AI Scenarios: From Doomsday Destruction to Do-Nothing Bots!

     When Chat GPT made its debut on November 30, 2022, it unleashed the hype of AI, and in the three years since, AI has taken on an outsized role not just in markets, but also in our lives. For much of the time, the AI story has been told by its advocates and its salespeople, and the companies in the AI ecosystem have benefited. Not surprisingly, given that its narrators benefit from this growth, that story has emphasized the positive, with dazzling AI use cases and optimistic extrapolation of the productivity gains from its adoption. In the last few months, we have seen cracks emerge in the AI story, with investors wondering when, and in what form, the immense investments in AI architecture will pay off, and how if they pay off, the businesses that they disrupt will fare. That disquiet has played out as negative market reactions to new AI investments at Meta and Amazon, a markdown in software company market capitalizations and in a sell off last week, in response, at least partially, to an AI scenario assessment from Citrini Research, a publisher of macro and stock research. Given that I know very little about the technology of AI, and that my macroeconomic knowhow is pedestrian,  my intent in this post is less about promoting my favored AI scenario, and more about providing a framework for you to develop your own.

The Citrini AI Assessment - Report and Responses

    The Citrini AI assessment came out on February 22, 2026, and it starts with a preface stating that it is presenting a scenario, not a prediction. I do have issues with that opening, but I will come to them later, but the report itself laid out a story for AI that unfolds with a dark end game for the economy, where by June 30, 2028, the AI disruption has unsettled businesses and displaced workers, with unemployment rates rising above 10% and the market down almost 40% in response. There have been other AI doomsayers, but many of those doomsday scenarios are built around the storyline that AI will not live up to its promise, and the pain comes from having over invested trillions of dollars in building its architecture. In contrast, the Citrini AI  story is built on the expectation that not only does AI work well at doing tasks currently performed by white collar professionals, across a range of firms, but its adoption happens very quickly. The pain in the Citrini story comes from that disruption creating substantial job losses, and especially so among higher-earning workers, and the resulting loss of income driving these job losers to cut back on consumption. The ripple effects play out across businesses, with default risks and spreads rising, private credit collapsing and the market and economy pricing in the pain.

    I do think that there are major flaws in the steps leading to the economic implosion in the Citrini assessment, but credit should be given where it is due. I have always been troubled by how much we have worshiped at the altar of disruption in this century, putting the founders of disruptors on pedestals and preaching disruption's virtue. In keeping with Joseph Schumpeter's description of capitalism as built around creative destruction, I do believe that a vibrant and dynamic economy needs a shake-up and challenging of the status quo, but disruption comes with costs to the businesses that are disrupted, and to the people who work in them. There is much to celebrate, as consumers, in terms of choice and price from the growth of online retail, but that does not take away from the devastation that has been wreaked on brick-and-mortar retail and its constituent parts. Ride sharing has brought car service from its nineteenth century ways into the twenty first century, but at the expense of yellow cabs and conventional car service businesses. The reason that many AI advocates took issue with the Citrini report was precisely because it bought into their sales pitch of how AI bots can not only do what lawyers, bankers, software engineers and consultants do, but also do them better, and then asked the question of "what then?.

   The Citrini AI scenario must have hit some targets, because in the days since, we have been flooded with scenarios countering Citrini and arriving at different outcomes. While I was not surprised to see Goldman Sachs, Moody's and JP Morgan jump in with their AI scenarios, with more benign outcomes for the economy, where the job loss and income effects from AI are modest and temporary, I was surprised to see Citadel wade into the argument, with a direct rebuttal to Citrini, which sees a much more positive end game from AI disruption, and is built around three pillars. The first is the current data on jobs and layoffs in the businesses most directly targeted by AI, such as software, where they note that while jobs have been shed, the job losses have been modest, and AI adoption trends don’t see breakouts consistent with the speedy disruption predicted by Citrini. The second is history, where they look at disruptions in the past (PCs, the internet) and note that none of them have been speedy or have created the job losses or economic collapses predicted in the doomsday scenario. The third is grounded in macroeconomics, where they point to the inconsistency of assuming  that a large positive productive shock, from AI’s success, will play out out as large negative shock to the economy and market in which it happens. 

Completing the AI story

    The problem with all of these AI scenarios is that they are rooted in the weakest of responses to uncertainty, which is to either pick a scenario and to describe it in detail, without establishing, at least in qualitative terms, how likely that scenario is, in the first place, or to list out a whole host of scenarios, without making judgments on likelihood on eany of them. It is entirely possible that what Citrini was presenting was a "worst-case" scenario (I read through the report and could not get a sense of if this was so, and the subsequent responses from Citrini have only muddied the waters), a "low likelihood" scenario or the "likely scenario" of how AI will unfold. If it is a likely scenario, and you buy into the pitch, the investment and personal consequences will be dramatic, since it is entirely possible that, if you are a white-collar worker, you may have lost your job by June 2028, and your savings, if invested in stocks, would have taken a beating. If it is a "low likelihood" scenario, and you are exposed, because of your job, age and portfolio composition, you should consider buying protection, but if it is a worst-case scenario, it is almost entirely useless, except for shock value.

Point Estimates and Probabilities

    For much of its history, financial analysis has been built around point estimates, where you identify key drivers, estimate the effects on your bottom line (earnings, cash flows) and make your best judgments. Thus, when valuing a company, you estimate the earnings growth on base year earning, how much you will reinvest of those earnings to grow to get to cash flows, and discount those cash flows back at a risk-adjusted rate to get to value. The problem with point estimates, where almost everything is uncertain is that you will be wrong 100% of the time, though you may still make money, if you are wrong in the right direction.

    Financial analysts and economics have been slow in adopting and using probabilistic approaches, where point estimates are replaced by distributions, and a single judgment on outcome by a distribution of outcomes. One reason, at least early on, was that economists and financial analysts often did not have rich enough data or powerful enough tools to use decision trees, simulations or scenario analysis in making their macroeconomic and investment judgments, but that is no longer true. Another reason may be that many in this group are uncomfortable with statistical distributions or probability estimates and stay away from using them, because of that discomfort. The third reason, at least for a subset of analysts, is a concern that being open about estimates and the errors in those estimates, which is visible to all in probabilistic approaches, will be viewed as a sign of weakness or lack of conviction on their part. I have a short paper on using probabilistic approaches, where I look not only at when you may want to use which approach (I look at decision trees, simulations and scenario analysis) but also have a short review of statistical distributions, if you are interested. 

    Since Citrini specifically titled their AI thought piece as a scenario, I will stick with scenario analysis in this post. In its most sloppy form, and one that has been around for decades, scenario analysis has taken the form of best case - base case - worst case scenarios, an almost useless exercise, since there are almost no risky investments that are going to pass muster under the worst case scenario, no matter how good they are, or are going to fail under the best case scenario, no matter how bad they are. A scenario analysis, done right, should look at scenarios that cover all possible outcomes on an investment or decision, and for completion, need probabilities attached to these scenarios, which can then be used by a decision maker to estimate expected values. That will be almost impossible to do if you are trying to work out future pathways to AI, since it is so early in the process and so little is known about outcomes. 

    There is an alternate path for scenario analysis that is less information-intensive and thus more feasible, and it draws on the 3P test that I  use when valuing companies, where my company valuation narrative has to start with the possible test (it can happen) to being plausible (which requires more backing) and then on to the probable (where you can estimate a likelihood). In the context of scenario analysis, this would require that you categorize scenarios into their the three groupings:

The discussion around where AI is going would become much healthier if scenario proponents were required to state where their proposed scenarios fall in this spectrum. Citrini, for instance, could have saved itself from some of the backlash, if the writer of the AI doomsday report had specified that it was a possible, but not quite plausible scenario.

The AI Disruption - Gaming the Outcomes

   In the last week, I have seen at least a dozen scenarios touted by individuals and entities, many of whom I respect, and I must confess that I am whipsawed. If, like me, you are drowning in these scenarios, with very different results and outcomes, the only way to retain your sanity and to take ownership of this process is for you to develop a framework where you can not only put each of these scenarios to the 3P test, but also to develop your own assessment of how AI will play out for businesses, investors and the economy.

1. The Disruption - Form and Speed

    The first set of questions that you need to address in the AI story relate to how the AI disruption will evolve, both in form and timing, and to then trace out the aftereffects. 

