On August 7, 2025, OpenAI officially released GPT-5, its latest flagship model, to the world.

A few weeks ago, I started receiving a stream of message about an Instagram post that I was allegedly starring in, where after offering my views on Palantir's valuation, I was soliciting investors to invest with me (or with an investment entity that had ties to me). I was not surprised, since I have lived with imitations for years, but I was bemused, since I don't have an Instagram account and have not posted on Facebook more than once or twice in a decade. In the last few days, those warnings have been joined by others, who have noted that there is now a video that looks and sounds like me, adding to the sales pitch with promises of super-normal returns if they reach out, and presumably send their money in. (Please don't go looking for these scams online, since the very act of clicking on them can expose you to their reach.)     I would like to think that readers of my books or posts, or students in my classes, know me well enough to be able to tell that these are fakes, and while this is not the first time I have been targeted, it is clear that AI has upped the ante, in terms of creating posts and videos that look authentic. In response, I cycled through a series of emotions, starting with surprise that there are some out there who think that using my name alone will draw in investors, moving on to anger at the targeting of vulnerable investors and ending with frustration at the social media platforms that allow these fakes to exist. As a teacher, though, curiosity beat out all of these emotions, and I thought that the best thing that I can do, in addition to the fruitless exercise of notifying the social media companies about the fakes, is to talk about what these AI imitators got right, what they were off target on and what they got wrong in trying to create these fakes of me. Put simply, I plan to grade my AI imitator, as I would any student in my class, recognizing that being objective in this exercise will be tough to do. In the lead-in, though, I have to bore you with details of my professional life and thought process, since that is the key to creating a general framework that you will be able to use to detect AI imitations, since the game will only get more sophisticated in the years to come. An Easy Target?     In a post last year, I talked about a bot in my name, that was in development phase at NYU, and while officially sanctioned, it did open up existential challenges  for me. In discussing that bot, I noted that this bot had accessed everything that I had ever written, talked about or valued in my lifetime, and that I had facilitated its path by making that access easy. I will explain my rationale for the open access, and provide you with the links if you want to get to them, hoping to pre-empt those who will try to charge you for that content. My Open Access Policy     I have said this before, but there is no harm in saying it again, but I am a teacher, first and foremost, and almost every choice I make in my profession life reflects that mindset. A teacher, like an actor or singer, craves an audience, and the larger and more enthusiastic that audience, the better. When I started teaching in 1986, my audience was restricted to those in my physical classroom at NYU's business school, and my initial attempts at expanding that audience were very limited. I had video recorders set up to record my lectures, made three copies of each lecture tape, and put them on the shelves at NYU's library for patrons to check out and watch. The internet, for all of its sins, changed the game for me, allowing me to share not only class materials (slides, exams) but also my lecture videos, in online formats. Though my early attempts to make these conversions were primitive, the technology for recording classes and putting them online has made a quantum leap. In spring 2025, every one of my NYU classes was recorded by cameras that are built into classroom, the conversions to online videos happened in minutes, right after the class is done, and YouTube has been a game changer, in allowing access to anyone with an internet connection anywhere in the world.     As the internet has expanded its reach, and social media platforms have joined the mix, I have also shared the other components that go into my classes more widely, starting with the data on industry averages that I need and use in my own valuations, the spreadsheets that contain these valuations and blog posts on markets and companies and any other tools that I use in my own analyses. While I am happy to receive compliments for the sharing and praise for being unselfish, the truth is that my sharing is driven less by altruism (I am no Mother Theresa!) and more  by two other forces. The first is that, as I noted in my post on country equity risk premiums last week, there much of what I know or write about is pedestrian, and holding it in secret seems silly. The second is that, while I am not easily outraged, I am driven to outrage by business consultants and experts who state the obvious (replacing words you know with buzzwords and acronyms), while making outrageous claims of what they can deliver and charging their customers absurd amounts for their advice and appraisals. If I can save even a few of these customers from making these payments, I consider it to be a win. My Sharing Spots     Everything that I have ever written, worked on or taught is somewhere online, almost always with no protective shields (no passwords or subscriptions), and there are four places where you can find them: Webpage: The oldest platform for my content remains my webpage, damodaran.com, and while it can be creaky, and difficult to navigate, it contains the links to my writing, teaching, data, spreadsheets and other tools.  