 — a 51.7% decline. This is not a count of all securities, ETFs or foreign cross-listings. A change of data provider in 2015 limits perfect historical comparability. World Bank/WFE, CM.MKT.LDOM.NO
Technology is not the only explanation. Ewens and Farre-Mensa find that changes in US private capital markets allowed startups to remain private for longer. This does not explain the entire subsequent fall in listed issuers. The implication for the future: exchanges must persuade companies that liquidity, visibility and access to a broad investor base justify listing costs and disclosure obligations. Ewens and Farre-Mensa, RFS 2020
3. Exchanges as reference points in a network of venues
Listing location and trade-execution location are different. SEC data put off-exchange volume in Nasdaq-listed stocks at 29.4% in February 2005 and 51.9% in January 2026. For NYSE-listed stocks, the figures were 13.0% and 47.0%. The chart contains three monthly snapshots, not annual averages. Off-exchange trading does not automatically mean unregulated activity. SEC release 34-105655, p.13, footnote 40
This division does not establish that exchanges will disappear. Foucault and Menkveld find that venue competition can deepen the consolidated order book with suitable routing. Zhu’s model shows that, under particular conditions, dark pools can improve price discovery on a displayed exchange even as some liquidity migrates away. These are conditional findings, not a claim that fragmentation always helps. One possible future is a network of cooperating and competing venues that still needs credible reference prices. JF 2008 · RFS 2014
4. Automation: better liquidity, but not every race is productive
Hendershott, Jones and Menkveld find that historical NYSE automation improved liquidity mainly for large stocks. This is not evidence about the effectiveness of today’s generative-AI agents. JF 2011
Aquilina, Budish and O’Neill find in a historical 2015 LSE sample that latency-arbitrage races accounted for about 20% of trading volume, with six firms responsible for over 80% of race wins and losses. Their estimated 17% reduction in liquidity costs from eliminating this mechanism is a counterfactual calculation, not realised savings or a current global measurement. QJE 2022, author’s materials
Budish, Cramton and Shim propose frequent batch auctions: orders arriving within the same short interval are processed together, reducing the value of tiny speed advantages and shifting competition toward price. This is a model-based proposal, not a universally adopted standard. A later model by Budish, Lee and Shim explains how revenue from selling speed can weaken operators’ incentives to reform. QJE 2015 · JPE 2024
AI can extend analytics and supervision. The FSB nevertheless identifies risks from common models, concentrated providers, cyberattacks and disinformation. Resilience and the quality of market rules may therefore matter more than speed alone. FSB 2024
5. Tokenisation changes records of rights, not market economics
Tokenisation represents asset claims digitally on a programmable platform. It may simplify ownership transfers and payment processing, but a token alone guarantees neither liquidity, shareholder rights nor safe settlement. The BIS stresses settlement money and coherent infrastructure; its framework does not always require a public blockchain. BIS Annual Economic Report 2025, chapter III
A 2026 ECB study finds an average 0.14-percentage-point lower issuance yield spread in a small, predominantly European sample of 23 tokenised bonds and 49 matched conventional bonds. It finds no statistically significant reduction in underwriting fees. This is an observational bond study, not proof of identical equity benefits or the disappearance of intermediaries. Born et al., ECB Macroprudential Bulletin 33, 2026
Regulatory changes are tangible: on 17 September 2026, the SEC announced temporary, conditional relief for certain venues trading tokenised NMS stocks. Conditions include equivalent investor rights and trading halts aligned with the primary listing exchange. This is not blanket approval for any product described as a “tokenised share”. SEC, 17 September 2026
6. Longer sessions and faster settlement have a price
On 10 April 2026, the SEC approved a Nasdaq rule change providing for 23 trading hours a day, five days a week. The approval record alone does not establish full operational launch. This is also different from 24/7 trading. SEC, 34-105199
Settlement — the final transfer of securities and money — is a separate issue. For most transactions covered by the US rule, T+1 has applied since 28 May 2024: the next business day, not precisely 24 hours. T+0 means the same day; it need not mean immediate atomic settlement. SEC, T+1
Capponi and Chang’s payment-network model identifies a trade-off: faster settlement reduces counterparty exposure but can diminish netting benefits and increase liquidity needs. Optimal speed depends on the network and funding conditions. This is a preliminary working paper, not a universal empirical finding or Fed policy. FEDS 2025-101r1, revised 2026
Editorial inference: around-the-clock availability does not guarantee an equally deep market at every hour.
7. What is most likely to survive?
Editorial conclusion: the evolution of exchanges is more plausible than the disappearance of their functions. This guarantees neither the survival of a particular operator nor returns on its shares.
In the base scenario, traditional exchanges coexist with alternative venues while integrating data, automation and new settlement arrangements. In a faster-transition scenario, tokenisation and longer sessions gain share if they deliver lower total costs, legal certainty and sufficient liquidity. In a stalled-transition scenario, cyber incidents, fragmentation or funding problems delay adoption. No numerical probabilities are assigned to these scenarios.
Assess progress through measurable outcomes: spreads and market depth, capital-raising costs, net listings after delistings, outages and secure ownership rights. Trading hours or the presence of blockchain alone cannot establish whether a market works better.
8. Method and literature
This is a selective research review, not a systematic meta-analysis. Access to Google Scholar results was restricted; publication identities and findings were checked through publishers, authors and research institutions. Peer-reviewed papers (QJE, JF, JPE, RFS) are distinguished from ECB research, a preliminary Fed working paper, and official data or announcements. Each chart preserves its observation period and statistical universe. Scenarios are editorial interpretation, without invented numerical forecasts.
Full titles, DOIs, sources and chart inputs: WMA source and calculation register. Educational material, not an investment recommendation.
Posts reflect the author’s personal views as of publication. They are not investment recommendations.