1. AI Disruption Magnitude - Worker Displacement versus Productivity-enhancing Tools:  If you listen to some of AI’s lead players, AI will have the capacity to replace workers across multiple businesses, as it develops strengths that go beyond the purely mechanical. One reason that the AI effect on unemployment is so large in the Citrini doomsday scenario is because AI’s reach in the scenario is not just restricted to replacing programmers in software but extends to replacing white collar workers in other technology businesses, financial intermediaries, banking and consulting. In contrast, Citadel’s more benign AI reading comes from AI displacing workers in a smaller subset of businesses, while providing tools in others. At the other end of the spectrum, there are still some who believe that when all is said and done, AI will provide tools to workers that may save them time, but will not be powerful or dependable enough to replace them.
2. AI Disruption Speed: Here again, there is disagreement, with some AI optimists believing that its disruption of regular businesses is imminent, whether displacing workers or in giving them tools. Others believe that AI adoption will take time, partly because the tools need work and partly because businesses and workers are slow to adapt to change. The Federal Reserve in St. Louis has created a tracker of AI adoption rates across users, and while it does not capture the depth of the AI adoption, it does provide a measure of how much familiarity and comfort that users are acquiring, with AI tools. 

With the caveats about survey data in place, there are interesting trends in these surveys. First, the use of Gen AI tools in non-work settings has grown more than its usage at work, an indication perhaps of how personal devices (phones, in particular) have changed technology adoption rates. Second, the time that AI has saved people, at least so far, has been modest, ranging from less than 1% in the accommodation and food businesses to about 4% in information and management of companies. Overall, this graph suggests that AI usage is neither as explosively fast growing nor as much of a time-saver, as its proponents suggest that it is. The pushback, though, is that these are surveys of the general population, and that there are data points indicating that the disruption effects are more substantial including the substantial write down in market capitalizations of software companies and layoffs at tech companies. The announcement by Block, the fintech company founded by Jack Dorsey, that it would it be letting go of almost 40% of its workforce, for instance, and blaming AI's rise for the action, was viewed as an indicator of AI's disruption potential. That is a noisy signal, though, since many tech companies have bloated work forces, and AI gives them easy cover, when correcting past mistakes. 

It is true that there is no crystal ball that you can use to gauge the magnitude and speed of AI disruption, but every AI scenario that you see starts with a judgment on one or both. 

2. The Disruption Aftershocks

    Disruptions create aftershocks, some positive and some negative, and while we often avert our gaze and attention from the latter, a full assessment requires considering both. With AI, the positive effects take the form of higher productivity, as it either allows people to do their jobs more efficiently (with AI tools) or actually replaces people and does their jobs instead, in effect allowing for more output with less labor. Relating back to the different pathways that AI disruption can take, both in form and in form and speed, I would hypothesize that these disruption benefits will be a function of how AI disruption plays out.

Proposition 1: The disruption benefits from AI disruption will be greater from people displacement than from AI productivity tools

Proposition 2: The productivity effects from AI disruption will decrease, at least in economic value terms, the longer it takes for the AI disruption to unfold.

The negative effects of AI, in economic terms, will come from the immediate displacement of people, if AI replaces labor, or from the decrease in employees needed to get tasks done, if AI tools make existing employees more efficient. Here again, I would hypothesize that these disruption costs will be  function of how the disruption plays out.

Proposition 3: The disruption costs from AI disruption will be greater from people displacement than from tools, as those laid off lose income and spending power. 

Proposition 4: The productivity costs from AI disruption will decrease, at least in economic value terms, the longer it takes for the AI disruption to unfold, since time will allow new entrants into labor markets to adjust to a disrupted business world.

Intuitively, the longer it takes AI to find roots in business, the more time it gives workers time to adjust, retrain or move on. As you can see, the scenarios where AI displaces existing employees and happens quickly are the ones with the biggest benefits and the biggest costs, and the scenarios where AI supplies tools to existing employees and happens slowly has the least benefits and costs. Building on this theme, I see the net effect of AI disruption playing out as follows:

If AI disruption displaces existing workforces, across many businesses, and happens quickly, the net effect is likely to be negative, at least in the near term, since the economy will not only have to absorb major layoffs quickly, but also because those laid off will be higher-earning white collar workers. While that maps on to the Citrini doomsday scenario, there is still much to debate about which industries will see the most job displacement and how quickly these workers will find other jobs. There is also a discussion that should follow, even in this negative net-benefit scenario, of how quickly the economy (and workers) will adapt, and if and whether net benefits will turn positive in the long term. If AI job displacement is on a limited scale, and/or takes time to unfold, both the benefits and the costs of the AI disruption become smaller, but the net benefit is more likely to be positive, in the short and long term. Finally, the AI disruption takes the form of tools that make workers more efficient, but not efficient enough to reduce workforces, both the benefits and costs of AI become much smaller. In fact, if these tools take a long time to craft and displace little or no labor you get the AI disruption fizzle, with very small benefits and costs.

3. The 3P Test

    Staying true to my earlier assertion that scenarios without probability estimates are not useful, I will try to put the various AI scenarios that I mapped out in the last section on the  3P continuum.

Let me start with the two possible, but not quite plausible scenarios. The first is the a speedy, massive AI disruption, where AI displaces worker across most businesses, and does so quickly, as visualized by Citrini. It can happen, but given the history of disruption, the limits of AI technology and inertia in the process, it is implausible. At the other extreme, it is possible that AI provides tools to workers that improve productivity marginally, with many ending up being more distractions than tools for productivity, effectively emptying its destructive potential, but that too strikes me as implausible, given what we are seeing in terms of AI capabilities. The most plausible scenarios are ones where AI displaces workers in some industries, such as software and some financial intermediaries, and provides tools that help workers to varying degrees in other businesses. As for probable, I think that disruption will reduce workforces in a subset of businesses, that its tools will include some game changers and that it will take longer to unfold, at least when it comes to monetization, than its advocates think. 

    My justification for why AI disruption will take time is based on a mix of factors. The first is that my (limited) knowledge and experience with AI products is that while they sometimes work magically well and quickly, they do have kinks, coming partly from being unable to separate good data from bad, and partly from their imperfect attempt to be imitate humans. The second is history, where no disruption has ever unfolded without delays and drawbacks; remember that the dot com disruption almost lost its moorings during the market bust in 2001. The third is human nature, where much as employees and managers claim to want to move on to new and better options, they remain attached to old technology and products; typewriters and mimeographs took a while to disappear after PCs stormed the workplace and flip phones persisted well into the smartphone era. 

    There are two reasons why I do think that AI disruption is still going to be significant, in the long term. The first is that some of those making the argument that AI will not displace jobs in the long term are assuming that AI in it more advanced form will look like ChatGPT on steroids or be primarily mechanical in its applications. Even my limited exposure to AI's advanced tools suggests that they have far greater capabilities, and their capacity to mimic human intuition and thought processes is unsettling. The second is the blanket assumption that workers in most white collar jobs will not be easily replaced because they bring training, brainpower and experience into those jobs that will be difficult to replicate. Many white collar workers are bright people with specialized knowledge, but the businesses that hire them put them in straight jackets, pushing mechanics over intuition and rule-driven thinking over principle-driven assessments. In short, it is the nature of the jobs that we have created in many white collar settings  that makes them vulnerable to disruption, not the intelligence or training of the people holding those jobs.

    It is worth noting that in my probable scenario, AI will unfold at different rates in different businesses, and if I were pushed to distinguish between the businesses that will be targeted most (and soonest) from the businesses where it will take more time, and have less impact, I would look at four factors:

There may be some confirmation and hindsight bias in this table, but there is a good reason why software, a relatively young industry, with young companies and employees, has been one of the first targets for AI disruption. it is profitable, its products and services are logical and rule-based and much of it has no regulatory or system protection. Within software, though, i would expect software that requires more user interface to be more resilient to AI disruption than software that operates in the background. This table, though, can help determine which white collar jobs will be most exposed to AI disruption, and which least, and perhaps also explain why blanket statements about job displacement in banking, consulting and law are overwrought. With banking and law, a substantial portion of the work done is to meet legal or regulatory requirements, not fill operating needs. I have written about the inanity and uselessness of fairness opinions in M&A, where bankers opine on whether an acquiring company is paying a "fair" value for a target, but this practice persists because these fairness opinions provide cover against lawsuits that ensue when deals fall apart. My guess is that the Delaware courts are not quite ready for an AI fairness opinion bot to take the stand and defend a deal, even if the quality of its work is better than a human banker. With consulting, where cookbook solutions are more the norm than the exception, it is worth remembering the clients pay consulting fees not for the advice, but so that they have someone else to blame, when things go wrong, and there too, an AI bot will not have the same outsourcing power as an army of bankers with Harvard MBAs from McKinsey.