Teaching: I teach two classes at Stern, corporate finance and valuation, and have four other classes - a lead-into-valuation accounting class, a made-for-finance statistics class, a class on investment philosophies and one on corporate life cycles, and I described these classes in a post on teaching at the start of 2025. You can find them all by going to the teaching link on my webpage, https://people.stern.nyu.edu/adamodar/New_Home_Page/teaching.html including my regular classes (class material, lecture notes, exams and quizzes and webcasts of the classes) in real time, as well as archived versions from previous semesters. In addition, the online classes are at the same link, with material, post- class tests and webcasts of sessions for each class. This is also the place where you can find links to seminars that I teach in the rest of the world, with slides and materials that I used for those classes (though I have been tardy about updating these). Data: At the start of every year for the last three decades, I have shared my analysis of data on publicly traded companies, breaking down the data into corporate finance and valuation categories. This link, https://people.stern.nyu.edu/adamodar/New_Home_Page/data.html, will take you to the entry page, and you can then either access the most recent data (from the start of 2025, since I update only once a year, for most datasets) or archived data (from previous years). My raw data comes from a variety of sources, and in the interests of not stepping on the toes of my data providers, my data usually reflects industry averages, rather than company-specific data, but it does include regional breakdowns: US, Europe, Emerging Markets (with India and China broken out individually, Australia & Canada & New Zealand) and Japan.   Spreadsheets: I am not an Excel ninja, and while my spreadsheet-building skills are adequate, my capacity to make them look polished is limited. I do share the spreadsheets that I use in my classes and work here, with my most-used (by me) spreadsheet being one that I use to value most companies at this link, with a webcast explaining its usage. Books: I have written eleven books and co-edited one, spread out across corporate finance, valuation and investing, and you can find them all listed here. Many of these books are in their third or fourth editions, but with each one, you should find a webpage that contains supplementary material for that book or edition (slides, answers to questions at the end of each chapter, data, spreadsheets backing the examples). This is the only section of the spreadsheet where you may encounter a gatekeeper, asking you for a password, and only if you seek access to instructor material. If you are wondering what is behind the gate, it is only the powerpoint slides, with my notes on each slide, but the pdf versions of these slides should be somewhere on the same page, without need for a password. Papers: I don't much care much for academic research, but I do like to write about topics that interest or confound me, and you can find these papers at this link. My two most widely downloaded papers are updates I do each year on the equity risk premium (in March) and country risk premiums (in July). Much of the material in these papers has made its way into one or more of my books, and thus, if you find the books unaffordable, you can get that material here for free. Blog posts: I will confess that when I write my first blog post on September 17, 2008, I had no idea what a blog was, what I was doing with it, and whether it would last through the following week. In the years since, this blog has become my first go-to, when I have doubts or questions about something, and I am trying to resolve those doubts for myself. In short, my blog has becoming my therapy spot, in times of uncertainty, and I have had no qualms about admitting to these doubts. During 2020, as COVID made us question almost everything we know about markets and the economy, for instance, I posted on where I was in the uncertainty spectrum every week from February 14, 2020 (when the virus became a global problem, not one restricted to China and cruise ships) to November 2020, when the vaccine appeared. You can get all of those posts in one paper, if you click on this link. While my original blog was on Google, in the last two years, I have replicated these posts on Substack (you need to be a subscriber, but it is free) and on LinkedIn. If you are on the latter, you are welcome to follow me, but I have hit my connections limit (I did not even know there was one, until I hit it) and am unable to add connections. YouTube: For the last decade, I have posted my class videos on YouTube, grouping them into playlists for each class. You can start with the link to my YouTube channel here, but if you are interested in taking a class, my suggestion is that you click on the playlists and pick on the one that corresponds to the class. Here, for instance, are my links to my Spring 2025 MBA valuation class and my Spring 2025 Corporate Finance class. Starting about a decade, I have also accompanied every one of my blog posts with a YouTube video, that contains the same material, and you can find those posts in its own (very long) playlist.  