In a month full of big stories, perhaps the biggest business story was the fall from grace of Jensen Huang and the consequences for Nvidia, investors, and the rest of the AI trade. The initial story, with its potent mix of personal fallibility and security concerns, related to Jensen’s affair with Alice Huang, who news stories labeled a Chinese spy, though the evidence on the ground is still thin. I had planned to write about the story then, but on the road in Asia and with teaching commitments, I put it off for a few days, and over those days, the story took a turn, with stories of dalliances with Nvidia employees. Last week culminated with Nvidia’s board asking Jensen to step down, and his agreeing, perhaps bringing to an end one of the more eventful and productive stints as a CEO in corporate history. I have no desire or standing to debate the moral aspects of this story, and I will let others take sides on whether the board should or should not have fired Jensen. In this post, I will focus on the question of what Jensen stepping down as CEO might mean for Nvidia’s value, and whether after the market correction over the last three weeks, the market has overreached.
What does a CEO do?
What does a CEO do for a business? Given the many tasks that go into creating, building, and running a business, it should come as no surprise that the list is long, but I will classify them broadly into groups:
Storyteller: Every business is built around a story, and that story drives its valuation and sometimes its pricing. While that story is grounded in the company’s products and the markets that it seeks, it is the top manager’s job to craft and tell the story to investors, employees and to consumers. Clearly, the significance of this component of a CEO’s story will shift, depending on where the company is in its life cycle, with its importance being greatest at either end of the life cycle.
Business manager: It is of course CEO’s job to manage the business that they are put in charge off, but that can span the spectrum from building a business model, for young companies, to working on ways to scale up a business for high growth firms to defending a business model for mature businesses to scaling down the model, as the company enters the declining phase.
People leader: Organizations are composed of people, including suppliers, employees and customers, and it is the top management’s job to lead these people, though the nature of that leadership will also shift as a company ages. Early on, when a company is still seeking out a business model and faces failure risk, it is to inspire people with vision and direction. As companies grow and mature, it is to create a corporate culture that retains employees, as competitors try to lure them away. In decline, when a company is shedding employees and shrinking, it is finding a way to do so as humanely as possible and find a soft landing, without destroying morale.
Public face: For better or worse, the CEO represents the public face of the company with investors, regulators and others, and the role they play will vary across groups and across the life cycle. With investors, CEOs frame the vision/narrative for young companies and are key to drawing in fresh capital, whereas with mature firms, especially if they are publicly traded, managers interactions are with institutional investors, where they try to manage expectations and frame results. With regulators and politicians, the CEO’s role is sometimes to play defense, when their companies are in the crosshairs and facing regulatory questioning, and sometimes to play offense against competitors.
Succession planner: There is a final aspect of a CEO’s job that is often ignored but can be key to extending a company’s life cycle, and that is preparing for a succession, where a new management is readied, for a handover, when it comes due. Clearly, the importance of creating succession plans increases as top managers age, and especially so at companies that are transitioning from one stage of the life cycle to the next.
In sum, while the list of roles that CEOs play and the functions that they provide is long, the role and functions that get emphasized will shift as companies move through the life cycle. Early in the life cycle, it is the visionary component that is front and center, moving on to business building and people leader as the company's business model takes form, and as the business scales and matures, the primary mission becomes playing defense (against competition, regulators and governments), and managing succession.
How much does a CEO matter?
A hundred years of formalized business education has been built on the premise that CEOs matter, and that having good management at the top of a business can make the difference between success and failure. While I do not disagree with that broad premise, the effect of good management can be greater at some firms and smaller at others, with five determinants:
Macro versus micro: Businesses that are driven more by macroeconomic movements, such as those in interest rates or commodity prices, will be less affected by who manages them than companies whose success can be traced to company-specific actions on which products to produce, how to price them and where to market them. Thus, management will matter more at a consumer product company than at a commodity company since the former’s success will depend upon its management decisions a great deal more with the latter.
Competitive advantage: The nature of the competitive advantages possessed by a business can also affect how much management matters. In companies with long-standing and legacy competitive advantages, management has more of a caretaking role, than at companies that constantly need to reinvent their competitive advantages. While we do not intend to downplay the challenges that managers at Coca Cola and Aramco face, when managing the company, they face less of a challenge than managers at Costco, a company in an intensely competitive business facing disruption.
Transition: Management matters more when businesses are at transition points, i.e., start-ups that are seeking out venture capital for the first time, or companies on the verge of an initial public offering or mature/declining firms ahead of a restructuring, since management actions can not only make a significant difference between success and failure, but investor reactions can also be governed by how much they trust management.
Upside and Downside: There is one parameter that management at companies can be judged upon, and that is whether they operate in businesses where success is keyed by whether you create upside or by protecting against downside, with the former falling under the category of taking advantage of opportunities and the latter under risk management. In young companies, it is the former that will dominate, but as companies mature, the latter will become more important.
Corporate Life Cycle: At the risk of generalizing, management matters more at start-ups and very young companies, since it must deliver not just on vision but on business-building than at mature firms, that have strong competitive advantages and are in good financial position, where management can be on autopilot. As firms go into decline, management will matter again, since management in denial or desperation can cost shareholders a great deal.
In sum, the CEO (and top management) always matters, but the extent to which it matters will vary across companies, with the life cycle playing a significant role.
The Value of a CEO
Given my obsession with value, it should come as no surprise that my discussion of CEOs begins by looking at the many ways that they can affect value. As I identify the multiple key person value drives, note that not all key people affect all value drivers, and the value effects can also vary not only widely across key people, but for the same key person, across time. At the risk of being labeled as a one-trick pony, I will use my intrinsic value framework, and by extension, the It Proposition, where if it does not affect cash flow or risk, it cannot affect value, to lay out the different effects a CEO can have on value:
For personnel at the top, and I include founders and CEOs, the effect on value comes from setting the business narrative, i.e., the story that animates the numbers (revenue growth, profit margins, capital intensity and risk) that drives value., and that effect, as I have noted in my earlier discussions of narrative and numbers, can be all encompassing. The effects of people lower down in the organization tend to be more focused on one or two inputs, rather than across the board, but that does not preclude the effect from being substantial. A salesperson who accounts for half the sales of a business and most of its new customers will influence value, through revenues and revenue growth, whereas an operations manager who is a supply chain wizard can have a large impact on profit margins. As someone who teaches corporate finance, I have always tried to pass on the message, especially to those who are headed to finance jobs at companies or investment banks, that of all of the players in an organization, finance people are among the most replaceable, and thus least likely to be key people, except perhaps at firms where distress is a real and imminent danger and finding financial pathways to survival (through restructuring and reorganizing) becomes the defining value trait.
With this framework, valuing a key person or persons becomes a simple exercise, albeit one that may require complex assumption. To estimate key person value, you value the company twice, once with the key persons included, with all that they bring to it’s cash flows and value, and then again, without those key persons, reflecting the changes that will occur to value inputs:
Value of key person(s) = Value of business with key person - Value of business without key person
The CEO valuation approach, while general, can not only yield different values for different CEOs, but also generate a value effect that is negative for a CEO whose influence has become malignant.
The framework can also help explain how the value of a CEO can evolve over time, from a significant positive at one stage of an organization to neutral later or even a large negative, explaining why some key people get pushed out of organizations, including those that they may have founded.