4. Cui Bono?

    Most of the AI scenarios yield net benefits, and even in the most damaging scenarios, where the AI disruption benefits are overwhelmed by its costs, at least in the short term, you could argue for net positive benefits in the long term. That is good news, but it should taken with a grain of salt, since the distribution of these net benefits across businesses and society will be unequal, and it is possible that the net benefits accrue to a few businesses (and individuals), leaving the rest (businesses and individuals) with net costs.

• The interests of the AI companies and the rest the economy/market will diverge on AI disruption, with the former benefiting if the disruption is across many businesses and happens quickly, and the latter benefiting from a slower disruption restricted to a few businesses. This will be the case even if AI tools add to productivity, since the lower costs that companies acquiring these tools will have as a consequence, may not translate into higher profits, especially if their competitors can pay and acquire the same tools.
• The last few major disruptions, starting with the internet, moving on the China and then the smartphone, have all tilted the playing field in many businesses towards larger companies, making businesses more winner-take-all. It is likely that the AI disruption will play out in similar ways, with the winners winning big, and lots of companies losing out. 
• At the individual level, it is not just plausible, but also likely, that a strong AI disruption will make wealth and income inequality worse, with founders of AI businesses joining the ranks of the  deca-billionaires and centi-billionaires.

There is one final cost that may not be explicit in economic terms, at least immediately, but one that has to enter the discussions, As AI threatens to displace workers in white collar businesses, it is worth remembering that a job is not just an income-generator, but also a source of self esteem and worth. When software engineers, who pride themselves on their coding skills, bankers, who have spent decades becoming excel ninjas, and consultants, who have found inventive ways of packaging cookbook solutions and presenting them as new and inventive, find that AI can do what they have spent a lifetime perfecting almost effortlessly, the psychic damage will be significant. The fact that blue collar workers lost their jobs to the internet and China disruptions faced a similar predicament and were largely ignored also means that there may be more than a hint of schadenfreude in society's response to white collar job losses.

The AI Personal Threat

     If you are looking at these side costs and threat to jobs that will come from the AI disruption, and wondering whether we should opt out, by regulating or restricting its reach, I am afraid that the choice is out of our hands. The genie is out of the bottle, and the only pathway that you have, if you operate in a space where AI is ubiquitous, is prepare for a reality where AI tools can automate and do much of what you do on a daily basis, but where you have to create a niche or moat that still makes you necessary.

    Just about two years ago, I wrote about an AI entity called the Damodaran Bot, that was being developed by Vasant Dhar, my colleague at NYU, and noted that having made all that material that I have developed in my lifetime (classes, books, writing, models, videos) publicly available, I was completely exposed to AI disruption. I have watched that bot develop, with quirks and occasional hiccups, to a point where it can replicate much of what I do almost effortlessly. At the time, though, I did write about what I could do to keep the moat at bay, including the following:

• Generalist vs Specialists: I am a dabbler, an expert in nothing and interested in lots of different things, and I do think that gives me an advantage over a bot that is trained to focus on a topic and drill down. The specialist advantages stem from mastering the vast content in a discipline, but those advantages are diluted with AI entities that can also see that content, but the generalist advantage of using multi-disciplinary thinking with be more difficult for AI to replicate.
• Left and Right Brain: I value companies, and early in my valuation life, I decided that financial modeling was not the right path to value businesses, and that good valuations bridge stories and numbers. If the legend of the right and left brains holds, where the left brain controls logic and numbers and the right brain drives your imagination, a bot will have a tougher time replicating what you do, if you use both sides. That said, I have seen the Damodaran Bot get much better at story telling in the two years that I have watched it, and I need to up my game.
• Reasoning muscle: When faced with questions in the days before the internet, you often had no choice but to reason your way to answers. That may have been time consuming, and your answers might even have been wrong, but each time you did this, you strengthened your reasoning muscles. As we move into a period, where the answer to every question is  online, on Google Search and ChatGPT, we are losing the need to exercise those reasoning muscles, and exposing ourselves to being outsourced by our bots. 
• An idle mind: I am not a voracious reader nor a listener to podcasts, and since I don't have much real work to occupy me, I also have plenty of vacant time, with nothing to do. I use that time to daydream and ponder about questions that capture my imagination, including why someone would pay billions of dollars for a sports franchise (like the Washington Commanders), how to deal with the risk of lava from a volcano hitting a spa and ruining its valuation and how streaming has broken the entertainment business. None of these posts include deep insights, but my guess is that the Damodaran bot would have trouble keeping up with my wandering mind.

With the admission that is may not be enough, and that my bot may soon be able write my books and posts, teach my classes and analyze/present data better than I can, I think that you should all be acting as if a bot with your name is looking over your shoulder and trying to learn what you do, and think about what you can do to keep that bot at bay. 

    There is always the possibility that you are arming yourself for a disruption that fizzles, but I will draw on Pascal's wager to explain why you should prepare for an AI imitator or bot, even if you don't believe that it is imminent:

Pascal, a French mathematician, used the wager to explain why be believed in God, even if he was  doubtful of a heavenly presence, because the expected value from believing in God exceeded the expected cost from not believing. In the context of AI, acting as if an AI presence and competitor is present will make you better at whatever you do, as a teacher, banker, consultant or software engineer, and that will persist, no matter what AI's impact is ultimately. Good luck!

YouTube Video

Data Links

1. Federal Reserve in St. Louis AI Adoption Tracker

AI Scenarios

1. The Citrini AI Doomsday 
2. The Citadel Response to Citrini
3. Dario Amodei on the Upside (and Dangers) of AI
4. Daron Acemoglu on the Economics (and Adoption) of AI
5. It's all going to be good - the Jeremy Siegel view on AI
6. The State of AI: Deloitte, McKinsey

Data Update 8 for 2026: Time for Harvesting - Dividends and Buybacks

    In the data update posts this year, I have wended my way from the macro (equities collectives, the bond market and other asset classes) to the micro, starting with hurdle rates and returns in posts five and six and the debt/equity choice in my seventh post. In this post, I will look at the decision by businesses on how much cash to return to their owners, and in what form (dividends or buybacks), and how that decision played out globally in 2025. I will argue that dividend policy, more than any other aspect of corporate finance, is dysfunctional both for the firms that choose to return the cash and the investors who receive that cash. It is also telling that there are many who seem to view the very act of returning cash as a sign of failure on the part of firms that do so, even though it is the end game for every successful business.

The Dividend Decision

    I start my corporate finance classes with a description of three core decisions that every firm has to make in the course of business, starting with the investment decision, where you try to invest in projects and investments that earn more than your hurdle rate, moving on to the financing decision, where you decide on the mix of debt and equity to use in funding those investments, and ending with the dividend decision, where firms decide how much cash to return to their owners. In the case of privately owned businesses, this cash can be withdrawn by the owners from the business, but in publicly listed companies, it takes the form of dividends or buybacks. In keeping with the notion that these are the cashflows to equity investors, and that those cash flows should represent what is left after (residual) after all other needs have been met, dividends should reflect that status and, at least in principle, be set after investing and financing decisions have been made:

That utopian view of residual cash being returned to shareholders is put to the test by two real-world realities that often govern corporate dividend policy:

1. Inertia: In many companies, dividend policy is set on auto pilot, with dividends this year set equal to dividends in the last year. It is for that reason that the word I would use to describe dividend policy, at least when it comes to conventional dividends, is 'sticky', and you can can see that stickiness at play at US companies, if you track the percentage of companies that increase dividends, decrease dividends or leave them unchanged every year.
   
   
   
   In every single year, from 1988 to 2025, the percentage of companies that pay the same dividends that they did in the previous year outnumbers companies that change dividends, and when dividends are changed, they are more likely to be increased than decreased.
2. Me-tooism: In most companies, managers look to peer group dividend policy for guidance on how much, if any, to pay in dividends. Thus, if you are a bank or a utility, it is likely that you will pay high dividends, because everyone else in the sector does so, whereas technology companies will pay no or low dividends, because that is industry practice. While there are good reasons why some industry groups pay more dividends than others, including more predictable earnings and lower growth (and investment needs), hewing to the peer group implies that there will be outliers in each group (fast-growing banks or a mature technology companies) that will be trapped into dividend policies that don't suit them.

