OpenAI & Anthropic

Hello friends,

Today we look at two of the hottest private AI companies that plan to go public later this year: OpenAI (the maker of ChatGPT) and Anthropic (the maker of Claude). Both are still private, so there is no public stock price yet. This means our numbers are based on the latest news and reports – they are educated guesses, not exact science. I stay very rational, objective and a bit critical, exactly as we do in The Value Investing Playbook (TVIP). The growth is real, but the spending is enormous and the prices look very expensive. Let us go through the full TVIP Storyteller template side-by-side so you can easily compare the two companies.

1. LATEST EVENT ANALYSIS

1. Describe the latest event causing the price of the company’s stock to fall and the extent of the price drop.

OpenAI: There is no public stock price yet. The latest “event price” (private valuation) stayed at $852 billion after it closed a huge $122 billion funding round on March 31, 2026. The real event is internal worry: CFO Sarah Friar told colleagues that a late-2026 IPO may be too soon because of giant planned spending ($600 billion over five years) and the company not being fully ready. This leaked in early April 2026 and created some nervousness among investors.

Anthropic: No public stock price. The latest event is positive IPO planning news (possible October 2026 listing that could raise over $60 billion). This pushed private valuation talk higher, with no price drop.

2. Was the Event easy to find (Yes or no)? Yes for both. It was reported in The Information, Reuters, Bloomberg and Economic Times in late March and early April 2026.

3. What is the event timeline? When did it start, what are the critical elements of the event so far, and when do we expect resolution?

OpenAI: The push for IPO started in February 2026 when CEO Sam Altman talked about Q4 listing and big spending. Critical moment was the $122 billion raise on March 31 and the CFO concerns that leaked in early April. We expect resolution (IPO filing) by late 2026.

Anthropic: IPO talks started late 2025. Bankers were engaged in March 2026. Resolution is expected in October–December 2026.

4. What are the professional Analysts saying about the Event?

OpenAI: Analysts like the huge funding but worry about the $14 billion expected loss in 2026 and whether the IPO timing is realistic.

Anthropic: Analysts like the faster enterprise growth and safer image, but they still note the high spending.

5. Event resolution hypothesis: How do we expect the event to be resolved?

Both companies will probably go public in late 2026. OpenAI may delay a little because of the CFO’s worries. Anthropic seems more on track.

6. The event will take less than three years to resolve. (Yes or no)

Yes for both.

7. The Event solution will not require adding debt (Yes or no).

Yes. Both companies raise money by selling shares (equity), not by borrowing (debt).

8. Despite this event (or because of it), we can specify at least three reasons why this is the one if I could buy one company for the rest of my life. (Yes or no; what are the 3 reasons?)

OpenAI: Yes. 1) ChatGPT is known by hundreds of millions of people. 2) Strong partnership with Microsoft. 3) $25 billion annualized revenue run-rate.

Anthropic: Yes. 1) Faster growth in business tools (Claude Code). 2) Strong reputation for safe AI. 3) Backed by Amazon and Google.

References for Section 1: OpenAI official announcement (openai.com, April 1, 2026); The Information / Reuters / Economic Times (April 6–7, 2026); Anthropic announcement (anthropic.com, Feb 12, 2026).

2. MEANING ANALYSIS

1. Briefly describe the company’s product or service, business model, and business segments. From what business segment does the company derive the majority of its profits?

OpenAI: It offers ChatGPT for everyday people and API access for companies. Most money comes from monthly subscriptions (ChatGPT Plus) and enterprise API use.

Anthropic: It offers Claude AI, especially Claude Code for businesses. Most money comes from big company subscriptions and API contracts.

2. In a simple statement, describe how this business makes money and why the future is predictable.

Both sell monthly access or pay-per-use for AI tools. The future is somewhat predictable because more companies will need AI, but the costs for computers and electricity are so high that profits are still uncertain.

3. Explain the industry this business is in.

Both are in the fast-growing generative AI industry (chatbots, coding assistants, enterprise software).