Valuing Jensen at Nvidia
Until a few weeks ago, investors were valuing Nvidia on the assumption that Jensen Huang would continue to run the company for an extended period. As the founder of the company and the architect of its rise to a mega cap company, he was admired for his business-building prowess, but over the last five years of the AI boom, he has been far more than the CEO of a company providing the chips for the AI build-up. Investing in some of the companies in the AI chain and providing financing (directly or as guarantees) for others, he has nurtured the AI cap ex boom from which Nvidia benefits. In this section, I will start with an update on the Nvidia CEO story, before moving on to an assessment of where Nvidia is in the life cycle, I will value Nvidia, with Jensen heading it and no sandal on the horizon, as a base value, before exploring how the scandal and his stepping down as CEO could affect that value.
Updating the Nvidia CEO Storyline (so far)
The first story about Jensen's affairs broke on October 29, with the story of his relationship with Alice Huang, with mentions about a China connection. In the following week, the story stayed in the headlines, with news of more affairs with employees who worked at Nvidia and reported to him, a violation of corporate rules. With the White House chiming in, the board had no choice but to act, and after an emergency meeting, Jensen announced on November 12 that he was stepping down as CEO. During this period, the market reacted to the news, with Nvidia losing a trillion and a half dollars in market cap:
The board's appointment of Collette Kress, Nvidia's CFO, as interim CEO has done nothing to quell the scandal, as the debate has turned to who the company would hire as a permanent CEO, with the twist that Jensen Huang could return in that role.
Accepting that Jensen Huang has been integral to Nvidia’s success so far does not mean that you can conclude that his exit as a CEO will decrease Nvidia’s value significantly, because that will depend on things that still are undecided.
The first is that while there is an interim CEO in place, a replacement has not been announced yet, and to estimate what Nvidia will lose in value from Jensen’s exit requires that we know the identity of the person who will replace him at the top of the company (and an assessment of how well he will fill that role)>
The second is that it is worth remembering that even after he steps down as CEO, Jensen will own a sizable stake in the company, with allies in the company, and it is unclear how much he will work with the replacement CEO; the more willing he is to provide guidance and direction, the smaller the value effect should be.
The third is that this will be a test of the strength of the management team that Jensen has built at the company, and whether it can run the company without him directly calling the shots.
Finally, this is no longer just an intra-company issue, since the Chinese spy angle has opened the doors to congressional hearings and legal consequences, both of which can have effects on Nvidia’s operations, at least in the near term.
Hanging over all of these uncertainty is the question of whether some of Nvidia's biggest investors and aggrieved employees will engineer a return for Jensen, and what that will do for the company.
Where is Nvidia in the life cycle?
In terms of chronological age, Nvidia is not a young company, at least by tech standards. It was founded in 1993 by Jensen and went public in 1999 and while it has enjoyed a niche status, it was AI that pushed the company to the top of the ladder, as its operating metrics (revenues, earnings) and market cap soared in the investing boom that has followed. In the chart below, I look at Nvidia’s entire history as a publicly traded company, reporting revenues, gross profits and operating income from 1997 to 2026:
There have been periods in Nvidia’s history where it looked like it was converging on being a mature growth company, but it has found ways to rediscover high growth markets in gaming and crypto, in the last decade, and AI in this one. In fact, the last five years, as Nvidia has grown its revenues at 69.42% a year from 2021 to 2026, and seen its gross and operating margins explode, it has become a unicorn – a company that has managed to look like a young growth company, but doing so at scale and profitability.
As founder of Nvidia, and CEO for its entire existence, Nvidia’s story is intertwined with Jensen Huang’s and it is undeniable that the company is constructed in his image – opportunistic in seeking out new markets and nurturing relationships with other players in the ecosystem for their benefit. I know that the board’s decision to ask Jensen to step down as CEO has evoked strong reactions on both sides. Those in support of the board’s action note that Jensen’s affairs with employees who reported to him was a flagrant violation of corporate policy, coupled with the security lapses in his other affair, merited the response. Those who oppose the action argue that the it is unclear whether Ms, Huang was a Chinese spy and that Jensen is far too critical not just to Nvidia as a company but also to the AI disruption that the company should keep him on. Both sides, though, should agree that Jensen has been an extraordinary chief executive and that Nvidia would not be the company it is today, without him at the helm.
Valuing Jensen Huang’s value to Nvidia
The first step in assessing Jensen’s value to Nvidia is to value the company with him still at the helm, an none of the scandal that has unfolded in the last three weeks.. In the picture below, I have an updated valuation of Nvidia, with updated numbers through the second quarter of 2026, from October 29, 2026:
My estimate of the value of equity for Nvidia, prior to the events of the last few weeks, is $4.8 billion, roughly $200 per share, a significant jump over my estimates of its equity value from a year prior, but a reflection of the company’s capacity to keep growing at scale and preserve profit margins. In arriving at this estimate of value, I have incorporated the effects of Nvidia’s financing of other AI players, with Nvidia’s equity holding in Anthropic added to value and the present value of almost $500 billion in commitments that Nvidia has, either on its own or for partners, in debt. On October 29, the day that the story of Jensen's affair broke, the conpany's market capitalization was $5.2 trillion, about 8% above my estimated value.
Now that Nvidia's market capitalization has been marked down to $3.5 trillion, you may be tempted to conclude that the stock is significantly undervalued now, but that assumes that Nvidia's value has not changed as it has lost its CEO, and struggles with the aftermath of the scandal. To assess the effect of Jensen’s affairs and his departure from Nvidia, we look at two dimensions. The first is on the short term impact, with revenue growth and operating margins affected over the next year, and the resulting effect on value. The second is the long term impact, where revenue growth and margins in the long-term change, with a much greater effect on value:
In the short term, the Jensen affair can have three consequences. One, and I think that this is unlikely, is that Nvidia sees a drop off in orders from customers, affecting near term revenues. The second, perhaps more likely, is that a transition to a new CEO might delay or hold up decisions that affect revenue generation or create costs in the near term, thus putting pressure on profit margins. The third is that the legal and legislative blowback, especially with the China twist to the story, can have regulatory or legislative consequences that affect next year’s operations. No matter what the reason, if the effects of Jensen's departurre are truly temporary, the percentage change in Nvidia’s equity value is contained, with a 5- 10% drop in most scenarios.
In the long term, there is the possibility that without Jensen at the helm, Nvidia will operate differently, perhaps pulling back on financing of customers and easing the relentless marketing that has made them the unstoppable force in the AI business. The long-term effect on value will depend on how much revenue growth rates and margins decline from the base case:
While the wide range of value effects may seem like a copout, where you will land up on this table will depend on where you fall in terms of the uncertainties that I listed at the start of the section:
Will Jensen Huang Return? While it seems unlikely that the board of directors at Nvidia will reverse their decision to fire Jensen, the door is not completely closed. Spurred on by Jensen's stated willingness to come back and pushback from large shareholders in the company who feel that the company needs him, there is a chance that he will return. (Kalshi estimates a 10% chance).