When maintaining or increasing dividends become the end game for a business, you unleash dividend monsters, where investing and financing decisions are skewed to meet dividend needs. Thus, a firm may turn away good investments or borrow much more than it should because it feels the need to sustain dividends. 

I have long argued that dividends, in their sticky form, are unsuitable as cash returns to shareholders, but for much of the last century, they remained the primary or often only way to return cash to shareholders. While buying back stock has always been an option available to US companies, its use as a systematic way of returning cash picked up in the 1980s, and in the years since, stock buybacks have become the dominant approach to returning cash for US companies:

As you can see, in the last decade, more than 60% of cash returned to shareholders took the form of buybacks. The primary reason, in my view, is that buybacks, unlike dividends, are flexible, with companies often reversing buybacks, if macro circumstances change, as was the case in 2008 and 2020. There are other reasons that have been offered for the explosive growth in buybacks, but none of them are as significant. There are some who have argued it is stock-based compensation for managers that is pushing them away from dividends to stock buybacks, but that rationale makes more sense for stock options, where stock prices mater, than for restricted stock. In fact, even as more companies shift to restricted stock as their stock compensation mechanism, buybacks have continued to climb, and they are just as high at companies that have no or very low stock based compensation as at companies with high stock-based compensation. Investor taxes are alway in the mix, since investors are often taxed at different rates on dividends and capital gains, but changes in tax law in the last two decades have reduced, if not eliminated, the tax disadvantages associated with dividends, cutting against this argument. 

    I know that there are many investors, especially in the value investing camp, and quite a few economists, who believe that the shift away from dividends to buybacks is unhealthy, albeit for different reasons. I will return to many of the myths that revolve around buybacks later in this post.

A Rational Cash Return Policy

    If you were designing a sensible cash return policy, it has to start with an assessment of how much cash there is available for a firm to return. Since that "potential dividend" should be the cash left over after taxes are paid, reinvestment has been made and debt repaid, it can be computed fairly simply from the statement of cash flows, as free cashflow to equity:

Note that free cash flow to equity starts with equity earnings, converts those earnings to cash flows by adding back depreciation and other non-cash charges, and then netting out capital expenditures and changes in working capital, with increases (decreases) in working capital reducing (increasing) cash flows. It is completed by incorporating the cash flows from debt, with debt issuances representing cash inflows to equity investors and debt repayments becoming cash outflows. Can free cash flows to equity be negative? Absolutely, and it can happen either because you are a money-losing company, too deep in the hole to dig yourself out, or even a money-making companies, with large reinvestment needs? Obviously, paying out dividends or buying back stock when your free cash flows to equity is violating the simple rule that if you are in a hole, you need to stop digging.

    If your free cash flow to equity is positive, you can choose to return it to shareholders, either in the form of dividends or buybacks, but you are not obligated to do so. In fact, if you have positive free cashflows to equity and you choose to return none or only a portion of that cash flow, the difference accumulates into a cash balance. If you choose to return more than your free cashflow to equity, you will either have to deplete an existing cash balance, or if you run out of cash, go out and raise fresh capital.

A company that systematically holds back on cash that it could have returned will, over time, accumulate a large cash balance, but that, by itself, may not trigger a shareholder response, if shareholders trust the company's managers with their cash. After all, cash invested in liquid and riskless investments, like treasury bills and commercial paper, is a neutral (zero NPV) investment, and leaves shareholders unaffected. If you don't trust management to be disciplined, though, you may punish a company for holding too much cash, effectively apply a "lack-of-trust" discount to the cash. The picture below provides a framework for thinking through the cash return decision, and how it will play out in markets.

As you look at the interplay between earnings, investment needs and potential dividends, you can already see why you should expect cash return policies to change over a company's life cycle:

The cash returns you see in this graph should largely map on to common sense, with start-ups and very young companies, often money-losing and requiring substantial reinvestment to grow, having negative free cash flow to equity (thus requiring equity infusions). Young growth companies  are usually self-funding because internal cash flows may rise to cover reinvestment, but these cash flows are not enough to pay dividends. Mature growth companies have enough cash to return, but stick with buybacks, because they value flexibility. Mature stable companies represent the sweet spot for dividend paying, since they have little in reinvestment needs and large predictable earnings and cash flows. As with everything else in the aging process, companies that refuse to act their age, i.e., young companies that choose to pay dividends or buy back stock or mature companies that insist on holding on to cash, damage themselves and their shareholders.

Dividends in 2025

   I will start the assessment of how much companies returned to shareholders in 2025 by looking at conventional dividends paid by companies, using two metrics. The first metric is the dividend payout ratio, where I divide dividends paid by net income, but only if net income is positive; if net income is negative, and dividends get paid, the payout ratio is not meaningful:

As you can see the median payout ratio is about 35% (59%) for US (global) companies, but in both samples, most companies do not pay dividends. There is a sizable subset of companies (12% of US and 14% of global companies) that pay out more than 100% of earnings as dividends, with multiple reasons for that oversized number including a bad earnings year, a desire to increase financial leverage and partial liquidation plans all coming into play.

    The second metric is the dividend yield, computed by dividing dividends paid by market capitalization, or dividends per share by the market price per share. In the graph below, I look at the distribution of dividend yields across companies in the graph below, in 2025:

Again looking at only dividend paying firms in the US and global samples, the median dividend yield was 1.10% for the former and 2.43% for the latter, with major divergences across sub-regions; note that the percent of dividend paying firms  in the United States has dropped below 30% and even globally, less than half of firms pay dividends. The dividend yield ties into the cost of equity discussion that I initiated in my fifth data update, where I described the cost of equity as the rate of return that investors expect to make on their equity investments. In the United States, for instance, that expected return was about 8.50% at the start of 2026, which would indicate that if you are an equity investor, it is price appreciation that you are dependent on, for the bulk of your equity return.

    The dividend yield for equities has declined over time, with the drop off being most noticeable in the United States. The graph below looks at the dividend yield on the S&P 500 from 1960 to 2025, and how that number has become a smaller and smaller portion of the overall expected return on stocks (which I compute with the implied equity return approach):

In 1960, about half of your expected return on stocks came from dividends and that statistic has trended downwards for the last few decades, and in 2025, it represented less than 15% of the total return on stocks. 

    As a final part of this analysis, I looked at dividend yields and payout ratios, broken down by sector, for both US and global companies:

As you can see, the sectors with the highest percentage of firms paying dividends are financials, real estate and utilities, for both US and global companies, and consumer product companies join in that group, for global companies. In terms of payout ratios, the same three sectors dominate, with energy and real estate returning more than 200% of net income as dividends, in 2025, and posting dividend yields in excess of 6%. Technology companies and communication services have the lowest percent of dividend paying companies and the lowest dividend yields and payout ratios.

    The drop in dividend yields over time for the market, the decline in dividend paying firms and the concentration of dividend paying firms in some sectors has put old time value investing to the test. Ben Graham's strategy of principal protection was built around buying large dividend paying firms and holding on for the long term and it has hit a wall. Any investing strategy built around dividends will result in a portfolio composed of mature and declining firms, and even if you accept that reality, those firms are increasingly concentrated in real estate, banking and utilities. 

Buybacks - Myths and Realities

    As buybacks have soared in the United States, misconceptions and myths about buybacks have also surged, with some myths used to back up the argument that buybacks are unhealthy and should therefore be banned and others presented as the basis for buybacks as good, representing cannot-lose strategies to beat  the market. I will start with the myths that are used to argue against buybacks first, before moving on to those that are used to justify it:

1. Myths in favor of the argument that buybacks are bad and should be restricted or stopped

Myth 1.1: Buybacks are a US phenomenon

Reality 1.1: Buybacks are becoming a global phenomenon

    When US firms first started buying back stock in the 1980s, it is true that is was almost entirely or primarily a phenomenon restricted to the US, with large parts of the world restricting or banning the use of buybacks to prevent price manipulation by companies. That is no longer the case, and companies around the world have taken to buybacks, as a flexible alternative to dividends, have adopted the practice. In 2025, I looked at dividends and buybacks from companies around the world:

Companies in the United States are still in the lead in the buyback race, buying back $1.153 trillion in stock in 2025, close to 60% of overall cash returned. Canada, the UK, and Japan are not far behind with more than 35% of cash returned taking the form of buybacks, and the EU and environs, often the slowest to adapt to change, saw almost 29% of cash returned in buybacks. For a variety of reasons, including poor corporate governance and regulatory restrictions, Africa & the Middle East, Eastern Europe and much of south and southeast Asia return relatively little in buybacks.