X (Twitter): Some of you have strong feelings about X, with some of those feelings reflecting your political leanings and others driven by the sometimes toxic posting on the platform. I have been a user of the platform since April 2009, and I have used it as a bulletin board, to alert people to content being posted elsewhere. In fact, outside of these "alert" posts, I almost never post on X, and steer away as far away as I can from debates and discussions on the platform, since a version of Gresham's law seems to kick in, where the worst and least informed posters hijack the debate and take it in directions that you do not want it to go. I cannot think of a single item of content that I have produced in the last decade that is not on one of these platforms, making my professional life an open book, and thus also accessible to any AI entity. The Damodaran bot that I wrote about last year has access to all of this material, and while I signed off on that and one other variant, there are multiple unauthorized versions that have been works-in-progress.  The Commonalities     My content has taken many forms including posts, videos, data and spreadsheets, and is on multiple platforms, but there are a few common features that they share: Low tech: I am decidedly low tech, and it shows in my sharing. My website looks like it was designed two decades ago, because it was, and contains none of the bells and whistles that make for a modern website. My blog remains on Google blogger, notwithstanding everything I have been told about how using WordPress would make it more attractive/adaptable, and my posts are neither short nor punchy. Every week, I get people reaching out to me to tell me that my YouTube videos are far too long and verbose, and that I would get more people watching with shorter videos and catchier descriptions, and much as I appreciate their offers to help, I have not taken them up on it., In addition, I shoot almost every one of my videos in my office, sometimes with my dog in the background, and often with ambient noise and mistakes embedded, making them definitely unpolished.  On twitter, I have only recently taken to stringing tweets together and I have never used the long text version that some professional twitter users have mastered. In my defense, I could always claim that I am too old to learn new tricks, but the truth is that I did not start any of my sharing as a means to acquiring a larger social media following, and it may very well be true that keeping my presence low-tech operates as a screener, repelling mismatched users. Process over product: In my writing and teaching, I am often taken to task for not getting to the bottom line (Is the stock cheap or expensive? Should I buy or sell?) quickly, and spending so much time on the why and how, as opposed to the what. Much as my verbosity may frustrate you, it reflects what I think my job is as a teacher, which is to be transparent about process, i.e., explain how I reasoned my way to getting an answer than giving you my answer. Pragmatism over Purity: Though I am often criticized for being an “academic”, I am a terrible one, and if there were an academic fraternity, I would be shunned. I view much of an academic research as navel gazing, and almost everything I write and teach is for practitioners. Consequently, I am quick to adapt and modify models to make them fit both reality and the available data, and make assumptions that would make a purist blanch.  No stock picks or investment advice: In all my years of writing about and valuing markets and individual stocks, I have tried my best to steer away from making stock picks or offering investment advice. That may sound odd, since so much of what I do relates to valuation, and the essence of valuation is that you act on your estimates of value, but here is how I explain the contradiction. I value stocks (like Meta or Nvidia or Amazon or Mercado Libre) and I act (buy or sell) those stocks, based on my valuations, but it is neither my place nor my role to try to get other people to do the same. That said, I will share my story and valuation spreadsheet with you, and if you want to adapt that story/spreadsheet to make it your own, I am at peace with that choice, even if it is different from mine. The essence of good investing is taking ownership of your investment actions, and it is antithetical to that view of the world for me or anyone else to be telling you what to buy or sell. No commercial entanglements: If you do explore my content on any of the platforms it is available on, you will notice that they are free, both in terms of what you pay and how you access them. In fact, none of them are monetized, and if you do see ads on my YouTube videos, it is Google that is collecting the revenue, not me. One reason for this practice is that I am lazy, and monetizing any of these platforms requires jumping through hoops and catering to advertisers that I neither have the time nor the inclination to do. The other is that I believe (though this may be more hope than truth) that one of the reasons that people read what I write or listen to me is because, much as they may disagree with me, I am perceived as (relatively) unbiased. I fear that formalizing a link with any commercial entity (bank, consultant, investor), whether as advisor, consultant or as director, opens the door to the perception of bias. The one exception to the "no commercial entranglements' clause is for my teaching engagements, with the NYU Certificate program and for the handful of valuation seminars I teach in person in the rest of the world. I am grateful that NYU has allowed me to share my class recordings with the world, and I will not begrudge them whatever they make on my certificate classes, though I do offer the same content for free online, on my webpage. I am also indebted to the people and organizations that manage the logistics of my seminars in the rest of the world, and if I can make their life easier by posting about these seminars, I will do so.     