If replaced, how good will his replacement be? If Jensen is out as CEO, the final choice on who will be Nvidia’s next CEO is up in the air with internal names (including Colette Kress, the interim CEO and Ian Buck, VP for hyper scale), CEOs of other chip companies (Renee Haas, CEO of ARM, Lisa Su, CEO of AMD as well big name ex-CEOs of other large companies (Tim Cook at Apple) all in the mix. The board will be balancing the need for someone who understands the AI chip business with that of a crisis leader, who can manage the company through trying times. Rather than focus on resumes or reputations, the board should be looking for a good fit for Nvidia, as it stands today. Tim Cook, in my view, is not a good fit for Nvidia today, unless this mushrooms into a crisis and he takes the role as a temporary CEO. While knowledge of the chip business and Nvidia's internal workings is a good ingredient to have in the replacement, the real skill will be finding someone who can preserve the delicate balance of preserving an astounding profit model, built on marketing Nvidia as the indispensable piece for compute power, and a growth model, that may require financially supporting potential customers for its chips.
If replaced, how much will Jensen be willing to work with his replacement? Closely linked to the choice of replacement CEO will be the role that Jensen plays in Nvidia in the months or perhaps even years to come. If the replacement CEO is someone that has worked with Jensen before, and feels comfortable listening to advice (and pushback) from him, it is likely that the value effect will be much more muted.
How strong are the layers of management at Nvidia under the CEO? In the thirty four years that Jensen has run Nvidia, he has built a team, with a relatively flat organizational structure. From all accounts, the people that he has chosen to run Nvidia seem smart and competent, and many are intensely loyal to the company. The test will be whether they can find ways to deliver the same results without the spark of Jensen’s intuition and long term vision.
Does the China story have legs? For me, the wild card in this story is the China twist. If Alice Huang does have official ties to the Chinese government and Nvidia has been using business fronts in Asia to sell its chips to the Chinese government, in violation of US law, the consequences will not only be more damaging but also persist no matter who is CEO and perhaps become worse if Jensen returns to his post.
If you believe that the board will find a competent replacement CEO for Nvidia, that Jensen will stay involved, even informally, in big decisions, that the strong management team will stay intact and that the China twist is overblown, the value drop will be small. If, as the weeks go through, the process becomes ugly and contentious, there is a real risk that you could get a CEO mismatch, burn bridges with the company’s founder and lose management personnel, resulting in a much greater loss of value, with any China-related scandals adding to the loss. If you want to try your hand on how much value Nvidia loses from this sordid episode, please open this spreadsheet and make your own judgments on short and long term effects on growth, margins, reinvestment and risk. In fact, the template can be used to value top management at any company, and by extension, it is a value of control.
As you wait for these uncertainties to resolve, you cannot sit still, since the market will likely continue to try to adjust based on expectations. I do think that if the market continues to mark down the stock, there will be a point at which you will find the stock to be a buy, notwithstanding unresolved uncertainties. For me, that value is about 30% below my estimated value of $4.8 trillion, yielding a breakeven value per share of around $140 per share, and I have a limit buy waiting at that price. If it happens, I will have Nvidia back in my portfolio, and if it does not, there are other stocks to buy!
A Governance Postscript
Jensen Huang has been a great CEO, but last week was a reminder that he is human, and that the same can be said about Elon Musk, Dario Amodei, Sam Altman and the other luminaries who are ensconced at the top of the biggest companies in the AI space. Thus, it is a given that they too will make mistakes, some of which they will be reluctant to confront, and even more unwilling to reverse. The essence of corporate governance is that even the very best managers need restraints to prevent overreach, and that giving them complete power in perpetuity is a recipe for regret. There is also a truth about long tenure as CEO that few are willing to acknowledge. A CEO that is newly hired may be humble and willing to listen, but the longer that CEO stays entrenched, the more likely it is that he or she will accumulate power, create boards that are rubber stamps and acquire an ego, completing the transition to Imperial CEO.
Finally, Nvidia’s travails should make us more appreciative of founders who build companies that can outlast them and find ways to make themselves dispensable. You may not like Bill Gates, but he left Microsoft, while still in his forties, and the company barely missed a beat and has gone on to scale new heights. The pictures of Jeff Bezos, at his wedding or in his speedos on a yacht may make you cringe, but to the extent that it does not make you stop buying on Amazon Prime or sell your Amazon stock, he has succeeded in building an entity that can outlive him.
I valued Anthropic in a post on November 2, 2026, and while estimating a value of $1.475 trillion for the equity in the company, I noted that the share count was a placeholder, while we waited for bankers to make a final judgment on how many shares would be outstanding in the company. I promised that I would return to complete the per share valuation, when the details were made available, and an updated prospectus filed last week filled that gap. The world has not stayed still in the fifteen days since my original valuation, and both company-level and macro developments require an update to my valuation. The prospectus was updated to include third quarter numbers, and the midterm elections have potentially altered the economics of the AI business, as data centers may become more difficult and expensive to build and AI regulations may become more stringent. In this post, I will start with a general discussion of how share count is complicated by the existence of multiple classes of shares and stock-based compensation, and argue that the absence of standardization has made share count the Wild West in investing, with consequences for both intrinsic valuation and pricing. I will then update my valuation of Anthropic and assess the per share value, based on the prospectus that was just filed.
The Elusive Share Count
When I first started valuing companies, getting from aggregate value to equity value per share was a trivial exercise, where you divided the aggregate value by the number of shares outstanding. This was, of course, a period when almost every US company had only one class of shares and stock-based compensation was still in its nascent stages, and lightly used. That is no longer the case, and while tedious, I think that the process of understanding share count has to start with the basics of what goes into that number for publicly traded companies.
Primary share count: The primary share count is obtained by looking at the shares in the company, across all classes of shares. That information is usually embedded in the balance sheet or in the footnotes, but with different numbers usually for authorized, issued and outstanding shares. The authorized share count is the least relevant of the numbers and is just a measure of the total number of shares that the company can issue, with an increase beyond it requiring a fresh authorization. The shares issued are a measure of the shares that the company issued to the market and while shares outstanding match shares issued for most firms, the presence of treasury stock (stock acquired back by a company, but held by the company) can cause outstanding shares to be lower than issued shares.
Restricted shares: Restricted shares comprised shares that have been issued by the company to its employees, though in rare cases, they can also be issued to the public. The restrictions take two forms. With employee issuances, there can be a vesting requirement, where an employee has to work for a specified period to be able to lay claim to the shares. The other is that even after vesting, these shares sometimes can come with restrictions on liquidity; i.e., the holders of these shares cannot trade the shares easily.
Options: The third claim on equity can come from options granted by the company to employees and sometimes lenders, with the latter manifesting as conversion options in convertible bonds or convertible preferred stock. The former are part of employee compensation and come with specified maturities, and often require vesting, just like restricted stock. The latter are included with bonds, and used to raise money from investors, with the conversion option usually running for life time of the bond, giving holders of the bond the right to swap their bond for shares in the company. Options give you the right to buy shares in the company at a fixed price (the exercise price), and as a consequence, they are different from primary and restricted shares in two ways. The first is that if the company's stock price drops below the exercise price (making it out of the money), it does not make sense to exercise the option, thus keeping it out of the share count. The second is that even when the stock price is higher than the exercise price (making it in-the-money), holders may not exercise the option, because there may be an advantage to holding the option for longer, hoping for a bigger payoff. This creates a time premium in the option, where the option is worth more than what you would get from exercise today. When you use an option pricing model, you are in effect trying to incorporate the current stock price, the exercise price, the likelihood that you will exercise the option and the expectations of future stock prices into a number.