Myth 1.2: Buybacks are wasteful and reduce corporate investment

Reality 1.2: Buybacks redirect corporate investment from mature companies to growth businesses

    The argument that buybacks are wasteful often come from using a firm as a self-contained economic unit, and noting that money used on buybacks cannot be reinvested back into the firm. That is absolutely true, but the cash that goes into buybacks goes to investors and mostly goes back into the market, as investments in other companies. While there are clearly exceptions, where companies that should be investing back into their businesses use that cash to buyback stock, the companies that are the biggest buyers of their own stock are mature firms with insufficient investment opportunities and the companies that have the cash redirected into them need that cash to fund their growth. You can see this play out, when you look at stock buybacks broken down, by age decile (based upon corporate age) for US and global companies:

As you can see, younger companies are not only less likely to buy back stock, but also return less cash in dividends and buybacks, at least as a percent of market capitalization than older companies. Using the life cycle perspective, this suggests that cash is rotating out of older, more mature businesses into younger businesses. I would argue that the difference between geographies where buybacks are rare and geographies where buybacks are common is not in how much corporate investment there is, but in where that investment is directed, with the former investing investing back into declining businesses and the latter funding higher growth and newer businesses.

Myth 1.3: Buybacks are funded with debt are are making companies too highly levered

Reality 1.3: Buybacks are primarily funded with free cash flows to equity and even as buybacks have surged, debt ratios have decreased.

    I am not a great believer in case studies precisely because anecdotal evidence is spun into backing priors and preconception.s There are, of course, firms that have dug themselves into a hole by buying back immense amounts of stock, and funding those buybacks with debt, but the aggregate debt ratios for US non-financial service firms, with debt to capital ratios measured against both book and market, have declined over the last four decades, even as buybacks have surged. 

If your response is that not all companies buy back stock, and that debt ratios has risen at companies that buy back stock, a comparison of debt ratios (debt to EBITDA and debt to capital) for US firms that bought back stock in 2025 versus those that do not dispels that argument:

If firms are borrowing money to fund buybacks, it is clearly not showing up in the statistics, since companies that bought back stock had much lower debt loads than the companies that did not, a simplistic comparison, but one that carries heft.

Myth 1.4: Buybacks are value-destroying because companies tend to buy back their own stock when prices are too high

Reality 1.4: Buybacks, at any price, can neither add nor destroy value. They can just transfer value

    Warren Buffett was late to the buyback party, but when he initiated buybacks at Berkshire Hathaway, he introduced a constraint, which is that he would do buybacks only if he believed that the company's stock price was less than intrinsic value. He, of course, had the credibility to make this assertion, but most companies don't impose this constraint and there is evidence that they often buy back their shares when stock prices are higher than they are lower. That does seem like value destruction, but a cash return can neither add nor destroy value, but it can transfer wealth. In the case of stock buybacks at too high a price, wealth is transferred from those who remain loyal shareholders in the firm to those who sell their shares. While there is hand wringing about this, you have a choice, as a shareholder, in a buyback, to sell or hold on, and if you believe that the buyback is at too high a price, you should sell your shares back.

2. Myths in favor of the argument that buybacks are good and generate excess returns for investors 

Myth 2.1: Buybacks are value-adding because companies that buy back their own stock when prices are lower than fair value are taking positive net present value investments.

Reality 2.1: Buybacks, at any price, can neither add nor destroy value. They can just transfer value.

    This is the inverse of the argument that buybacks are value destroying and they are both grounded in a misclassification of buybacks as projects, rather than cash return, competing with investment projects for the company's dollars. The truth again is that a stock that buys back stock at lower than fair value is transferring wealth from those who sell back to those who remain, and here again, if you are on the wrong side of wealth transfer, it was your choice to sell back that made you the loser. 

Myth 2.2:  Buybacks are almost always good for stock prices, since there are fewer shares outstanding after buybacks, and that should increase the price per share.

Reality 2.2: A buyback can increase, do nothing or decrease value per share, depending on the price at which it is done and its effects on leverage.

    Buybacks reduce share count (the denominator) but the cash that leaves the firm also reduces fir value (the numerator). A fair-value buyback will create offsetting effects, leaving value per share unchanged, though there can be a secondary effect on value, if the buyback, by reducing equity, changes the debt to capital mix and cost of capital for a company:

It is true that empirical evidence backs up the notion that stock prices benefit from buybacks, but that may be from the selection bias of under levered firms with large cash balances being the biggest players in the stock buyback game. 

In general, almost all of these myths come out of treating buybacks as something new and different, rather than a variant on dividends. In general, companies that should not be paying dividends, either because they lack the cash or the future is uncertain, should not be buying back stock either. 

Dividend Dysfunction

    At the start of this post, I noted that dividend policy is dysfunctional at many firms, driven by inertia (we've always paid dividends or we've never paid dividend before) and the desire to hew to peer group policies. As a result, there are many companies around the world that adopt dividend policies that, at least of the face of it, take explaining including:

1. Money-losing companies that pay dividends: While there are some companies that offer justifications grounded in worries about sending bad signals or hopes of a bounce back in earnings, many get stuck with dividend policies, because of inertia or peer group pressure, that can drive them into ruin.
2. Money-making companies that refuse to pay dividends: Here again, there can be good reasons for holding back including concerns about whether you can sustain earning and expectations that you will need to invest more in the future, but in some cases, it can unwillingness to initiate dividends in an industry where no one else pays dividends.
3. Negative FCFE companies that return cash (dividends or buybacks): In addition to hopes for a bounce back in FCFE, companies may continue to return cash, even with negative FCFE, because they are trying to increase debt ratios or shrink their businesses over time. 
4. Positive FCFE companies that return no cash: Companies that have positive FCFE that don't return cash may hold back that cash because of the desire to reduce debt ratios or because they ahve investment plans.

The graph below lists out the number of companies in each group, broken down by geography:

Across the globe in 2025, almost 18% of money-losing companies paid dividends, as did about 70% of money-making companies. With FCFE as your indicator, about 37% of companies that returned cash (in dividends and buybacks) in 2025, had negative FCFE, as did 66% of companies with positive FCFE.

Conclusion

    There are a whole host of misalignments between what companies return to their shareholders, either as dividends or in buybacks, and what they can, as potential dividends. That suggests to me, and perhaps I am wrong, that investment strategies that are built around cash return, whether they be dividends or buybacks, are likely to go off the tracks. Furthermore, any strategy that is built entirely around dividends, as is the case with strategies where you load up on high dividend yield stocks or buy a handful of heavy dividend payers, such as the Dogs of the Dow, misses the essence of equity investing. A stock is not a bond, where dividends replace coupons, and you get some price appreciation on top, and treating it as such will only create disappointment.

YouTube Video

=

Data links

1. Dividend statistics, by industry (US and Global)
2. Buyback statistics, by industry (US and Global)
3. Dividends and Buybacks - History for US firms

Spreadsheets

1. Buyback stock price calculator

Data Update Posts for 2026

1. Data Update 1 for 2026: The Push and Pull of Data
2. Data Update 2 for 2026: Equities get tested and pass again!
3. Data Update 3 for 2026: The Trust Deficit - Bonds, Currencies, Gold and Bitcoin!
4. Data Update 4 for 2026: The Global Perspective
5. Data Update 5 for 2026: Risk and Hurdle Rates
6. Data Update 6 for 2026: In Search of Profitability
7. Data Update 7 for 2026: Debt and Taxes
8. Data Update 8 for 2026: Dividends and Buybacks

Data Update 7 for 2026: Debt and Taxes

   In my fifth data update, I examined hurdle rates in 2025, and in my sixth data update, I looked at the profitability and return metrics for firms. Both hurdle rates and profitability metricsmcan be affected by how much debt companies choose to have in their financing structure, and it enters explicitly into my cost of capital calculations, both through the costs of equity/debt and the mix of the two, and into my accounting return calculations, for net margin and return on equity. In this session, I start with an examination of the trade off that all businesses face when it comes to choosing between debt and equity to fund their operations, and then look the debt choices that companies made in 2025. As with every other one of my data updates this year, AI enters this conversation not only because of the huge investments that are being made into AI architecture, but also because a non-trivial portion of this investment is coming from debt, with private credit as a key contributor.

