MSX US Stock Daily Watch: OpenAI's annualized revenue reveals a $20 billion discrepancy! The AI industry chain faces a revenue quality test

[MSX Research Institute · Daily U.S. Stock RWA Watch] is a signature daily report produced by MSX, a leading RWA trading platform. Leveraging strong macroeconomic research capabilities, we capture the core pulse of global traditional U.S. equities, liquidity changes, and the RWA tokenization market, helping you position ahead for quality assets.

Today's Watch

A new disclosure regarding OpenAI's revenue has triggered collective volatility across the AI supply chain.

According to the Financial Times, citing financial materials provided to investors, OpenAI's annualized revenue run rate was close to $50 billion as of the end of September 2026, significantly below the nearly $70 billion widely reported by the media at the end of September. The two figures differ by about $20 billion, a gap of nearly 29%.

Following the news, AI-related companies including Oracle, CoreWeave, Nvidia, AMD, and Broadcom broadly declined. The Nasdaq Composite fell about 1.3%, the S&P 500 dropped 0.5%, while the Dow edged up 0.1%.

It should be particularly noted that this is not OpenAI publishing a quarterly earnings report below expectations, nor does it mean the company suddenly lost $20 billion in revenue.

OpenAI remains a private company and currently does not have publicly disclosed, audited quarterly financial statements. The "$50 billion" and "$70 billion" discussed here are both annualized revenue run rates, extrapolating the most recent month's or quarter's sales pace over the next 12 months, rather than cumulative revenue the company has already recognized in 2026.

The discrepancy between the two figures largely stems from differences in accounting methodology.

Axios subsequently reported that the earlier figure of nearly $70 billion was a "grossed-up" adjustment of OpenAI's sales generated through cloud service partners, intended to facilitate investor comparisons between OpenAI and Anthropic. OpenAI's own methodology is closer to $50 billion, and different companies handle revenue from partnership channels such as AWS, Google Cloud, and Azure in somewhat different ways.

Bloomberg's follow-up report further showed that OpenAI's annualized revenue was approximately $50 billion at the end of September, but the company still expects to reach or exceed $70 billion by the end of 2026, driven primarily by enterprise business growth. In other words, $50 billion is closer to the current run rate, while $70 billion is currently more appropriately viewed as a year-end target.

Data in a Minute

OpenAI's annualized revenue run rate was close to $50 billion as of the end of September;

The figure widely cited by the market previously was close to $68 billion to $70 billion, a difference of about $20 billion, or nearly 29%;

The $50 billion primarily reflects revenue directly recognized by OpenAI;

The figure of nearly $70 billion includes a gross-up adjustment of partner sales for comparison with Anthropic;

OpenAI's overall annualized revenue run rate grew 77% in the third quarter;

The enterprise business annualized revenue run rate grew 107% in the third quarter;

OpenAI expects annualized revenue to reach or exceed $70 billion by the end of 2026; growing from $50 billion to $70 billion implies a further increase of about 40% in the fourth quarter;

In March 2026, OpenAI completed a $122 billion funding round at a post-money valuation of $852 billion; the company is reportedly in talks for at least $30 billion in new funding;

The Nasdaq Composite fell about 1.3%, and the S&P 500 dropped about 0.5%;

Nvidia fell about 3%, Oracle dropped more than 5%, and CoreWeave declined nearly 8%;

Nebius fell about 7.4%, AMD dropped about 3.3%, and Broadcom declined about 3.2%;

Micron fell about 4.3%, Coherent dropped about 9.5%, and AAOI declined about 13.6%;

Philadelphia semiconductor-related ETFs fell about 3.4% on the day, a decline notably larger than the broader market.

MSX View

This is not a "revenue miss" in the traditional sense. If a public company gave $70 billion in revenue guidance and ultimately only achieved $50 billion, that would of course constitute a serious miss. But OpenAI did not formally provide company guidance that it had "already achieved $70 billion in annualized revenue by the end of September." The earlier $70 billion figure came from media citing investor information, adjusted for partner channel revenue to create a comparable number.

OpenAI has also not announced any reduction of its end-of-2026 target. Bloomberg's latest report, on the contrary, shows the company still expects to reach or exceed $70 billion in annualized revenue by year-end. Therefore, the most accurate description is that "the market previously used an incorrect or inconsistent comparison methodology," rather than "OpenAI's revenue suddenly dropped by $20 billion."