The Imitation Game     Given that my end game in sharing is to give access to people who want to use my material, I have generally taken a lax view of others borrowing my slides, data, spreadsheets or even webcasts, for their own purposes. For the most part, I categorize this borrowing as good neighbor sharing, where just as I would lend a neighbor a key cooking ingredient to save them the trouble of a trip to the grocery store, I am at peace with someone using my material to help in their teaching, save time on a valuation or a corporate finance project, prepare for an interview, or even burnish their credentials. An acknowledgement, when this happens, is much appreciated, but I don't take it personally when none is forthcoming.  There are less benign copycat versions of the imitation game - selectively using data from my site to back up arguments, misreading or misinterpreting what I have said and reproducing large portions of my writing without acknowledgement. To be honest, if made aware of these transgressions, I have gently nudged the culprits, but I don't have a legal hammer to follow up. The most malignant variations of this game are scams, where the scammers use my content or name to separate people from their money - the education companies that used my YouTube videos and charge for classes, the data sites that copy my data or spreadsheets and sell them to people, and the valuation/investment sites that try to get people to invest money, with my name as a draw. Until now, I have tried, as best as I can, to let people know that they are being victimized, but for the most part, these scams have been so badly designed that they have tended to collapse under the weight of their own contradictions. It is clear to me that AI is now going to change this game, and that I will have to think about new ways to counter its insidious reach. To get a measure of what the current AI scams that are making the rounds get right and wrong, I did take the time to take a closer look at both the Instagram post and the fake video that are making the rounds.  What they get right: The Instagram post, which is in shown below, uses language that clearly is drawn from my posts and an image that is clearly mine. Not only does this post reflect the way I write, but it also picked Nvidia and  Palantir as the two firms to highlight,  the first a company that I own and have valued on my blog, and the second a company that I have been talking about as one that I am interested in owning, at the right price, giving it a patina of authenticity. The video looks and sounds like me, which should be no surprise since it had thousands of hours of YouTube videos to use as raw data. Using a yiddish word that I picked up in my days in New York, I have the give the scammers credit for chutzpah, on this front,, but I will take a notch off the grade, for the video's slickness, since my videos have much more of a homemade feel to them. What they struggled with most: The scam does mention that Palantir is "overhyped", a word that I use rarely, and while it talks about the company’s valuation, it is cagey about what that value is and there is little of substance to back up the claim. Palantir is a fascinating company, but to value it, you need a story of a data/software firm, with two channels for value creation, one of which looks at the government as a customer (a lower-margin, stickier and lower growth business) and the other at its commercial market (higher margin, more volatile and higher growth). Each of the stories has shades of grey, with the potential for overlap and conflict, but this is not a company where you can extrapolate the past, slap numbers on revenue growth and profitability, and arrive at a value. This post not only does not provide any tangible backing for its words in terms of value, but it does not even try. If these scammers had truly wanted to pull this off, they could have made their AI bot take my class, construct a plausible Palantir story, put it into my valuation spreadsheet and provide it as a link.  What they get wrong: To get a sense of what this post gets wrong, you should revisit the earlier part of the post where I talk about my sharing philosophy, and with as much distance as I can muster, here are the false notes in this scam. First, this scam pushes people to join an investment club, where I will presumably guide them on what to buy or sell. Given that my view of clubs is very much that of Groucho Marx, which is that I would not be belong to any club which would admit me as a member, the notion of telling people which stocks to buy cuts against every grain of my being. Second, there is a part of this scam where I purportedly promise investors who decide to partake that they will generate returns of 60% or higher, and as someone who has chronicled that not only do most active investors not keep up with the market, and argued that anyone who promises to deliver substantially more than the market in the long term is either a liar or fraud, this is clearly not me.  