Debt versus Equity: Choices and Tradeoff

    The discussion of the tradeoffs that businesses face on whether to borrow money (debt) or use owner's funds (equity) has to start with a clear distinction between what it is that sets them apart. While that distinction may seem trivial, since accountants do break financing down into debt and equity on accounting balance sheets, accountants are not always consistent in their categorization, and I think that understanding what sets debt apart from equity can help catch these inconsistencies. There are three dimensions where debt and equity deviate:

1. Nature of claim: Debt gives its holders a contractual claim on the cash flows, insofar as the terms of interest and principal payments are laid down contractually at the time of the borrowing. Note that these contractual claims cover both fixed rate debt, where the interest payments are fixed over the lifetime of the debt, and floating rate debt, where the interest payments will change over time, but in ways that are specified by the bond/loan agreements. Equity gives its holders a residual claim, i.e,, a claim on cash flows, if any, that are left over after other claim holders have been paid.
2. Priority of claim: This follows from the first distinction, but debt holders get first claim on the cashflows, when the firm is in operation, and on liquidation proceeds, if the firm ever goes bankrupt. It is this priority of claims that should generally make debt safer than equity in almost every enterprise that employs both.
3. Legal consequences: A company that fails to pay dividends to its equity investors, no matter how deeply set their expectations of receiving these dividends, may see its stock price drop, but it cannot be held legally accountable for the failure. A company that fails to make its contractual obligations on debt can not only be sued, but can be pushed into bankruptcy, effectively ending its business life.

There are three other distinctions, which do not always hold, but are usually true:

1. Tax Treatment: In much of the world, the tax code is tilted in favor of debt, with interest payments being tax deductible and cash flows to equity (dividends or buybacks) coming out of after-tax cash flows, but there are three caveats. The first is that the tax savings from debt kick in only when a company is generating a taxable profit, though laws on tax loss carry-forwards can allow even money-losing firms to get tax benefits, albeit with a delay. The second is that there are parts of the world, such as the Middle East, where the tax code explicitly bars interest tax deductions, though companies find work arounds sometimes to get the benefits. The third is that there are a few countries that try to even the playing field by either giving a tax deduction to companies for some payments to equity investors (interest on capital as a tax deduction in Brazil) or to investors directly by allowing them credits for corporate taxes paid, when they receive dividends.
2. Role in management: In most businesses, equity investors are given supremacy when it comes to managing the company, exercising that power through either direct ownership or corporate governance mechanisms (such as boards of directors). Again, there are exceptions, as is the case where lenders are given seats on boards of directors or veto power over major operating decisions, but these exceptions are usually triggered when companies violate covenants in loan agreements. 
3. Maturity: Debt usually has a finite maturity, though as we saw with the Google hundred-year bond issuance just a few weeks ago, that maturity may be well beyond the lifetime of the buyers of the bond. Equity, in contrast, is, at least on paper, an instrument with no finite due date, and may have cash flows that last into perpetuity. 

The figure below captures the differences between debt and equity in the context of a financial balance sheet:

With these distinctions in place, and given that businesses have a choice of using either debt or equity to fund their operations, let us look at the trade off, starting with what the fictional (but often used) reasons for using one source of funding over the other: 

One of the most common (bad) reasons that I hear business owners and CFOs of even large companies give for borrowing money is that debt is cheaper than equity. On the face of it, that is of course true, but it is an illusion, at least without the tax benefits kicking in. If the debt is fairly priced, i.e., you are being charged an interest rate that reflects your default risk, borrowing money will make your equity more risky and leave your cost of capital unchanged (if you have no default risk) or raise it (if you have default risk). Intuitively, your cost of capital is designed to capture the risk in your operations, and playing games on the financing side cannot change your operational risk. Among risk-takers, a common reason for using debt is that it will increase your return on equity, and while that again is technically true, it will also raise your cost of equity and magnify the impact of both your successes and your failures. Thus, if you want to borrow money to magnify the payoff to you, as an equity investor, from a successful trade or investment, you should do so, but dispense with the illusion that this is a free lunch.  Those who avoid debt have their own share of illusions, starting with the argument that the interest payments on borrowed money will lower net income. That is true, but since you have less equity invested, you may still come out as a beneficiary. They also argue that debt will increase default risk, and lower their bond ratings, but of which are likely to happen, but the objective in business is not to maximize bond ratings, but to increase value; a BBB-rated firm that borrows money and gets tax advantages can be worth more than the same firm with a AAA rating and no debt.

    So what are the real trade offs? The first and biggest benefit of debt is its tax treatment, with the tax benefits adding to firm value. Note, and this is said with no moral or ethical judgment attached to it, that this increase in value is coming from taxpayers and not from your operations becoming more valuable. A secondary benefit may come from imposing discipline on managers in public companies, with the need to make interest payments operating as a restraint on a headlong rush into poorly performing investments. On the other side of the ledger, the biggest concern you should have when you borrow money is that it increases the risk of bankruptcy, which if it happens, truncates business life, and even it does not, concerns about it happening can alter how customers, suppliers and investors interact with a business. The other cost that you face when you borrow money is that equity investors and lenders have very different interests, with equity seeking upside and lenders worrying about downside, and the costs of that conflict of interests plays out in covenants and restrictions on operating activity. The figure below summarizes these real trade offs.

The tax benefits versus bankruptcy cost trade off on debt is a simple and very powerful explainer of how much companies should borrow, but in the real world, there are companies that sometimes override the tradeoff and choose to borrow far more or far less than you would expect them to, and they are not necessarily being irrational. Here are three reasons why companies may choose a sub-optimal financing mix:

1. Shields against bankruptcy: If the biggest restraint on borrowing more is the fear of default, anything that reduces or eliminates that fear will cause companies to borrow more money. That default protection can come from governments acting as implicit or explicit guarantors of corporate debt, as was the case with Korean companies in the 1990s, or from seeing other companies in trouble being bailed out by the government, because they were too big to fail. 
2. Control versus Value: While businesses have the option of using either equity or debt to fund operations, raising fresh equity usually requires giving up ownership of the business to venture capitalists (at a private business) or to other public market investors (for public companies). For founders and family groups that value control over almost everything else, this can result in firms borrowing money, even though the fundamentals do not support the action. This can explain why Middle Eastern firms, many of which get no tax benefit from debt, may choose to borrow money to fund operations, usually with higher costs of capital, as well as the existence of venture debt, an almost absurd notion from a corporate finance standpoint, since you are lending to start-ups and young money-losing companies with unformed business models and 
3. Subsidized debt: If a business has access to debt with below-market interest rates, given default risk, it may make sense to borrow money at these subsidized rates. These debt subsidies are often granted to companies that are seen as delivering on a social purpose (green energy in the last decade) or a political/security interests (defense and infrastructure businesses), and you should therefore not be surprised if they all carry too much debt.

On the other side of the ledger, there are three reasons why companies may borrow less than they should:

1. Restrictive covenants: In markets where debt comes primarily from bankers, it is possible that the covenants that come with this debt are so onerous that businesses will choose to leave tax benefits on the table in order to preserve operating flexibility; this may explain why technology companies, even those with large and stable cash flows, often choose not to borrow money or if they have to, go directly to bond markets.
2. Overpriced equity: Financial markets make mistakes, and sometimes those mistakes may work in your favor as a company with your stock price soaring well above what you think is justifiable, given your fundamentals. In that case, you may choose to use equity, even if you have debt capacity, using your own overpriced shares as currency in funding acquisitions.
3. Regulatory constraints: In some countries and/or sectors, there may be regulatory restrictions on borrowing that cap how much debt you can take on, even though you have the capacity to carry more in debt. Those restrictions can take the form of limits on book debt ratios or on how much interest expense is tax deductible, as a function of revenues or EBITDA.

The picture below captures these frictional considerations:

In sum, the choices between debt and equity play out differently at different companies, depending not only on the characteristics of the company (tax rate, default risk etc.) but also on the management team making that decision on whether to borrow money. If you are an optimizer, by nature, you may this discussion too diffuse, since it points you in a direction (more or less debt) and not to a specific debt mix, but that is easily remedied, if you use the cost of capital as your optimizing tool to find the mix of debt and equity that minimizes your cost of capital. 

Download optimizer spreadsheet

Rather than try your patience by belaboring that process, I can point you in the direction of how that is done in my corporate finance class sessions, and with this tool.