But this does not mean the market reaction was entirely unreasonable. When a company's valuation, funding, and infrastructure commitments all reach such a massive scale, even mere ambiguity in accounting methodology is enough to raise the risk discount investors demand. If OpenAI enters the public market in the future, it cannot continue to rely solely on annualized revenue metrics without a unified definition; investors will demand to see actual recognized revenue, gross margins, operating cash flow, and long-term payment obligations.

$50 billion still represents an extremely rapid pace of commercialization. OpenAI's annualized revenue was about $20 billion at the beginning of the year and approached $50 billion by the end of September, equivalent to roughly 150% growth over nine months. The 77% growth in overall run rate and 107% growth in enterprise business in the third quarter also indicate that demand has not stopped due to intensifying competition or falling token prices. Enterprise customers are becoming one of OpenAI's most important sources of growth.

From a business quality perspective, the enterprise business deserves more attention than ordinary consumer subscriptions. Consumers can cancel their ChatGPT subscriptions at any time, whereas enterprise customers typically require data integration, security reviews, employee training, and workflow modifications. Once a model is embedded into a company's coding, customer service, data analytics, and office systems, the cost of switching vendors rises, making long-term API consumption more likely.

Therefore, what the market needs to watch next is not just user numbers, but whether enterprise contracts can translate into sustained usage, stable renewal rates, and higher gross margins.

Moreover, falling token prices are both a pressure and a growth tool. The continued decline in current model prices and per-token costs may indeed reduce revenue generated from equivalent usage. But falling computing costs do not necessarily mean the AI market is shrinking. Lower prices also incentivize enterprises to extend AI applications to more employees and more workflows, enabling applications that were previously not economically viable to finally be deployed.

This is similar to the early development of cloud computing: storage and computing unit prices declined over the long term, but usage grew even faster, ultimately driving continuous expansion of overall market revenue. For OpenAI, the core question is not how much a single token can sell for, but whether the growth rate of token usage can continue to outpace price declines, while model inference costs fall even faster. If unit prices fall 20% and usage grows 100%, revenue and gross profit may still improve. Therefore, judging AI demand cannot rely solely on model pricing; it also requires simultaneous observation of API call volumes, enterprise seat counts, paying users, inference costs, and per-customer spending.

Currently, OpenAI has become a major source of demand for global AI infrastructure investment. Its growth expectations are tied to the future orders of GPU, data center, power, optical communications, and cloud computing companies. This is why the market reaction was first concentrated in computing power providers such as Oracle and CoreWeave.

These companies need to invest substantial capital upfront to build data centers and purchase GPUs, with the related revenue to be recognized gradually over the coming years. If OpenAI's commercialization pace lags behind the speed of infrastructure deployment, investors will need to reassess equipment utilization rates, contract execution timelines, and customer payment capacity. Although Nebius did not see any new changes to its OpenAI contracts due to this disclosure, it still fell along with the sector as a highly valued, capital-intensive AI cloud service provider. Coherent and AAOI similarly did not decline because OpenAI directly cut optical module orders, but rather because the market lowered its valuation expectations for the pace of AI data center expansion as a whole.

This shows that the AI supply chain is currently highly correlated. When the revenue figures of a leading model company change, risk quickly transmits from the application layer to cloud services, GPUs, servers, power, and optical communications sectors.

About MSX:

MSX is a leading RWA trading platform dedicated to providing secure, efficient, and transparent access to global financial markets. As one of the world's earliest on-chain U.S. stock trading platforms, MSX has always been at the forefront of the industry, leading market transformation through continuous innovation.

The platform deeply integrates blockchain technology with a compliance framework, offering comprehensive spot and derivatives trading of nearly 400 tokenized stocks and Pre-IPO assets, perfectly bridging the gap between traditional finance and the digital asset industry.

Centered on the core mission of "enabling quality assets to circulate freely," MSX has now built a diversified digital financial services system covering U.S. stock spot and perpetual contracts, crypto-to-crypto trading, Pre-IPO, and the MSX Research Institute, aiming to provide global investors with round-the-clock, high-performance access to quality assets.

Official Website: https://msx.com/

Chinese TG: https://t.me/MSXOfficial

Global TG: https://t.me/MSXGlobal

X: https://x.com/MSX_CN

Now available simultaneously on the global App Store / Google Play.

Risk Disclaimer: Macroeconomic conditions and U.S. stock markets are highly volatile. This content is provided solely for academic and research observation reference by the MSX Research Institute and does not constitute any investment advice.