In sum, there is good news and bad news in this grading assessment. The good news is that this AI scam gets my language and look right, but it is sloppily done in terms of content and capturing who I am as a person. The bad news is that it if this scammer was less lazy and more willing to put in some work, even with the current state of AI, it would have been easy to bring up the grades on content and message. I will wager that the Damodaran Bot that I mentioned earlier on in this post that is being developed at NYU Stern would have created a post that would have been much more difficult for you to detect as fake, making it a Frankenstein monster perhaps in the making. The worse news is that AI technology is evolving, and it will get better on every one of these fronts at imitating others, and you should prepare yourself for a deluge of investment scams. An AI Protective Shield     I did think long about writing this post, wondering whether it would make a difference. After all, if you are a frequent reader of this blog or have read this post all the way down to this point, it is unlikely that you were fooled by the Instagram post or video. It remains an uncomfortable truth that the people most exposed to these scams are the ones who have read little or none of what I have written, and I wish there were a way that I could pass on the following suggestions on how they can protect themselves against the other fakes and scams that will undoubtedly be directed at them.  "Looks & sounds like" not good enough: Having seen the flood of fake AI videos in the news and on social media, I hope that you have concluded that “looks and sounds Iike” is no longer good enough to meet the authenticity test. This remains AI’s strongest suit, especially in the hands of the garden variety scammer, and you should prepare yourself for more fake videos, with political figures, investing luminaries and experts targeted. Steer away from arrogance & hype: I have always been skeptical of the notion that there is “smart” money, composed of investors who know more than the rest of us and are able to beat the market consistently, and for long periods. For the most part, when you see a group of investors (hedge funds, private equity) beating the market, luck is more of a contributor as skill, and success is fleeting. In a talk on the topic, I argued that investors should steer away from arrogance and bombast, and towards humility, when it comes to who they trust with their money, and that applies in spades in the world of AI scams. Since most scammers don’t understand the subtlety of this idea, screening investment sales pitches for outlandish claims alone will eliminate most scams. Do your homework: If you decide to invest with someone, based upon a virtual meet or sales pitch, you should do your homework and that goes well beyond asking for their track records in terms of performance. In my class on investment philosophies, I talk about how great investors through the ages have had very different views of markets and ways of making money, but each one has had an investment philosophy that is unique, consistent and well thought through. It is malpractice to invest with anyone, no matter what their reputation for earning high returns, without understanding that person’s investment philosophy, and this understanding will also give you a template for spotting fakes using that person’s name.  Avoid ROMO & FOMO: In my investing classes, I talk about the damage that ROMO (regret over missing out) and FOMO (fear of missing out) can do to investor psyches and portfolio.  With ROMO (regret over missing out), where you look back in time and regret not buying Facebook at its IPO price in 2012 or selling your bitcoin in  November 2013, when it hit $1000, you expose yourself to two emotions. The first is jealousy, especially at those who did buy Facebook at its IPO or have held on to their bitcoin to see its price hit six digits. The second is that you start buying into conspiracy theories, where you convince yourself that these winners (at least in the rear view mirror) were able to win, because the game was fixed in their favor. Both make you susceptible to chasing after past winners, and easy prey for vendors of conspiracies. With FOMO (fear of missing out), your overwhelming concern is that you will miss the next big multi-bagger, an investment that will increase five or ten fold over the next year or two. The emotion that is triggered is greed, leading you to overreach in your investing, cycling through your investments, as most of them fall short of your unrealistic expectations, and searching for the next “big thing”, making you susceptible to anyone offering a pathway to get there. Much as we think of scammers as the criminals and the scammed as the victims, the truth is that scams are more akin to tangos, where each side needs the other. The scammer’s techniques work because they trigger the emotions (fear, greed) of the scammed, to respond, and AI will only make this easier to do. Looking to regulators or the government to protection will do little more than offer false comfort, and the best defense is “caveat emptor” or “buyer beware”.  YouTube Video Links Webpage: https://pages.stern.nyu.edu/~adamodar/New_Home_Page/home.htm  Blog:  (1) Google: https://aswathdamodaran.blogspot.com  (2) Substack: https://aswathdamodaran.substack.com  (3) LinkedIn: https://www.linkedin.com/in/aswathdamodaran/  YouTube https://www.youtube.com/channel/UCLvnJL8htRR1T9cbSccaoVw X: https://x.com/aswathdamodaran?lang=en