For some older and more mature companies, it is possible that the only shares outstanding are the primary shares, and getting a share count is easy. For most companies, though, and especially young firms that use stock-based compensation heavily (because they don't have the cash to pay employees), restricted shares and options outstanding can make a significant difference. For these companies, there are often two limiting numbers that define the extremes of share count. The first is just the primary share count, which will be the lowest of the share count numbers. The second is the fully diluted share count, where you count not just shares outstanding today, but all shares that could be outstanding if the restricted stock is vested and all options (employee and conversion) get exercised, giving you the highest share count. There is an intermediate solution of partially diluted shares, where rather that count all restricted stock and options, you may count only the vested shares (for restricted shares) and the in-the-money options (for employee options).
Share Count in Intrinsic Valuation
Since intrinsic valuation is about estimating how much you value individual shares in a company to make a judgment on whether a stock is under or over valued, the share count clearly matters, since having more shares outstanding will yield a lower value per share. There are three ways in which you will see analysts and investors deal with this overhang:
The Bludgeon: In the bludgeon approach, you let simplicity be the driver of your choice and go with either the primary shares or the fully diluted shares, with the values per share you get reflecting the outer ranges of your per share value estimates. The primary shares approach, with the lowest share count, will yield the highest value per share, and in companies with restricted stocks and options outstanding, that value per share will be too high.
Primary share value per share = Aggregate Equity Value/ Primary shares outstanding
The fully diluted shares approach, with the highest share count, will yield the lowest value per share, and many investors prefer it, contending that it is better to be conservative.
Fully diluted share value per share = Aggregate Equity Value/ Fully diluted shares outstanding
The problem with this approach, though, is that it counts all options, in and out of the money, as part of the share count, and not only assumes immediate exercise, but that the company will not receive the exercise proceeds, leading to an even greater underestimation. It is true that using a partially diluted share count will provide a partial solution, by removing out-of-the-money options but the remaining problems will persist.
The Refined Bludgeon: The second approach (called the Treasury stock approach) tries to remedy the failure to include exercise proceeds from option exercise in value per share, by adding those proceeds to the value of equity, and doing so will push your value per share, relative to a fully diluted share value:
Treasury stock value per share = [Aggregate value of equity + Strike Price × Options outstanding]/ Partially diluted shares
Here, you use only the options that are in the money, because out-of-the-money will give you a value boost by bringing an exercise price higher than the stock price that is fantasy. The problem with this approach is that it ignores the time premium on options, effectively undervaluing those options, and will result in a value per share that is too high.
The Scalpel: As I noted earlier, options are different from any other share category, because they may or may not be exercised, depending on the stock price, and there is uncertainty about when they will be exercised, even when in-the-money. Consequently, the right approach in valuation is to estimate the value of options, using an option pricing model) and then subtract that value from aggregate equity value to arrive at a equity value for just shares outstanding (primary and treasury). Dividing that residual value by a share count that includes only primary and restricted shares will come closest to capturing the correct value per share:
Option-adjusted value per share = [Aggregate value of equity − Value of options outstanding]/ (Primary + Restricted Shares issued)
It is true that to value options, you will need information on the exercise price and maturity, but that is required disclosure now at US companies. With this approach, you no longer have to separate in-the-money from out-of-the-money options, since both will have value, and the option pricing model will adjust the value for the exercise price (relative to stock price) and the time left on the option.
The picture below summarizes these contrasting estimates of value per share, using different share counts:
Again, rather than view these as competing estimates, I would suggest viewing the option-adjusted value per share as the number closest to the truth, and the rest as approximations, with the primary and the fully diluted estimates not even in the running.
Having different classes of shares with divergent voting rights and claims to cashflows adds a layer of complexity to this process, since to bring those differences into play, you have to make a judgment on how much voting rights and cash flow claims matter, and incorporate those judgments into values per share that will be different for different share classes, with voting shares valued at a premium relative to non-voting shares.
Anthropic: Intrinsic Value per share
The Anthropic updated prospectus lays out the gory details, and they are gory, of share count in this company:
As you can see, there are multiple classes of shares and I am not even bringing in the class T shares that the trust holders or the founder share, with 50.01% of the voting rights in perpetuity. There are five details relating to shares outstanding that are worth emphasizing.
The IPO will result in the issuance 801 million additional class A shares that are added to the share count.
There are 1026 million restricted shares, and given that they have a four-year vesting period, none are vested yet.
There are 1,774 million options outstanding, all likely to be deep in-the-money (with a strike price of $2.91, well below the expected offering price of $70-78 per share), with 1,082 million vested already. There are eight years of option life left, indicating that there will be a time premium, albeit a small one.
The company has made a commitment to donate shares under a matching plan, once it goes public, with 933.7 millions tagged for the plan
The conversion options in the convertible bonds should normally be separated out and treated as options, but if I read the prospectus correctly, it looks like they will be converted shortly after the IPO into shares.
With those caveats in place, the share count ranges from 14,875.18 million shares, if you count only shares outstanding (and don't count the donation shares, the shares on conversion of the convertible bonds, restricted shares or options) to 17,676.18 units, if you have a fully diluted count. You have intermediate solutions, with variants based on whether you include vested units or not), but the value per share will vary widely, depending on the count.
To arrive at the value per share, I started with a valuation of the aggregate equity in the company, using the storyline I developed more fully in an earlier post, but I made four changes;
I updated the base year numbers (for the last twelve months) to include the third quarter (of 2026) data that was provided in the updated prospectus. The effect on overall value is muted, because while there was a slowing down in the annualized revenue run rate in the third quarter, that was already known and incorporated into the earlier story and valuation.
I adjusted the cash proceeds down from $100 billion, a rumored value at the time of my original valuation, to about $60 billion, based on an expected offering price of $74 and the disclosed shares that they plan to issue (801 million). That lowers the post-IPO value of equity by the difference of $40 billion.
I do not want to overreact to the election results, but I do think it is prudent to assume that building data centers will become more expensive (captured by lowering the sales Anthropic will get per dollar of capital invested in the first five years from $3 to $2.75. This increases expected reinvestment and lowers cash flows).
It is unclear whether and how a new Congress will regulate AI, but here again the prudent path seems to be that the drag on margins will get larger, from 5% to 6%, lowering the target margin to 26.5%.
The resulting value for the firm and its equity is below (updated to incorporate the $83 offering price set after this post was written):
The resulting aggregated equity value is $1,351,556 million, with the drop in estimated value mostly a result of the updated compute costs and regulatory drag. To convert that number into an equity value per share, I use all of the approaches from the last section:
Note that while there is a range of values here, from $76.46 (fully diluted) to $90.86 (primary), the value that I would use is the option adjusted value per share of $76.. The offer price has not been set, at the time of this post, but the real comparison should be to the price that Anthropic will open trading at tomorrow. (Right after this post was written, the news story reported that the offer price has been set at $83/share. With that offering price, the value per share drops slightly for all of the calculations).