Debt and Equity in 2025

    With this tradeoff on debt and equity in mind, let's turn to the data, and in particular, I plan to focus on the choices that companies made globally, on the financing question, in 2025. I will start by looking at the two forces that should have the greatest relevance in this decision, the tax benefits of debt and the default risk, and then look at the mixes of financing across sectors, industries and regions.

The Tax Landscape

    Any discussion of taxes has to start with reality checks. The first is that governments need tax revenues, to fund their spending, and corporations and businesses are a target, partly because they affect taxpayers (and voters) indirectly, rather than directly (as is the case with income and sales taxes). The second is that businesses do not like to pay taxes, and try to minimize the taxes they pay, mostly through legal means, with accountants, transfer pricing specialists and tax lawyers abetting, though they sometimes step over the line into tax evasion. When measuring the tax burden that businesses face, we have to distinguish between three measures of tax rates:

1. Marginal Tax Rates: The marginal tax rate reflects the tax rate you face on the last dollar of your taxable income, and thus comes from the statutory tax code of the domicile that the business operates in. While there are a few companies that try to report these tax rates, you are more likely to uncover them by going into the tax code. Fortunately, the leading accounting firms keep updated estimates of these marginal tax rates in the public domain, as do some tax watchdogs, and I used  The Tax Foundation for this year's updates across countries, and the numbers are in the picture below: 

   

   Download corporate tax rates, by country

   While your eye may be drawn to differences in corporate tax rates, across countries, these differences have narrowed, as the countries with the largest economies (and taxable business) are converging around a marginal tax rate of 25%. There are regional differences, with Latin America and Africa home to some of the highest corporate tax rates, and Eastern Europe and Russia home to some of the lowest. Clearly, there are exceptions within each region, with Ireland the leading outlier in Europe, with a marginal tax rate of 12%, and Paraguay in Latin America, with a marginal tax rate of 10%.
2. Effective tax rates: The effective tax rate is an accounting measure, reflecting the taxes paid and taxable  income line items in the income statement, which follows accrual accounting principles. The effective and marginal tax rates can deviate for many reason, including corporate income earned in other countries, tax deferral strategies and even differences between tax and reporting books. I estimated effective tax rates for the companies in my database, and report the averages, by sub-region of the world, in the table below:

   

   Corporate Marginal and Effective Tax Rates, by Country

   In the aggregate, the effective tax rates were lower than the marginal tax rates in about 60% of the companies in my sample, and the difference is a rough proxy for the effectiveness of a tax system, with marginal tax rates running close to or behind effective tax rates in more effective tax regimes. By that measure, India has the least effective tax code among the regions, with an effective tax rate of 22.33% and a marginal tax rate of 30%, followed by the United States and Japan, though the caveat would foreign sales in lower tax locales, in each of these cases. The tax rate statistics, broken down by industry, for global companies, is at this link, if you are interested.
3. Cash tax rates: The cash tax rates also come from accounting statements, with the information in the statement of cash flows used to convert accrual taxes paid to cash taxes paid, and are reflective of what companies actually pay to governments during the course of the year. In 2025, the average cash tax rate across companies with taxable income was 25.86% (21.02%) for global (US) firms, about 1% higher than the effective tax rate in both cases.

For the debt question, it is the marginal tax rate that is most relevant, at least for computing tax benefits, since interest expenses save you taxes at the margin; interest expenses get deducted to get to taxable income, and it is the last dollars of taxable income that thus get protected from paying taxes.

The Default/Distress Landscape

    In a world where companies never default, and you still get tax benefits from borrowing, companies push towards higher and higher debt ratios. In the real world, default acts as a brake on debt, with higher default risk translating into lower debt ratios. While default risk is company-specific, the exposure for default risk, across all companies, will vary over time, largely as a function of how well the economy is doing. The ratings agencies (Moody's, S&P and Fitch) track defaults on a year-to-year basis, and in 2025, they all recorded a drop in default rates across the globe, with US companies driving much of the decline. S&P, in its review of 2025 default and distress, reported that a drop in corporate defaults from 145 in 2024 to 117 to 2025, with the US share of defaults declining from 67% to 62%.  To provide historical context, I looked at corporate default rates on loans (using data from FRED) on a quarterly basis going back to 1986:

Corporate loan default rates

While the low defaults in 2025 were a positive sign for lenders, especially given the economic turmoil created by tariffs and trade wars, there were some worrying trends as well. In May 2025, Moody's estimate of the probability of default at US companies spiked to 9.2%, its highest value since the 2008 crisis. On the bond ratings front, you had more ratings downgrades than upgrades during the year, and almost $60 billion in corporate bonds slipped below investment grade during the  year.  Breaking down all rated companies, by S&P ratings class, and by region, at the end of 2025:

Source: S&P Cap IQ

The US has the highest percentage of listed companies with bond ratings, but even in the US, only 11.43% of companies carry that rating, and that percentage is far lower in other parts of the world. Among rated companies, the US has the highest percentage of below investment-grade ratings, suggesting that in much of the rest of the world, there is a self-selection that occurs, where only companies that believe that they will get high ratings are willing to go through the ratings process. Finally, at the start of 2026, there are only AAA rated-companies left in the world, at least according to S&P, in Johnson & Johnson and Microsoft. Looking at 2025, through the lens of default, the numbers look comforting, at least on the surface, with the number of defaults decreasing, but there was disquiet below, as bond buyers wrestled with the consequences of a changing economic world order, and worries about another crisis lurking in the wings. 

Debt Burden in 2025

    With the background data on tax rates and default risk in place, I will turn to measuring the debt in publicly traded firms, in 2025, and differences in debt burdens across companies, sectors and regions. That mission requires clarity on how to measure debt burdens, and the picture below offers the choices:

Broadly speaking, debt burden metrics can capture debt comfort, i.e., the buffer that businesses have built in to meet their debt obligations and debt level, where you look at debt as a percent of overall funding. In the former group, there are two proxies that you can use to gauge the borrowing buffer  - the interest coverage ratio, measuring how much companies have as operating income, relative to their interest expenses, and the debt as a multiple of EBITDA, capturing how many years it will take a company to pay off its debt, if current EBITDA is sustained. In the latter, I will look at debt as a percent of capital invested, using both accounting measures of capital invested (book value) and market value measures.

1. Debt comfort

    When companies borrow money, the contractual claims from that debt usually take two forms. The first is interest expenses, and ongoing claim that gives you tax benefits but has to be covered out of income generated each year, and the second is repayment of principal, which comes due at maturity. The interest coverage ratio focuses entirely on the former, and interest payments are scaled to how much a company generates in operating income:

Interest coverage ratio = Earnings before interest and taxes/ Interest expenses

This ratio is simple, with high values associated with less default risk and more safety, at least from a lending perspective. It is still powerful, and it remains the financial ratio that best explains differences in bond ratings across non-financial service companies, and I use it to estimate synthetic bond ratings for firms in my corporate financial analysis.

    The problem with interest coverage ratios is that they ignore the other contractual obligation that emerges from debt, which is principal payments due, and the ratio that is most often used to measure that exposure scales total debt at a firm to its earnings before interest, taxes and depreciation:

Debt to EBITDA = Total Debt/ EBITDA

With this ratio, lower values are associated with less default risk and more safety, because a firm, at least if it wanted to, could pay off its debt in fewer years with its operating cash flows.

    In the table below, I look at interest coverage ratios and debt to EBITDA values, by sector, for US and global companies, using the same approach I employed in my last update and reporting a ratio based on aggregated values as well as the distribution of the ratio across companies:

As you can see, with both the US and global groupings, technology companies have the largest safety buffers when it comes to debt, with very high interest coverage ratios and low debt to EBITDA, whereas real estate and utilities have the least buffers, with low interest coverage ratios and high debt to EBITDA. As always, the contrast between the aggregated and median values indicate that larger companies, not surprisingly, operate with stronger buffers than smaller companies in almost every sector grouping. Finally, the debt comfort numbers are not computed for financial service companies, for the same reasons that we did not compute costs of and returns on capital for these firms - debt to a bank is raw material and not capital.