And ten more thoughts on AI.

At the start of July, I updated my estimates of equity risk premiums for countries, in an semiannual ritual that goes back almost three decades. As with some of my other data updates, I have mixed feelings about publishing these numbers. On the one hand, I have no qualms about sharing these estimates, which I use when I value companies, because there is no secret sauce or special insight embedded in them. On the other, I worry about people using these premiums in their valuations, without understanding the choices and assumptions that I had to make to get to them. Country risk, in particular, has many components to it, and while you have to ultimately capture them in numbers, I wanted to use this post to draw attention to the many layers of risk that separate countries. I hope, and especially if you are a user of my risk premiums, that you read this post, and if you do have the time and the stomach, a more detailed and much longer update that I write every year. Country Risk - Dimensions     When assessing business risk from operating in a country, you will be affected by uncertainty that arises from almost every source, with concerns about political structure (democracies have very different risk profiles than authoritarian regimes), exposure to violence (affecting both costs and revenues),  corruption (which operates an implicit tax) and legal systems (enforcing ownership rights) all playing out in business risk. I will start with political structure, where the facile answer is that it less risky to operate a business in a democracy than in an authoritarian regime, but where the often unpalatable truth is that each structure brings its own risks. With democracies, the risk is that newly elected governments can revisit, modify or discard policies that a previous government have adopted, requiring businesses to adapt and change to continuous changes in policy. In contrast, an authoritarian government can provide long term policy continuity, with the catch being that changes in the government, though infrequent, can create wrenching policy shifts that businesses have to learn to live with. Keeping the contrast between the continuous risk of operating in a democracy and the discontinuous risk in an authoritarian structure in mind, take a look at this picture of how the world looked in terms of democracy leading into 2025: Source: Economist Intelligence Unit (EIU) It is worth noting that there are judgment calls that the Economist made in measuring democracy that you and I might disagree with, but not only is a large proportion of the world under authoritarian rule, but the trend lines on this dimension  also have been towards more authoritarianism in the last decade.         On the second dimension, exposure to violence, the effects on business are manifold. In addition to the threat that violence can affect operations, its presence shows up as higher operating costs (providing security for employees and factories) and as insurance costs (if the risks can be insured). To measure exposure to violence, from both internal and external sources, I draw on measures developed and updated by the Institute of  Economics & Peace across countries in 2024: Institute of Economics & Peace The Russia-Ukraine war has caused risk to flare up in the surrounding states and the Middle East and central Africa continue to be risk cauldrons, but at least according to the Institute's measures, the parts of the world that are least exposed to violence are in Northern Europe, Australia and Canada. Again, there are judgments that are made in computing these scores that will lead you to disagree with specific country measures (according the Peace Institute, the United States and Brazil have higher exposures to violence than Argentina and Chile, and India has more exposure to violence than China), but the bottom line is that there are significant differences in exposure to violence across the world.          Corruption is a concern for everyone, but for businesses, it manifests in two ways. First, it puts more honest business operators at a disadvantage in a corrupt environment, since they are less willing to break the rules and go along with corrupt practices than their less scrupulous competitors. Second, even for those businesses that are willing to play the corruption game, it creates costs that I would liken to an implicit tax that reduces profits, cash flows and value. The measure of corruption that I use comes from Transparency International, and leading into July 2025, and the heat map below captures corruption scores (with higher scores indicating less corruption), as well as the ten most and least corrupt countries in the world:  Transparency International As you can see from the map, there are vast swaths of the world where businesses have to deal with corruption in almost every aspect of business, and while some may attribute this to cultural factors, I have long argued that corruption almost inevitably follows in bureaucratic settings, where you need licenses and approvals for even the most trivial of actions, and the bureaucrats (who make the licensing decisions) are paid a pittance relative to the businesses that they regulate.           