Note that I have studiously avoided trying to estimate the value per share, by share class, in spite of the fact that the shares differ in voting rights. That is because of the company's uncommon voting structure, where the voting rights reside in a single share owned by the founders. The closest parallel I can think of for this structure are the golden shares that governments put into government-owned utilities and companies that were privatized in the 1980s and 1990s. That golden share gives governments veto power over major decisions, but not power over the day-to-day operating choices made by a company. Anthropic's founder share is far more powerful, and I do think that there is a danger in loading up this much power in a single share. I don't want to give governments and regulators ideas, but I can visualize an activist regulator, finding Anthropic at fault for an AI accident, and using public safety as a rationale for expropriating the founder's share to make the company a government-controlled utility. I am sure that the founders will try to get the courts to step in, claiming (rightly) expropriation, but why tempt fate?
Share Count in Pricing
You may be uninterested in intrinsic value and focused on pricing Anthropic in the market, but you will still run into share count in multiple places. The first is in the market cap, a number that gets used widely to rank firms and talk about valuations, where using different share count measures can lead to very different market cap estimates. The second is in computing pricing multiples, where market cap becomes an input into estimating enterprise value or is used a stand alone. The third is in computing per share numbers for earnings, revenues or any other metric, since per share requires the use of a share count; the primary earnings per share will be very different from fully diluted earnings per share.
Almost every financial data news service will report the market capitalization as part of the core descriptives for a company, and companies are often ranked based on market capitalization. Given that reality, it is surprising how much variability there is across services on reported market capitalizations for some of the highest profile companies in the market:
Market capitalizations, end of trading day on 11/18/26
The differences are small in this group, but as you look at younger and high-growth firms, especially with large stock-based compensation overhangs, the differences will widen. If you are comparing across companies, you may believe that using the same per share metric (primary, diluted, partially diluted etc.) for all the companies in your group will yield comparable numbers but that is not true, since there can be wide variations in the use of restricted stock and options across companies. Since these market capitalizations are fed into enterprise value and earnings per share calculations, it should come as no surprise that EV multiples and PE ratios will vary across services, for the same company at the same point int time.
So what is the correct market capitalization for a company? I can offer an answer, but I can also guarantee you that there is no single service that I know of that following this practice. It is to take the option-adjusted approach in valuation and flip it around to get a market capitalization:
Option-adjusted Market Capitalization = (Share price × Number of primary and restricted shares outstanding) + Value of employee options and conversion options, at current stock price
This will yield values that allow you to compare companies with very different option counts, as well as differences in how in-the-money the options are at each company.
Once Anthropic goes public tomorrow and has a listed market price, you should expect to see these share count inconsistencies come into play, as services will report different market capitalizations for Anthropic, using the share price, and different share counts. As with intrinsic valuation, the purely primary share based estimate and the fully diluted estimates will both be off for the company, with the former too high a number and the latter too low a number. You could settle the difference by valuing the options (my spreadsheet has a worksheet that will do that for you), and coming up with the correct value.
Conclusion
Most investors and analysts pay little heed to share count, and how it is computed, and with mature firms, that oversight may have only minor consequences. As your search leads you to look at firms earlier in the life cycle, I would recommend paying attention to the footnotes, especially related to restricted stock and employee options, and if you have convertibles, to how the conversion option is being accounted for. With Anthropic, it is the accounting for the surge in the value of the conversion option, as its pricing increased in private markets, that led to the $74 billion one-time and non-cash charge to earnings in 2026. I have much to do tomorrow that is not market-related and I will not be watching the Anthropic open, but hopefully, you will and keep me posted!
When I posted my valuation of Anthropic a few days ago, I noted that turbulence lay ahead, both for the company and the market. I did not expect that turbulence to be this forceful and arrive this soon, but markets don't have an obligation to wait until it is convenient for investors, to adjust to new realities. With the election results on November 3 acting as a catalyst, investors initially responded positively with stocks up on Wednesday, but reversed course sharply on Thursday, with the NASDAQ down close to 6% and the S&P 500 declining 4%. This drop has again brought to the forefront questions about the AI trade and whether it has run its course.
In this post, I will start by referring to my earlier note on AI as a business, and use the ideas in it to deconstruct the many players that are lumped together as AI companies, from those spending the money building the data centers (hyperscalers) to the firms benefiting from selling the builders the ingredients for these data centers (chip makers, electrical equipment, power and even real estate) to the companies that are spending considerable amounts , leasing these data centers for long periods (with the LLMs in the lead). I will argue that while all of these groups benefit, at least in general, from AI's success at creating large and profitable markets, we are at a point in the AI disruption, where their interests and payoffs will diverge, with implications for both the businesses and investors. I use last week's market carnage in technology and examine how the market treatment of these different groups may be a reflection of a more discerning market, less willing to mark up and mark down all AI companies, and more likely to treat them differently, as their business interests diverge.
The AI Transition
My journey on AI started with a selfish and narrow intent, which was to evaluate an investment I had made in Nvidia in 2018, for reasons entirely unrelated to AI. I thought that Nvidia, at the time, was a innovative chip company with more pricing power than many in its peer group, and felt that it was undervalued, given its prospects. While the stock outperformed my portfolio early on, it was the ChatGPT debut on November 30, 2022, and the subsequent AI buzz that put its stock price on steroids. That left me with a challenge, albeit a welcome one, of deciding whether I should continue to hold on to a stock that had already become a twenty-bagger or sell some or all of it. Since much of Nvidia's supersized success was being propelled by AI demand, that required me to take my first look at how big the AI market would be, and how well Nvidia was positioned to take advantage of that market. If you have read my multiple posts on Nvidia over the last three years, you may be aware that I sold all my shares in the company in four installments, with the last sale happening earlier this year. As many of you have been quick to point out, my decision to gradually decrease my Nvidia position has led me to leave money on the table, but I have no regrets.
The Players
In the process of trying to understand AI and disentangle the many players in the AI space. I break them down broadly into four groups:
AI Suppliers: While AI is viewed as a technology business, we have learned over the last few years that it is much more capital intensive than any other tech endeavor in history. The data centers and infrastructure that are needed to power AI agents and models has created a surge in demand for ingredients. The biggest portion of the cost, at least so far, has been on high-powered chips, with Nvidia claiming a lion's share, and enough left over demand for other chip companies to benefit. This infrastructure build out has increased demand for electrical equipment, with power and real estate companies also gaining as data centers get built.
AI Builders: Alphabet, Amazon, Meta and Microsoft, four companies that were part of FANGAM, the acronym that predates the Mag Seven, bought into the AI hope, and joined by a handful of other players (Oracle, Coreweave), they have invested hundreds of billions of dollars into AI cap ex (data centers and infrastructure). The market cut these hyperscalers slack early on, rewarding them for their capital spending, even though that spending was denting profitability at these companies and demolishing cash flows, and part of the AI trade question has been whether the market is running out of patience, waiting for payoff.
AI Products: In the last two years, companies which are on the frontlines of creating AI products and services, have became the focal point of investors, albeit in the private investing world, with talk of public offerings pricing them in the trillions of dollars. SpaceX became the first of these to list, but Anthropic's IPO is imminent and OpenAI will join, sooner rather than later. There are other companies (Palantir is the most prominent), albeit only a handful, that are have made AI a central component of the products and services they sell, but that number will be added to, as private businesses in the space go public.
AI Customers: There is a fourth group, perhaps more critical than any of the above, that will determine the value creation potential and long-term standing of this business, and that includes all of the potential customers, individual as well as businesses, for AI products and services. While the premise for AI hype is that its reach (both in terms of numbers and monetization) will be huge, it is undeniable, even if you are an AI diehard, that there is little of substance to show at the moment. Even if you take the ARR numbers that are being leaked by the lead players, the collective revenues from AI products and services right now is about $250 billion, a small number relative to what has been invested in the space. The entire AI ecosystem is built on the premise that this group will become much larger, both in terms of reach across businesses and money spent in the trillions of dollars.
