OpenAI Burns US$14B/Year at US$850B Valuation: The Capital Logic Behind "Losing More, Worth More"

You use ChatGPT every day, but don’t realize you’re helping prop up the valuation of a company that burns US $14 bi l l i o nann u a l l y . O p e n A I^{'} s f u l l - y e a r r e v e n u e i sr o ug h l y * * U S$ 3.7 billion**, but its expenses exceed US$14 billion. In other words, for every dollar it earns, it spends nearly four dollars.

By the logic you and I were taught growing up, this company should have collapsed long ago, right? But what’s the reality? Its valuation has rocketed from US $29 bi l l i o n l a s t y e a r t o U S$ 850 billion now — up nearly 30 times. The more it loses, the more it’s worth. This isn’t a financial fairy tale; it’s real capital operations happening right now.

You’re surely asking yourself: what’s the logic here? Is it a bubble, a scam, or just a game whose rules I don’t understand?

This article shares information for financial education purposes only and does not constitute any investment advice. Each person’s financial situation, risk tolerance, and investment experience are different; what works for someone else may not work for you. Please make all investment decisions independently and consult a properly licensed Taiwan financial advisor and tax professional.

Core Logic: Valuation Has Never Been About How Much You Earn Today

Remember this: in capital markets, valuation has never been about how much you earn today — it’s about pricing how much you might monopolize tomorrow.

You may only understand half of this today, but after finishing this article you’ll understand it fully.

You’ve Used the Wrong Tool: Job-Hunting Logic vs Investment Logic

Let’s start from a scenario familiar to any Taiwanese office worker. You’re job-hunting on 104人力銀行 (a Taiwanese job platform) and see a startup offering higher pay than big companies, better benefits than government agencies, but it’s only three years old and has never made a profit. Would you go?

Most people’s gut reaction is no — it’s not stable, the company might collapse. That gut reaction may be right when job-hunting, but if you transfer the same intuition to investment markets and use it to judge whether a company is worth investing in, you’ll systematically miss the biggest wealth opportunities of the past 20 years.

Why? Because you’re using cash-flow logic to judge a target whose logic is monopoly pricing — these two frameworks are completely different dimensions. You’ve used the wrong tool.

Cash-flow logic vs monopoly-pricing logic

Rule 1: Loss vs Burn Are Two Different Things

Taiwanese ordinary investors have three core cognitive blind spots on this topic. The first blind spot, and the one beginners most often fall into, is treating loss and burn as the same thing.

You see OpenAI burning US$14 billion a year, and your first reaction is “this company is losing money.” But “loss” has a very precise financial definition. Taiwan’s FSC and IFRS define loss as the net negative result after a company subtracts costs and expenses from revenue over a period.

That definition itself is fine. The problem is that the structure of the burn determines whether this loss is consuming assets or building them.

What is most of OpenAI’s US$14 billion in spending? It’s compute investment, R&D salaries, model training costs, infrastructure build-out. These show up as expenses on the financial statements, but in capital-market valuation logic they are building a technical moat that others find harder and harder to replicate.

You can think of it as a company spending a fortune to build a highway others can’t enter. During construction it’s definitely spending money, definitely showing losses on paper, but once the highway is done, everyone who needs to pass through has to pay you, and you can keep collecting.

The Taiwan Financial Research and Training Institute (TABF) explicitly stated in its 2023 tech-stock valuation research report that for tech platform companies, a significant portion of early-stage losses belong to strategic front-loaded investment; the nature of such spending is building future pricing power, not pure value impairment.

Rule 2: Valuation Is Pricing Tomorrow’s Monopoly Rights

If you use “cash flow” logic to look at OpenAI, you’d say it’s a terrible company: revenue US $3.7 bi l l i o n, l oss U S$ 14 billion, net outflow over US$10 billion — by P/E ratio this company shouldn’t even have a positive valuation.

But if you use “monopoly pricing” logic, you see a completely different story:

OpenAI’s ChatGPT is currently the world’s largest AI dialogue product, with paid users numbering in the millions, and enterprise customers including tech giants such as Microsoft and Salesforce. As AI applications expand across customer service, content generation, programming, data analysis, and medical diagnostics, OpenAI has the opportunity to become the underlying infrastructure.

What capital markets are pricing isn’t OpenAI’s revenue this year — it’s the value of OpenAI potentially becoming the “operating system” or “utility company” of the AI era 5 to 10 years out. That imagination space is the core logic behind the US$850 billion valuation.

Extending the Highway Analogy

Imagine Google back in 2000. Google was burning cash every year then — investing in data centers, R&D on search algorithms, expanding the business. If you used cash-flow logic, Google’s valuation in 2000 should have been very low. But if you used monopoly-pricing logic, you’d see Google had the chance to become the internet’s gateway — and we all know what happened later.

OpenAI’s position in AI is similar to Google’s position in search back then. This isn’t saying OpenAI is guaranteed to succeed — it’s saying capital markets are willing to pay a high premium for this kind of “future possible monopoly” position.

The first blind spot is the one we just mentioned — treating loss and burn as the same thing. What you really need to look at isn’t the loss amount, it’s the loss structure. If the loss is used to build a moat, the loss is valuable; if the loss is used to maintain daily operations, that’s the dangerous kind.

The second blind spot is using job-hunting logic to evaluate investments. When job-hunting you need to know whether the company can pay salaries next month, so stability matters most; but when investing you need to know how much of the market the company can monopolize in the next 5 to 10 years, so imagination space matters most.