As a final component, I look at legal systems, especially when it comes to enforcing contractual agreements and property rights, central to running successful businesses. Here, I used estimates from the IPRI, a non-profit institution that measures the quality of legal systems around the world. In their latest rankings from 2024, here is how countries measured up in 2024: Property Rights Alliance In making these assessments, you have to consider not just the laws in place but also the timeliness with which these laws get enforced, since a legal system where justice is delayed for years or even decades is almost as bad as one that is capricious and biased.  Country Risk - Measures     The simplest and most longstanding measure of country risk takes the form of sovereign ratings, with the same agencies that rate companies (S&P, Moody's and Fitch) also rating countries, with the ratings ranging from Aaa (safest) to D (in default). The number of countries with sovereign ratings available on them has surged in the last few decades; Moody’s rated 13 countries in 1985, but that number increased to 143 in 2025, with the figure below listing the number of rated countries over time: Note that that the number of Aaa rated countries stayed at eleven, even while more countries were rated, and has dropped from fifteen just a decade ago, with the UK and France losing their Aaa ratings during that period. In May 2025, Moody's downgraded the United States, bringing them in line with the other ratings agencies; S&P downgraded the US in 2011 and Fitch in 2023. The heat map below captures sovereign ratings across the world in July 2025: Moody's While sovereign ratings are useful risk measures, they do come with caveats. First, their focus on default risk can lead them to be misleading measures of overall country risk, especially in countries that have political risk issues but not much default risk; the Middle East, for instance, has high sovereign ratings. Second, the ratings agencies have blind spots, and some have critiqued these agencies for overrating European countries and underrating Asian, African and Latin American countries. Third, ratings agencies are often slow to react to events on the ground, and ratings changes, when they do occur, often lag changes in default risk.     If you are leery about trusting ratings agencies, I understand your distrust, and there is an alternative measure of sovereign default risk, at least for about half of all countries, and that is the sovereign credit default swap (CDS) market, which investors can buy protection against country default. These market-determined numbers will reflect events on the ground almost instantaneously, albeit with more volatility than ratings. At the end of June 2025, there were about 80 countries with sovereign CDS available on them, and the figure below captures the values: The sovereign CDS spreads are more timely, but as with all market-set numbers, they are subject to mood and momentum swings, and I find using them in conjunction with ratings gives me a better sense of sovereign default risk.     If default risk seems like to provide too narrow a focus on countr risk, you can consider using country risk scores, which at least in principle, incorporate other components of country risk. There are many services that estimate country risk scores, including the Economist and the World Bank, but I have long used Political Risk Services (PRS) for my scores.. The PRS country risk scores go from low to high, with the low scores indicative of more country risk, and the table below captures the world (at least according to PRS): Political Risk Services (PRS) There are some puzzling numbers here,  with the United States coming in as riskier than Vietnam and Libya, but that is one reason why country risk scores have never acquired traction. They vary across services, often reflecting judgments and choices made by each service, and there is no easy way to convert these scores into usable numbers in business and valuation or compare them across services.      Country Risk - Equity Risk Premiums     My interest in country risk stems almost entirely from my work in corporate finance and valuation, since this risk finds its way into the costs of equity and capital that are critical ingredients in both disciplines. To estimate the cost of equity for an investment in a risky country. I will not claim that the approaches I use to compute equity risk premiums for countries are either original or brilliant, but they do have the benefit of consistency, since I have used them every year (with an update at the start of the year and mid-year) since the 1990s.      