The companies in all of these groups can be classified as being in the AI business, but they generate value in very different ways from that business.
The Linkages
There are links that hold these groups together, and I have tried to capture those linkages and how they play out as revenues, expenses and earnings in each group:
This simple, and perhaps even over-simplistic, breakdown of AI does, in my view, provide a good starting point for understanding why their interests and market pricing is more likely to diverge in the future, as the business takes form. The linkages between the AI groups come from three core transactions between these groups that determine not just their current earnings but also their long term value propositions:
Building the AI infrastructure: The AI builders are spending the money building the infrastructure and much of it is being legitimately classified as capital expenditures, though there is a discussion to be had about the life of that infrastructure. The reason that the investment life matters is that it determines the amortization schedule for the AI cap ex, and consequently the negative impact on earnings in the current year; the free cashflows are lower since they take a direct hit from the cap ex. The suppliers report these same capital expenditures as revenues in the current year, and when netted out against their operating costs, increase earnings at these companies. I wrote about how the net effect of the infrastructure buildout has been an increase in aggregate earnings at US companies in the last year, and has supported stock prices, even as interest rates have soared.
Leasing the AI infrastructure: The AI architecture, once built, is leased out by the builders to the businesses that use the data centers to create and support the products and services they offer. The builders report the lease payments as revenues, allowing them to earn a return back on their infrastructure build, but for the AI product and service businesses, these payments are either operating expenses, if annually reset, or financial expenses, if structured as contractual commitments over many years; the latter will create imputed interest expenses and depreciation.
Creating and selling AI products: Using the leased AI architecture, AI product and service businesses create products and services that they sell to consumers (individual and enterprise) and record the payment they receive, from either subscription or usage-based models, as revenues. With the LLMs (OpenAI, Anthropic), these products are AI agents and frontier models where users pay either in subscriptions or based upon usage, and net of the costs of creating these products, hopefully into profits.
The linkages between the companies in the AI ecosystem described in the section above are operating linkages that are part and parcel of doing business, but there are financial linkages, though, the motives for which I will examine later in this post, that are messier. In particular, AI companies have funded each other, with both equity and debt. Microsoft and Nvidia are shareholders in OpenAI, Alphabet and Amazon in Anthropic and Nvidia has also provided financial backstops for data centers being built out. These financial cross holdings create accounting noise (with mark ups of equity investments reported as earnings) and contribute to opacity.
The Timing of Benefits
Given that AI infrastructure can take years to come online, there will be timing differences across the groups, in terms of seeing the benefits from their business models, with AI suppliers being the first to see revenues and profits, followed by AI builders, with AI product & service businesses generally being last to see profitability.
1. Hope and Hype
In the first phases of a disruption, the visions of a huge market often drive actions on the part of all of the AI groups, but the way they play out and how they show up in results will vary:
The suppliers of the ingredients that go into AI infrastructure will be the first to see tangible benefits as the spending on AI architecture shows up immediately as revenues in income statement and as incremental profits.
For AI builders, no matter how merited the build-up (in terms of delivering value over time), the immediate effect on both earnings and cash flows will be negative. The reason the earnings effects has been muted with AI is because the biggest AI infrastructure investors (Meta, Microsoft, Alphabet and Amazon) have earnings and cashflows from legacy businesses and the cloud businesses (thus playing a role asiAI suppliers) that have provided a buffer, but the free cash flow impact has been real.
For companies in the AI product group, which are nascent businesses, there will be little to show in terms of revenues, since neither the infrastructure nor the products are ready to go. With R&D and employee compensation expenses, they will be money-losing machines.
For AI, the hope and hype phase was most pronounced in 2023 and 2024, and not surprisingly, the biggest surge in market capitalization during the period was at Nvidia, as its chips became the essential ingredient to get investments going.
2. Investment Buildup
In this phase, the AI infrastructure building continues at a torrid pace, but portions of it are complete and ready for use, and the consequences again will vary for the groups involved:
As growth in AI infrastructure continues, AI suppliers will continue to see growth in revenues and profit from the continued building, though growth rates will come down with scale. In addition, new entrants may start to compete on prices, reducing pricing power at least for some of the supplier groups.
For Ai builders, the portion of the AI that is ready to use can be leased out, allowing for a monetization of at least a portion of the AI investment. Since the investment continues to increase, and much of it is still not operational, on measures like return on capital, the AI buildup investment will not measure up. Markets will often be willing to forgive, at least with managers that they trust, on the assumption that things will get better over time.
As AI infrastructure comes online, AI product companies can start generating products and services, and to the extent that they are growing from small scale, growth rates can be exponential. Those revenues though will generally be dwarfed by expenses from product development, continued research, and employee compensation, leaving the companies with large losses.
With AI users, a mix of early adopters and experimenters, in a subset of industries where AI is viewed as having the most promise, will emerge. Given AI’s strengths, software businesses were among the first to be exposed to AI products, with consequences for competitive position and profitability, but AI agents also have found a foothold in white collar businesses.
Through much of 2025, AI suppliers continued to benefit, AI builders invested hundreds of billions of dollars in AI cap ex, seemingly without much pushback, the product companies drew attention with frontier models that dazzled users with their mix of power, albeit with rough edges, and markets continued to reward them all, based on its perception of market potential.
3. Business Building
A big end market notwithstanding, business building is hard and in this phase it is inevitable that questions will turn to unit economics, economies of scale and eventual profitability, with each group facing different degrees of questioning.
By this stage, AI suppliers should have fully formed business models, but revenue growth rates will ease, as new infrastructure spend becomes smaller, at least on relative terms. As builders try to develop cheaper supply sources, either by bringing some of it in-house or by looking at competitors, margins will start to compress, and you will see some suppliers focus on the infrastructure replenishment/replacement market, rather than on new infrastructure. Since markets are forward-looking, the pricing of these companies will reflect expectations of lower growth and the companies will look cheap on a pricing basis (low PE ratios, low EV multiples), relative to their history.
As more infrastructure becomes operational, AI builders will begin to see more revenues and earnings from existing investments, but the magnitude will depend largely on whether demand from AI product companies is running ahead of capacity (good for revenues and earnings) or behind (bad for revenues and earnings). As returns on invested capital pick up, markets will make their own assessments of whether they will end up adding or destroying value, and reassess the pricing of these companies accordingly.
AI product companies will jockey for position in the market, looking for ways to differentiate their products and preserve pricing power. With compute power being a key driver of both AI model power and growth in revenues, AI companies will try to not only get more AI infrastructure under control, but, if it is scarce, lock it for the long term. In the aftermath, AI product companies with a duration problem, where AI compute commitments are locked up long term, while their revenue streams are still in their formative phase and subject to short term volatility. Again, the forward-looking nature of markets implies that the pricing of these companies will look high, relative to current operating numbers, but that pricing will be volatile, as markets reassess prospects.
AI users will find themselves on the receiving end of sales pitches for AI products and services, with the key selling point being improved productivity, with the warning included that not buying these products will put them at a disadvantage, relative to competitors. For AI to find its business feet, you need the customer base for AI products and services to become both wider (cutting across many different industries) and deeper (where customers spend far more on AI, on a repeat basis).