2. Debt level 

    If you go back to the financial balance sheet structure that I started this post with, the debt measure that emerges is one that scales it to the equity invested in the firm (debt to equity) and to the capital invested (debt to capital). These measures have resonance in corporate finance in valuation, because they become drivers of the costs of equity and debt and ingredients in the cost of capital.That said, you can measure this ratio using book value debt to capital (or equity), where you stay with the values of debt and equity reported on accounting balance sheets or with market value debt to capital (and equity ratios), where you use market values for debt and equity. At the risk of sounding dogmatic, book value debt ratios should never come into play in financial analysis and it is market value ratios that matter for two reasons. The first relates back to all of the criticisms I had of accounting invested capital in the context of computing account returns - it is dated and skewed by accounting contradictions and actions. The second is that it is unrelated to what you are trying to measure in a cost of capital, which is what it would cost you to acquire the firm today, where it is market price that determines how much you have to pay, not book value. That said, there remain a fairly large subset of analysts and firms who swear allegiance to book value for a variety of reasons, most of which have no basis in reality. I report book and market debt to capital ratios for all publicly traded firms, broken down by sector for global and US companies:

As you can see, companies look significantly more debt-laden with book value numbers than with market value, and in sectors like technology, where accountants fail to bring the biggest assets on to the books, the difference is even starker. The results in this table reinforce the findings in the debt comfort table, with technology companies carrying very little debt (3-5% in market cap terms) and utilities and real estate carrying the highest. I also reported, on the aggregated numbers, the gross and net debt ratios, with the latter netting cash holdings from debt.

AI Investing and Debt

    In every data update post that I have written so far this year, AI has become a component of the discussion, reflecting the outsized role it played not just in market pricing during 2025, but also in business decisions made during the year. To see the connection between AI and debt, I will start with AI investing side, where hundreds of billions were spent by companies building AI infrastructure and large language models (LLMs) during 2025, with plans to spend more in the years to come. A sizable portion of this AI capital expenditure have come from big tech companies, with Meta, Alphabet, Amazon, Oracle and Microsoft all making large bets on the future of AI, and the extent of their investment is visible in the graph below, where I look at capital expenditures and cash acquisitions at these firms (with Broadcom added to the mix) from 2015 to 2025:

The shift at these firms from capital-light to capital-intensive models over this period has been staggering, with the collective investment in 2025 alone hitting $400 billion, with guidance suggesting that they are only getting started. It is worth noting that while big tech has garnered the AI cap ex headline, there are a whole host of other companies that are investing in AI architecture, which include real estate, data centers and power, and many of these companies are still not publicly listed. Going back to investment first principles, you can debate whether these companies can expect to generate positive net present value from their AI investments, and I have argued in earlier posts that it is very likely that they are collectively over investing, with over confidence and a fear of being left behind driving their both corporate investments and investor pricing, in keeping what you would expect when there is a big market delusion.

This big market delusion is a feature, not a bug, and we have seen it play out with dot com stocks in the 1990s, online advertising companies about ten years and even with cannabis stocks in the early years of their listing. The belief that the AI market will be huge, and have two or three big winners, is driving an investing frenzy not just at the big tech companies, but also in smaller start-ups and young firms, but the the market is not big enough to accommodate the expectations across all of these firms, and that will inevitably lead to a correction and clean up. 

    The AI investing boom enters the financing storyline, which is the focus for this post, because it needs immense amounts of capital. For many of the big tech companies, much of that capital has come from their existing businesses which are cash machines, although the AI cap ex will deplete the free cash flows available to return to shareholders. That said, though, the ramping up of capital investment has been so dramatic that even the cash-rich bit tech companies have turned to debt, as you can see in the graph below:

In 2025, the big tech companies collectively borrowed $160 billion, but given their cashflows and market capitalization, that debt does not put them at risk. For many of the smaller and lower-profile companies investing in this space, where internal cashflows are insufficient, there is a need for external capital, with some coming from equity and a significant portion coming from debt. It is in the context of the debt that I have to pick up on another storyline, which is the rise of private credit as an alternative to banks and the corporate bond market.  

As you look at the explosive growth of private credit in this graph, it is worth emphasizing that private credit has been available as an option for borrowers for as long as borrowing has been around, but its usage explode in the last two decades. As AI has increasingly taken a starring role in markets, evidence is accumulating that more private debt is being directed to financing the AI investment boom,. With more than $200 billion in private debt going to AI firms in 2025, AI-related debt is rising as a percent of private credit portfolios.  

    As private credit has grown as an option, core questions remain of what it brings to a market as  differentiating features that allow it to supplant more traditional lending alternatives, i.e. banking and the corporate bond market. Here are some of the reasons offered by private credit advocates for why it may be a preferred choice for entities, in general, and for those investing in AI architecture, in particular: 

1. Better default risk assessments: One of the arguments that private credit lenders make is that they have the technical know-how to use data, that banks and bond markets have been more averse to using or have been constrained from using, to get better assessments of default risk. Those assessments, assuming that they are right, allows private credit to lend to entities at rates that are lower than they would be charged, with conventional risk assessments. In principle, that is a solid rationale, but I am unclear about what data it is that traditional lenders are not utilizing that private credit can use, but it is possible that technology and access to the internals of borrowing entities may provide an edge. In fact, the only way to gauge whether this argument of better credit assessment holds up is with a credit shock, where defaults spike across the board.
2. Cashflows-based versus Asset-based lending: A second argument is that traditional lenders, and especially banks, are focused too much on the value of the assets that they are lending against and too little on the cash flows. It is true that bank lending in particular is too focused on asset value, but that focus would provide an opening for private credit in AI, only if AI data centers and architecture investments are poised to start delivering large and positive cash flows soon, and banks are holding back on lending them money. I am hard pressed to think of too many AI investments that have these near-term payoffs.
3. Speedier and more Flexible/Customized Responses: IThis may be the biggest selling point for private credit in the AI investment world, where the investing entities are not just spending billions on AI architecture, but are in a hurry to do so. The regulatory and institutional constraints built into bank lending will stretch the process out in time, and issuing bonds, even if it were an option, comes with its own delay components. In addition, the debt for AI investments may need far more customization than what banks and bond markets can offer, or are allowed to offer, giving private credit an advantage. The problem with speed and customization being the biggest sales pitches for private credit is that it can go with taking short cuts on due diligence and adding terms to loans that cut against prudence, and those can be fatal to lending businesses.

Clearly, these reasons for the presence of private debt have merit, but only to a subset of borrowers, mostly smaller and private, and without a long borrowing history, and for a subset of projects. None that these reasons resonate for the larger tech companies, which have options to borrow money quickly and at fair market rates both from banks and the bond market, and Google's recent hundred year bond issue is an indication of how much slack bond markets are willing to concede to these firms. When a private credit fund lends Meta for an AI investment, as Blue Owl did in this transaction, the skeptic in me sees either a below-market-rate loan or one with terms that no prudent lender would accept in a loan, and neither is a sustainable lending strategy in the long term. 

    The coming together of the two storylines on AI and private credit comes with a risk that may extend well beyond the players in these spaces. If you agree with my contention that companies are collectively over investing in AI, driven by the big market delusion, there will be a time when that delusion  dissipates and markets will have to correct. In an all or mostly-equity driven space, the pain will be borne by shareholders or owners of these companies, but while painful to them, its ripple effects will be limited. When debt enters the picture, as it has in the AI investment space, the effects of a correction will no longer be isolated to equity investors in these companies, and as private credit gets repriced (from the marking of debt down to reflect higher default risk), the pain to the rest of the economy increases. In effect, we will have a banking crisis created primarily by non-banking lenders behaving badly. We saw some of this start to happen in the last year, as the glow came off the AI rose, and S&P noted the stresses that it put on private credit players. Private credit has had a good run, in terms of delivering returns to investors in it, but it has, in my opinion, the relentless selling of it as an alternative investment class has made it much too big. A shakeout is overdue, which will separate the sloppy lenders from the good ones, and perhaps shrink private credit to healthier levels.

YouTube Video

Data links

1. Marginal and Effective tax rates, by country (January 2026)
2. Debt comfort ratios, by industry (US and Global)
3. Debt load ratios, by industry (US and Global)

Spreadsheet

1. Capital structure optimizer 

Paper on Big Market Delusion

1. The Big Market Delusion (with Brad Cornell)

Data Update Posts for 2026

1. Data Update 1 for 2026: The Push and Pull of Data
2. Data Update 2 for 2026: Equities get tested and pass again!
3. Data Update 3 for 2026: The Trust Deficit - Bonds, Currencies, Gold and Bitcoin!
4. Data Update 4 for 2026: The Global Perspective
5. Data Update 5 for 2026: Risk and Hurdle Rates
6. Data Update 6 for 2026: In Search of Profitability
7. Data Update 7 for 2026: Debt and Taxes
8. Data Update 8 for 2026: Dividends and Buybacks