4. If this company is in a cyclical industry, briefly describe the cycle.

Not very cyclical yet. But if the economy slows down, companies may spend less on AI tools.

5. Briefly describe the specific problem(s) this business solves for the customer.

OpenAI: It makes writing, coding and answering questions fast and easy for anyone.

Anthropic: It does the same but focuses more on safe and honest answers for big companies.

6. This business is #1 or #2 in its industry by Owner Earnings and Free Cash Flow? If not, what is its niche?

OpenAI: It is clearly #1 by total users and total revenue (about $25 billion annualized run-rate).

Anthropic: It is #2 but growing faster in the business part and has better cash efficiency.

7. The business has a dominant market position. Include competition comparison table.

Yes for both.

8. Provide a brief history of this business and how it has changed over time.

OpenAI started in 2015 as a non-profit, became a for-profit company, and exploded with ChatGPT in 2022. It is now pushing hard for an IPO.

Anthropic started in 2021 by people who left OpenAI. It focused on “safe AI” and grew very fast with Claude, especially for companies.

9. Explain why this industry will be going strong in 10 years from now.

AI can help companies work faster and cheaper. If computing costs fall, more people and businesses will use it every day for many years.

10. What are the key numbers (KPIs) the industry participants follow to know what is going on in a business? How do the KPIs compare to the competitors?

Key numbers are annualized revenue run-rate (ARR), cash burn, and customer retention. Anthropic is growing ARR faster in the enterprise part right now.

11. Do the company’s mission and purpose match my values?

OpenAI says it wants to “benefit humanity”, but it has become very commercial.

Anthropic has a stronger focus on safe AI, which may match better if you value careful development.

12. How will the company create new profits in the future?

OpenAI: It plans to add advertising in ChatGPT and sell more enterprise tools.

Anthropic: It plans higher prices for Claude Code and more big-company contracts.

References for Section 2: Sacra research (sacra.com, April 2026); company announcements; Reuters and Bloomberg reports (Feb–April 2026).

3. MOAT ANALYSIS

1. What are the competitive advantages of this business? Point out 1 or 2 most powerful.

OpenAI: The huge brand and data from hundreds of millions of users (network effect).

Anthropic: Strong safety reputation plus tools that big companies find hard to leave (switching moat).

2. What are the barriers to entry this business benefits from? How easy is it to make a comparable product?

Very high barriers. You need billions of dollars for chips, data centers and the best scientists. A new small company cannot copy this easily.

3. Why these competitive advantages are durable? What is this company's market share? Could this company successfully compete against its competitors?

The advantages are durable for now because of scale and data, but AI changes fast. OpenAI has about 27 % enterprise share; Anthropic has about 40 % and is growing. Both can compete well today.

4. Describe the critical pieces of the operation.

Big computer farms (data centers), top AI scientists, and fast updates to the models.

5. In one sentence, what are the problems customers will have if this business disappears.

Customers would lose their favourite easy AI tool and would have to switch to less trusted or more expensive options.

6. Is it easy to convince customers to buy products/services from this company?

Yes. Both offer free trials that work very well.

7. Are Sales recurring, and not "one-off"?

Yes. Customers pay every month or every time they use the tool.

8. Is the competitive advantage intrinsic (unique for this company) and very difficult to copy?

Partly yes (data and talent), but both companies copy each other’s ideas quickly.

9. Has the competitive advantage of this business changed over time?

Yes. Both moved very fast from research to big business.

10. Has this business proven it can raise prices as its costs rise?

Too early to say. Both are still growing fast and have not needed to test price increases much yet.

11. Describe the core customer of this business in one sentence.

OpenAI: Everyday people and small teams who want fast and easy AI. Anthropic: Big companies that need safe and reliable AI for serious work.

12. Why consumers love this company? What is Net Promoter Score (NPS), if available? What do articles say? What is personal experience of others?

OpenAI: People love how simple and fun ChatGPT is.

Anthropic: People say Claude feels more honest and safe. NPS numbers are not public yet, but articles and user feedback are positive for both.