There are three broad scenarios that can unfold. The first is that AI products and services work much better than expected, and that the demand for these products runs ahead of expectations. In this scenario, AI product companies will benefit first, since they will see growing revenues and higher pricing power, which create the unit economics and economies of scale for escalating profits, but the benefits will ripple to the AI builders, allowing them to increase capacity, which in turn will lift AI supplers. The second is that AI products and services work about as well as expected, with demand matching up to expectations. In this case, AI product companies should, at least in the aggregate, hold their own, but as with any competitive business, some will gain at the expense of others. There will be both business consequences, as weaker players drop out, and market implications, as the market adjusts the pricing for the survivors. The third is that AI products and services work less well than expected, with demand and monetization falling well below expectations. This will hurt not just the AI product companies who will struggle to meet revenue targets, but ripple through the other groups, as product companies either break or renegotiate compute commitments, hurting builders, who in turn, will shut down or abandon infrastructure capacity, creating pain for AI suppliers.
Commonalities and Competing Interests
Even though the businesses in the AI space have very different business models, there is one common element that holds them together; the value of all of the players in the ecosystem ultimately will be determined by the size of the AI market and its profitability. As I noted in my earlier posts on AI, if the end market for AI products and services is small, because of the imperfections as AI tools, regulatory restrictions or limited reach, the entire AI ecosystem (builders, suppliers and product businesses will all feel the pain. Conversely, if AI delivers on its promise as a powerful productivity tool that works across many businesses, all of them benefit, albeit in different forms and different degrees. If history is a guide, it is the product companies that will be the primary benefactors, but the central role of compute power in the AI business may give AI builders and suppliers more of a share. In addition, if one of the groups faces growth and survival challenges, those pains will spread into the other groups as well. That may explain why you see so much circular financing in the AI ecosystem, where financially better- positioned and stronger players like Nvidia invest in or offer financing guarantees to weaker players, either to ensure survival or to provide an impetus for continued investment spending.
Those common interests notwithstanding, these businesses also have competing agendas, and decisions and actions that affect one group negatively can create benefit for others. Thus, if AI builders choose to build too much infrastructure, too quickly, it will hurt them and their shareholders, but that overbuilding will be a bonanza for the AI suppliers, as chips, equipment, power and real estate players all benefit from the overbuilding. Once built, it is in the AI builders' interests to lease out the infrastructure for longer periods than merited, locking in their revenue streams, but for the AI product companies, those long term commitments can increase default risk. In short, as the AI business matures, I would expect these competing interests will often overwhelm the common interests, creating divergence.
For much of AI's early existence, the commonalities have overwhelmed the competing interests, with the hope of a large market for AI products and services pushing up market pricing for all of the groups, with AI chip companies like Nvidia being rewarded for the expected growth in AI architecture, hyperscalers seeing their pricing marked up, when they spend more on architecture, and the compute power commitments atAI product companies being viewed as a competitive plus, rather than minus. That, in a nutshell, has been the AI trade, but as the AI disruption moves through the disruption, it is inevitable that this will change, with the market differentiating across groups, picking winners and losers.
AI Companies - Letting the Data Speak!
With that long lead-in, I think that we are better positioned to take a look at the AI trade, both in terms of where it is right now and where it may be going next. I do believe that the easy AI trade, where you could buy any company that was part of the AI ecosystem (supplier, buyer or product) and make money from a general optimism about AI’s potential is mostly done, and there are many traders who can point to profits they made from that trade. While last week’s market collapse in AI stocks may have drawn attention, the AI trade has been far shakier in 2026, as strength is some AI groups has masked weakness in others, reflecting a market trying to discern differences across companies in the space. To get an empirical measure of how the different AI groups have performed this year, I create samples of companies in each of the three AI groups (supplier, builder, user) and before you take issue with my choices, I will concede on four points:
Public vs Private: Data availability and the requirement that there be a market price restricts the sample to publicly traded companies. That not only reduces sample size and creates sampling bias, but it also has disparate effects across the different groups. In particular, two of the biggest players in the AI products grouping, OpenAI and Anthropic, as well as almost all of the smaller players (Mistral, Qwen and Deepseek) do not make it into the sample, because they stayed privately held, and SpaceX enters the public company sample in the middle of 2026.
Missed companies and changing sample: Since there is no separate accounting categorization for AI-related revenues and expenses, I relied on public descriptions of the companies operating in each of three AI groupings. Those descriptions are likely to be incomplete, but the companies that are missed are smaller and lower-profile companies, in terms of both operating metrics and market capitalization. In addition, as companies go public or exit the market, there will be changes in the sample.
Not pure plays: Most of the companies in this sample operate in other businesses, many of which pre-date their entry into AI. This is especially true of the hyperscalers, where Amazon, Alphabet, Meta and Microsoft have legacy businesses that generate large earnings and market value. That will make it difficult to isolate the AI effect on market capitalization, because changes in market cap at these companies will be driven not just by their AI actions, but by legacy business performance.
In multiple groupings: Some of the companies in this sample operate in more than one grouping. Meta, Alphabet and Microsoft, in addition to be major investors in AI cap ex, are working on their own versions of LLM (Muse for Meta, Gemini for Alphabet and Co-Pilot for Microsoft) to enter the product market. I have included SpaceX as an AI product company, but at the moment, almost all of its revenues on its AI arm come from leasing out its compute power to Anthropic and Alphabet, making it more of an AI builder than AI product company.
With those caveats in place, here are the public companies that I have in each grouping:
Not surprisingly, given where AI is in the disruption cycle, with AI architecture is in its build phase, and products still nascent, we have far more publicly listed companies that are involved in supplying the materials needed for the architecture - chips, electrical, power and real estate/construction, than in the other group. To get a measure of how each grouping has performed as the AI boom has unfolded, I looked at the aggregated market capitalization of companies in each group from 2022 to 2025:
As you can see, the market capitalization of each of the groupings surged over the three year period, with AI chips companies having the biggest increase, in both dollar and percentage terms, over the period; Nvidia was the biggest contributor. The AI product segment also surged, but the value of the publicly traded companies in this space is small, as the biggest companies (OpenAI, Anthropic and xAI) were still privately owned during this period. The stock returns of AI builders lagged other AI groups but, but their market cap increased 180% between 2022 and 2025, even as they were spending hundreds of billions in AI cap ex, with little to show as returns on that investment. Cumulatively, AI companies outpaced the rest of the market substantially over this period, rising as a percent of the market.
Have things changed in 2026? To answer that question, I looked at how these AI groups have performed during the year:
AI products group percentage change does not include SPCX
Looking at the year-to-date numbers for 2026, AI chips and AI real estate have run ahead of the market, and the rest of the AI groups lag the market; again, you should read the AI products and services grouping with caution, because it has only three companies in it, and SpaceX made its entry into the group mid-year. In fact, AI companies, in the aggregate, increased their share of overall US equity market capitalization for much of the year, peaking at about 37% on September 30, 2026. Since the end of the third quarter of 2026, though, every AI group has struggled, with the exception of AI real estate & construction. Last week, in particular, was a testing one for AI companies, with Thursday (November 5) standing out and in the figure below, I report the dollar value changes of each group by day:
On November 5 alone, AI stocks lost almost $1.8 trillion in market cap, with chip stocks taking the brunt of the hit. During the course of the the last week, the AI chip and AI electrical groupings took a beating, but most of the other groups held their own. In fact, the hyperscalers, with the four Mag Seven companies in their midst, actually saw an aggregate increase in market cap during the week. It is a small sample, and there is a danger that I am overreaching, but if the immediate aftermath of last Tuesday's elections is a slowing down of data center construction, it stands to reason that the first companies to be affected will be the AI suppliers, who are dependent on growing construction for their growth. In fact, given that shareholders in the hyperscalers have become increasingly concerned about the magnitude of AI cap ex spending, they may very well welcome a slowdown.