The third blind spot is only watching Taiwan’s public markets and ignoring private-market opportunities. OpenAI’s valuation surge happened primarily through private markets, including Microsoft’s US $10 bi l l i o nin v es t m e n t an d T h r i v e C a p i t a l^{'} s U S$ 6.6 billion lead round. Ordinary Taiwanese investors find it very hard to participate in these private rounds directly — that’s a structural information gap.

How Should You Approach “Losing More, Worth More” Investment Opportunities?

Step 1: Clarify Your Investment Logic

Are you investing with cash-flow logic or monopoly-pricing logic? If the former, you should avoid all loss-making tech stocks; if the latter, you have a chance to profit from companies like OpenAI.

This isn’t a question of right or wrong — it’s a choice. But you must clearly know which logic you’ve chosen.

Step 2: Participate Through Public-Market Substitutes

Although ordinary Taiwanese investors can’t directly participate in OpenAI’s private rounds, they can indirectly participate through public-market substitutes:

Microsoft: is OpenAI’s largest shareholder; OpenAI’s growth directly reflects in Microsoft’s market cap
NVIDIA: is the main supplier of AI compute; the more prevalent AI applications become, the higher the demand for NVIDIA hardware
TSMC: is the main foundry for AI chips like NVIDIA’s
Taiwan’s AI supply chain: includes Taiwan vendors in thermal, CoWoS packaging, HBM, and other fields

This isn’t saying you must buy these stocks — it’s telling you the paths to participate in the AI wave.

Step 3: Build Your “Future Monopoly” Screening Criteria

Next time you see a “losing more, worth more” company, screen with these 3 questions:

Is this company’s loss being used to build a moat?
Does this company have the chance to become the underlying infrastructure of some industry?
Is the total addressable market (TAM) of this market large enough?

If 2 or more of the 3 questions answer yes, the target may be worth your further research.

The 3-question screening flow for "losing more, worth more"

Advanced Content: Two Key Concepts You Need to Understand

Concept 1: Strategic Loss vs Operating Loss

Strategic loss refers to loss that a company actively chooses in order to seize market share, build a moat, and cultivate user habits. The characteristic of this loss is: as market share expands, the loss shrinks and may even turn into profit.

Operating loss refers to loss caused by an unworkable business model, runaway costs, or revenue not keeping up with spending. The characteristic of this loss is: even as market share expands, the loss continues to widen.

OpenAI falls under strategic loss, because each additional user brings in additional revenue, while compute investment has economies of scale — the more users, the lower the unit cost.

Concept 2: The Imagination Space of Future ARPU

When capital markets value OpenAI, what they’re really pricing is its future average revenue per user (ARPU) and total user count.

OpenAI’s current paid user count is in the millions, but if AI applications普及 to every office worker using an AI assistant, every enterprise using AI customer service, every programmer using AI coding tools, the total user count would be hundreds or even thousands of times the current level.

Same with ARPU. Today’s US $20/ m o n t h s u b scr i pt i o n co u l d in t h e f u t u r e u p g r a d e t o U S$ 100/month or more for enterprise plans. Multiply that imagination space and you get the source of the US$850 billion valuation.

4 Steps to Take Today

Step 1: Build Your Own Investment Logic

Write down explicitly whether you’re using cash-flow logic or monopoly-pricing logic to invest. If you choose cash-flow logic, then stick to buying blue chips and ETFs and don’t touch loss-making tech stocks. If you choose monopoly-pricing logic, be prepared to tolerate greater volatility.

Step 2: Research Public-Market Substitutes in the AI Supply Chain

Trace OpenAI’s growth path backward and identify which public listed companies will benefit from the AI wave. Microsoft, NVIDIA, TSMC, and Taiwan’s AI supply chain are all candidates worth researching.

Step 3: Use 3 Questions to Screen New Targets

Every time you see a “losing more, worth more” company, screen with “loss structure + moat + TAM”. If 2 or more of the 3 questions answer no, don’t touch it.

Step 4: Control Your Position, Don’t All-In

Even if you’re very confident in a company’s future, don’t all-in. AI is a long-term trend, but short-term stock price volatility will be very large. Keep each single position within a reasonable share of your portfolio.

Conclusion: You Don’t Need to Understand OpenAI, but You Need to Understand This Logic

You don’t need to know how OpenAI operates, but you need to understand the capital logic behind “the more it loses, the more it rises”. Because that logic isn’t just for OpenAI — it applies to every tech company that’s building a future monopoly position.

If today you don’t understand this logic, you won’t just not understand OpenAI — you’ll also, for the next 10 years of investment decisions, miss opportunity after opportunity, or worse, repeatedly pour money in at the wrong time for the wrong reasons and get trapped.

What capital markets price isn’t how much you earn now, but how much you might monopolize in the future. Once you can look at every target through this lens, your investment horizon will be completely different.

Last line for you: rather than looking at how much money a company is burning now, look at how deep and wide the moat it’s burning into is. The deeper and wider the moat, the stronger the future pricing power — that’s the real reason losing more means being more valuable.

Disclaimer: All content in this article is information sharing for financial education purposes only and does not constitute any investment advice. Each person’s financial situation, risk tolerance, and investment experience are different; what works for someone else may not work for you. All investment decisions should be self-assessed based on your financial situation and risk tolerance, and you should consult a properly licensed Taiwan financial advisor and tax professional. Any numbers and simulations mentioned in this article are estimates based on specific assumptions and do not guarantee actual investment outcomes. Investing always carries risk, and past performance does not guarantee future results.