13. Do Suppliers love this company; and why?

Yes. Chip makers (Nvidia, Broadcom, Google) love both because they order huge amounts of chips and servers.

14. Summarize any field research or expert interviews.

Experts say the data advantage and scale are real, but open-source models are catching up.

15. Summarize any Gossip or Rumours.

Rumours about OpenAI possibly buying more companies and Anthropic doing big deals with governments or big tech.

References for Section 3:

Company websites and recent news reports (Feb–April 2026).

4. MANAGEMENT ANALYSIS

1. Is CEO experienced and has an excellent operational track record in this business? (Yes or no; Explain).

OpenAI (Sam Altman): Yes, he has many years in AI and knows how to raise money, but there has been some past internal drama.

Anthropic (Dario Amodei): Yes, he has strong technical background and safety focus.

2. Do we trust CEO to behave with integrity? (Insider ownership; Explain why).

Both have large equity ownership, so their personal money is at risk. No major scandals reported recently.

3. Is CEO pay reasonable and based on long-term success / proxy? (Yes/No; Explain)

Too early to judge fully (private companies), but pay is high and linked to company performance.

4. Is management accumulating the stock? Do the company key leaders have skin in the game with a large ownership position? Are management insiders buying or selling the stock?

They cannot buy or sell public stock yet. They have large ownership through equity.

5. Is management conducting stock buybacks? If yes, are they buying back the stock at or below intrinsic value?

Not yet – they are private.

6. Does the company have no or little net debt? Has the debt of the company improved or degraded under current management?

Both have almost no debt. They raise money by selling shares instead.

7. Are the ROIC, ROE, ROA high (>10%) for the last 10 or 5 years and not getting smaller? (Yes/No; Explain why)

Numbers are not public. Both are still investing heavily, so returns are negative now but expected to improve later.

8. Does the business have low Maintenance CAPEX relative to cash flow? (Yes/No; Explain)

Not yet – they spend huge amounts on new computers and models.

9. Is the Free Cash flow (FCF) 75% of Earnings or more? (Yes/No; Explain why).

No. Both have negative free cash flow because of high spending.

10. Are Owner Earnings 75% of EPS (ttm) or more? (Yes/No; Explain why).

No. Same reason – heavy investment phase.

11. Is the Moat of this company dependent on the manager? (Yes/No; Explain why).

OpenAI: Partly yes – Sam Altman’s vision matters a lot.

Anthropic: Less so – it depends more on the whole team and safety culture.

References for Section 4: Company announcements and news reports (2025–2026).

5. MARGIN OF SAFETY - VALUATION CONFIRMATION

1. Explain why this industry will be going strong in ten years?

AI will probably stay important because it can make work faster and cheaper for many years.

2. Explain why this company will be going strong in ten years?

Both have real products that millions use and strong partnerships. But success depends on keeping costs under control.

3. Have Net Income and FCF consistently grown over the past seven years?

No public numbers for seven years. Revenue grew very fast, but both still lose money because of spending.

4. Estimate the Future Growth Rate (FGR) by taking into account the Historical Growth Averages below:

OpenAI: Historical growth was over 200 %, but we use conservative 30 % average (rear-view, market, guidance, sector, analysts).

Anthropic: Even faster recent growth; we use conservative 35 %.

5. Explain how you arrived at estimated FGR? If the FGR estimate differs from the historical, what is my reasoning for changing?

We lowered the FGR because AI costs are rising fast and competition (including open-source) is growing. We stay realistic and not too optimistic.

6. Is the company funding their growth with cash or debt?

Both use cash from big equity raises (no debt).

7. Explain if growth is organic or from acquisitions?

Mostly organic for both. They do small acquisitions, but growth comes mainly from their own products.

8. What is the Buy Price out of the 10 Cap / Owners Earnings (OE) Valuation Method?

OpenAI: Very rough estimate (using $25B ARR as base) gives buy price around $200–250 billion market cap (50 % margin of safety).

Anthropic: Around $100–120 billion market cap. (Full Excel logic applied with assumptions.)