An Assessment of AI as investment: The Big Market Delusion!
I would like to think that I have kept an open mind about AI's business potential, in my writings, and I have been taken to task by AI advocates, for not being optimistic enough about the size of the AI market, and by AI skeptics, who believe that I am inflating its size. In almost all of my writing about AI, though, I have argued that it is exposed to what I call the big market delusion, where the potential of a big markets in conjunction with overconfidence on the part of founders, VCs and CEOs, leads to over investment and a correction. The picture below is the generic version:
I am seldom certain about anything in markets, but I believe that AI, because of its reach and potential, almost certainly has led to over investment, by AI builders, over commitments to compute power by LLM companies and over pricing, by the market, at least in the aggregate. The correction, when it comes, will cause pain, but the pain will be distributed unevenly, and for those who believe that AI has potential but have found the companies in the space over priced, it will provide opportunities. Before you decide that this (the over investment, over pricing and subsequent correction) is a problem, it is worth noting that this phenomenon is a feature, not a bug, that is associated with uncertainty and big change, and that it is, in spite of the costs of the cleanup, is a pre-requisite for change.
This big market delusion suggests that you should either not buy any of the companies in the AI space or even sell short on these companies, but here is why you may not arrive at that conclusion:
Uncertainty about correction (when and what form): Even if you are right that investor and business overconfidence has lead to overreach, you need the market to come to agree with you, and it is difficult, if not impossible, to try to predict when in will happen, and what form it will take. The obvious response that many investors have is to just avoid AI companies for the moment, but since it is just a large portion of the market, this becomes a variant on market timing, with questionable payoffs.
Catalysts for correction: The catalysts that trigger the correction can be bad news about a leading Ai company, a macro or political event that makes investors reassess the landscape, and history offers little useful guidance. The big market delusion is ultimately built on a behavioral characteristic (over confidence) and the correction has as much to do with investor psychology, as with financial fundamentals. It is one reason why I am not quick to dismiss charts and technical indicators (price and volume) as investing voodoo, since these tools may be better attuned to detecting these mood and momentum shifts.
Acting in the fact of uncertainty:Since the correction process is difficult to time, if you are an investor of long standing in one of AI’s many winners, you should use each correction phase to sell a portion of your shares. You will give up upside, by doing so, but you will not be completely exposed, when the correction happens. If you decide to buy AI stocks after the correction, don’t wait for the bottom, since you may never hit it. Find the companies in the AI marketplace that you would like to own, set a trigger pricing or a valuation at which you would buy, and automate the process. Thus, if you like SpaceX, but only if its market cap is $1 trillion, put in a limit buy at that value, and let it get filled automatically.
Go where it is darkest: One of the themes of my valuation and investment philosophy classes is that the payoff to assessing fundamentals and valuing companies is greatest in sectors and at times of maximum uncertainty. If you are an AI advocate, and you have generated supersized returns over the last four or five years from just buying companies in the AI space, you are entering a phase, where even if your AI optimism holds, you should prepare yourself for more periods like last Thursday, where the market mood on AI darkens, and use those periods to be more discriminating in your AI investment choices.
For those of you have read this far expecting a clear signal of whether now is the time to sell or buy, I know that you may find my conclusions to diffuse to be actionable.
Implications for Anthropic IPO
Since I have spent much of the last few weeks valuing Anthropic, ahead of its IPO, and following up on that valuation, there is no avoiding the question of what the market and political turmoil in the two weeks will do that IPO. I have seen rumors floating around of a cut in the IPO pricing (from the initial estimate of $1.5 to $2 trillion to a lower number and some that the IPO may be delayed into next year.
As I noted in an earlier post, IPOs are priced, and in young businesses, with uncommon metrics. After not hearing talk of annualized revenue run rates for a few weeks, there was a leak that Anthropic's ARR for October hit $80 billion, which was touted as both good news (by some) and bad news by others. The truth is that pricing is a mood and momentum game, and it is undeniable that the mood has darkened not just for AI, but also for the market in the last two weeks. If you consider bankers who are terrified of over pricing a trillion-dollar plus offering and picking up the pieces, the pressures to lower the offer price only get stronger.
On the valuation front, it is not market mood that matters but events that change the fundamentals. The most critical of these were the election results, where voters sent a clear message about their aversion to data centers. In Texas, which is home to about a third of the new data centers coming online in the near term, it may not stop existing data center construction, but it will slow the growth of data centers, at least in the United States. It is also possible that the newly elected congress will write in tighter restrictions on AI, creating a bigger drag on margins. For those who are more focused on seeing the glass as half full, this will also vindicate Anthropic's rush to lock in compute power commitments, before the rules of the game changed, and the limits on data centers may give Anthropic a moat against competitors. Just to give you a measure of how sensitive the Anthropic valuation is to these changes, increasing how much Anthropic will have to invest to add compute power in future years (by lowering the sales to cap ratio from 3 to 2.5) and increasing the drag on margins from regulation (from 5% to 7.5%) reduces the estimated equity value to $1.25 trillion, even if you leave the target revenues at a trillion dollars.
I have no sources in investment banks, but I think that this IPO is too far along for bankers and the company to withdraw or delay, without markets reading worst case scenarios into their actions. That said, I would not be surprised to see Antrhopic priced closer than $1.25 trillion, with perhaps a smaller cash raise (with fewer shares being offered), allowing for that "price jump" on the trading day and a buzz that carries over.
Conclusion
As AI moves from the hype phase to business building, I hope that we stop talking about "the AI trade", as a catch all for AI companies, and start differentiating between the different groups for two reasons. The first is that as the AI business starts to take form, there will be developments that are good for one group and bad for others; the election results clearly affected AI suppliers more negatively than it did AI builders. The second is that when you have crises, the market will lose its head and mark down all AI companies indiscriminately, offering you the chance to buy those that you believe should not have been marked down. I also think that it is time that we stop making overly broad assessments of the AI business that lock us into our priors. Thus, if you start with the presumption that the entire AI phenomenon is a "bubble" (notice that this is the first time in the post that I have used that word), your mind is already made up and no AI company, no matter what group it belongs to, will be worth its price. That view, no matter how strongly held, would have kept you out of a segment that has added almost $18 trillion in market capitalization between 2023 and 2026, and even with a strong market correction, it is likely that staying out of the AI bubble would outweigh what you would have earned by investing in some of these companies. At the other extreme, if you are so dazzled by the technology that you see this as a sure win, you are setting yourself for disappointment, because not only are there no sure wins in equity markets, but the pathway to AI nirvana will be come with challenges that test your faith. In short, stay with your beliefs about AI but leave the door open for different opinions.