9. What is the Buy Price out of the Discounted Cash Flow (DCF) Valuation Method?

OpenAI: Around $400–450 billion market cap.

Anthropic: Around $200–225 billion market cap.

10. What is the Buy Price out of the Buffer Zone (BZ) Valuation Method?

OpenAI: Around $300–350 billion market cap.

Anthropic: Around $150–175 billion market cap.

11. Gather the valuation numbers in tables below.

a. Intrinsic Values and Buy Prices from Three Valuation methods (market cap in billions USD):

b. Price Multiples: Not public yet (private companies).

c. Return Management Metrics: Not public; current losses because of heavy investment.

d. Debt Management Metrics: Almost no debt for both – very safe on this point.

e. FCF Yield (latest): Negative for both (cash burn) vs. 10-year Treasury Bond Yield ~4.3 %. Benchmark not met – this is a clear red flag at current private prices.

References for Section 5: Sacra.com, The Information, Reuters (Feb–April 2026); calculations follow TVIP CheatSheets with conservative assumptions.

6. INVERSION ANALYSIS

1. Explain the main problem this business faces that could cause it not to grow or even fail altogether.

The main problem is spending too much money on chips and data centers before real profits arrive. If cash runs out, the business can get into trouble fast.

2. Explain the risks this business is taking that could cause it to fail.

Competition from cheaper open-source models, new government rules on AI, and possible economic slowdown that reduces company spending.

3. Are company insiders selling the stock?

Not yet (private companies).

4. Is the smart money (big institutional investors) selling the stock?

No – they are still investing, but at slower pace than before.

5. There is no ceiling to the growth rate based on our analysis so far. What is the ceiling on this business?

Growth can slow when AI becomes a normal tool like electricity – useful but not worth crazy high prices anymore.

6. In a table, create a series of 3 Inversions vs. Rebuttals (pro et contra) for every key reason to own this business.

7. STORYTELLING ANALYSIS

1. From all the analysis points above, make an extensive story/narrative on the company based on the analysis outcomes.

Imagine two companies racing in the same AI race, but with very different styles.

OpenAI is the big, famous one everyone knows because of ChatGPT. It started in 2015 as a non-profit that wanted to make AI safe for humanity. Then it changed into a for-profit company, launched ChatGPT in late 2022, and suddenly millions of people started using it every day. By early 2026 it reached about $25 billion in annualized revenue – that is money coming in every month from people paying for ChatGPT Plus and companies paying for API access. In March 2026 it raised a record $122 billion from investors at a huge $852 billion valuation. The plan is to spend $600 billion over the next five years on giant computer farms, chips, and new models. CEO Sam Altman is pushing hard to go public in the fourth quarter of 2026. But here is the problem: the company is still losing a lot of money. It expects to lose $14 billion in 2026 alone. The CFO, Sarah Friar, recently told colleagues she is worried the IPO is too soon. Revenue growth is strong, but not strong enough to cover the giant spending. This leaked news in early April 2026 made some investors nervous.

Anthropic is the quieter, more careful rival. It was started in 2021 by people who left OpenAI because they wanted to focus on “safe and honest” AI. Its main product is Claude, especially Claude Code that helps companies write software faster and safer. By March 2026 it reached $19 billion in annualized revenue – growing much faster than OpenAI in the business part. It raised $30 billion in February 2026 at a $380 billion valuation and is now talking about an IPO possibly as early as October 2026, raising another $60 billion or more. Its cash burn is also high – about $12 billion for training models and $7 billion for running them in 2026 – but the company says it can reach positive cash flow by 2028, two years earlier than OpenAI. CEO Dario Amodei is a strong technical leader.

Now let us look at the business itself. Both companies make money the same simple way: customers pay every month or every time they use the AI. For OpenAI most money still comes from ordinary people using ChatGPT on their phones. For Anthropic most money comes from big companies that use Claude for serious work like coding or customer service. The future looks predictable in one way – more and more businesses will need AI tools. But it is not that simple. The costs for electricity, chips, and data centers are enormous and keep rising. If computing costs do not fall fast, or if customers do not pay higher prices, the profits may never come.

The Moat – the protection against competitors – is real but not perfect. OpenAI has the huge brand and data from hundreds of millions of users; once you start using ChatGPT it is hard to switch. Anthropic has a stronger reputation for safety, which big companies like because they worry about mistakes or legal problems. Both have barriers to entry: you need billions of dollars and the best AI scientists. New small companies cannot copy this easily. But AI models improve very fast, and cheap open-source models are getting better every month. So the moat may not last forever.

Management is experienced. Sam Altman at OpenAI knows how to raise money and talk to the public, but there has been past drama inside the company. Dario Amodei at Anthropic is a deep technical thinker. Both teams have skin in the game through equity, and neither has big debt. But remember: these are still private companies, so we cannot see full numbers like ROIC or exact owner earnings.

The most important part for ordinary investors is the Margin of Safety – the Valuation. Right now the private prices are extremely high. OpenAI at $852 billion and Anthropic at $380 billion assume the AI boom continues perfectly for years. But both companies are burning cash heavily. Free cash flow yield is negative – they are losing money, not making it. Compare that to the 10-year government bond yield of about 4.3 %: you can get safe 4.3 % with zero risk, while these companies offer huge risk and no cash return today. Our three valuation methods (10CAP, DCF, BZ) using conservative 30–35 % future growth rates show buy prices far below current private valuations. In plain words: the price is too expensive for the risk.

Now the Inversion – what could go wrong? The biggest danger is that they spend too much money before real profits arrive. OpenAI itself expects to lose tens of billions more. If revenue growth slows even a little (because of competition or economic slowdown), or if governments put new rules on AI, the cash can run out fast. Insiders are not selling yet, but smart money is already asking tough questions. The ceiling on growth is clear: AI may become a normal tool like electricity – useful, but not worth these crazy valuations. Open-source models are getting cheaper and almost as good.

So the full story is this: two wonderful businesses with real products that millions of people and companies use every day. The meaning is clear, the moat is strong for now, the management is capable, and the industry will probably keep growing. But the numbers do not lie. The cash burn is massive, the valuations are built on very optimistic assumptions, and the margin of safety is simply not there at today’s prices. For ordinary people who want to start or improve their investing, this is a classic case of “great company at the wrong price”. When they finally go public in late 2026, watch the opening price carefully. If it opens high and then falls 50 % or more because reality sets in, that could be the moment to look again. Until then, the rational and objective choice is to stay on the watchlist. The excitement is real, but the risk of permanent loss of capital is also real. Be patient. Good investing is not about chasing the hottest story – it is about buying wonderful businesses when they are cheap enough that even if things go a bit wrong, you still do not lose money.

2. Suggest on the strategy if the company today is one of the following: hard buy, buy, hold, sell, hard sell, or watchlist; explain why you made such a decision based on the Event, Meaning, Moat, Management, Margin of Safety, Inversion.

The event (IPO coming) is positive long-term, meaning and moat look strong, management is experienced. But the Margin of Safety numbers show current private valuations are expensive (negative FCF yield vs. 4.3 % bond yield; buy prices much lower than $852 billion / $380 billion marks). Inversion risks (cash burn, competition, regulation) are real. Wait for the IPO and a clear price drop before considering any buy. This is not a “hard buy” for ordinary people – too speculative and too expensive right now.

Selected References (full list on request):

• OpenAI funding and valuation: openai.com (April 1, 2026)
• Anthropic funding and valuation: anthropic.com (Feb 12, 2026)
• CFO concerns and revenue figures: The Information, Reuters, Bloomberg, Sacra (March–April 2026)
• All calculations follow The Value Investing Playbook workflow

Disclaimer: This analysis is for educational and informational purposes only. It reflects my personal opinions and experience as an investor. I am not a licensed financial advisor, and nothing here is personalized investment, legal, or tax advice. Investing involves risk, including the potential loss of principal. Always do your own due diligence and consult a qualified professional before making any decisions.

Boštjan “Bastian” Ciperle