The process starts with my estimate of the implied equity risk premium for the S&P 500, and I make this choice not for parochial reasons but because getting the raw data that you need for the implied equity risk premium is easiest to get for the S&P 500, the most widely tracked index in the world. In particular, the process requires data on dividends and stock buybacks on the stocks in the index, as well as expected growth in these cash flows over time, and involves finding the discount rate (internal rate of return) that makes the present value of cash flows equal to the level of the index. On June 30, 2025, this assessment generated an expected return of 8.45% for the index: Download ERP spreadsheet Until May 2025, I just subtracted the US 10-year treasury bond rate from this expected return, to get to an implied equity risk premium for the index, with the rationale that the US T.Bond rate is the riskfree rate in US dollars. The Moody’s downgrade of the US from Aaa to Aa1 has thrown a wrench into the process, since it implies that the T.Bond rate has some default risk associated with it, and thus incorporates a default spread. To remove that risk, I net out the default spread associated with Aa1 rating from the treasury rate to arrive at a riskfree rate in dollars and an equity risk premium based on that: Riskfree rate in US dollars       = T.Bond rate minus Default Spread for Aa1 rating                                                             = 4.24% - 0.27% = 3.97% Implied equity risk premium for US = Expected return on S&P 500 minus US $ riskfree rate                                                             = 8.45% - 3.97% = 4.48% Note that this approach to estimating equity risk premiums is model agnostic and reflects what investors are demanding in the market, rather than making a judgment on whether the premium is right or what it should be (which I leave to market timers).        To get the equity risk premiums for other countries, I need a base premium for a mature market, i.e., one that has no additional country risk, and here again, the US downgrade has thrown a twist into the process. Rather than use the US equity risk premium as my estimate of the mature market premium, my practice in every update through the start of 2025, I adjusted that premium (4.48%) down to take out the US default spread (0.27%), to arrive at the mature market premium of 4.21%. That then becomes the equity risk premium for the eleven countries that continue to have Aaa ratings, but for all other countries, I estimate default spreads based upon their sovereign ratings. As a final adjustment, I scale these default spreads upwards to incorporate the higher risk of equities, and these become the country risk premiums, which when added to the mature market premium, yields equity risk premiums by country. The process is described below: Download spreadsheet The results from following this process are captured in the picture below, where I create both a heat map based on the equity risk premiums, and report on the ratings, country risk premiums and equity risk premiums, by country: Download equity risk premium, by country If you compare the equity risk premium heat map with the heat maps on the other dimensions of country risk (political and legal structures, exposure to violence and corruption), you will notice the congruence. The parts of the world that are most exposed to corruption and violence, and have capricious legal systems, tend to have higher equity risk premiums. The effects of the US ratings downgrade also manifest in the table, with the US now having a higher equity risk premium than its Aaa counterparts in Northern Europe, Australia and Canada. A User's Guide      My estimates of equity risk premiums, by country, are available for download, and I am flattered that there are analysts that have found use for these number. One reason may be that they are free, but I do have concerns sometimes that they are misused, and the fault is mine for not clarifying how they should be used. In this section, I will lay out steps in using these equity risk premiums in corporate finance and valuation practice, and  if I have still left areas of  grey, please let me know. Step 1: Start with an understanding of what the equity risk premium measures     The starting point for most finance classes is with the recognition that investors are collectively risk averse, and will demand higher expected returns on investments with more risk. The equity risk premium is a measure of the “extra” return that investors need to make, over and above the riskfree rate, to compensate for the higher risk that they are exposed to, on equities collectively. In the context of country risk, it implies that investments in riskier countries will need to earn higher returns to beat benchmarks than in safer countries. Using the numbers from July 2025, this would imply that investors need to earn 7.46% more than the riskfree rate to invest in an average-risk investment in India, and 10.87% more than the riskfree rate to invest in an average risk investment in Turkey.     It is also worth recognizing how equity risk premiums play out investing and valuation. Increasing the equity risk premium will raise the rate of return you need to make on an investment, and by doing so, reduce its value. That is why equity risk premiums and stock prices move inversely, with the ERP rising as stock prices drop (all other thins being held constant) and falling as stock prices increase.  Step 2: Pick your currency of analysis (and estimate a riskfree rate)     I start my discussions of currency in valuation by positing that currency is a choice, and that not only can you assess any project or value any company in any currency, but also that your assessment of project worth or company value should not be affected by that choice. Defining the equity risk premium as the extra return that investors need to make, over and above the risk free rate, may leave you puzzled about what riskfree rate to use, and while the easy answer is that it should be the riskfree rate in the currency you chose to do the analysis in, it is worth emphasizing that this riskfree rate is not always the government bond rate, and especially so, if the government does not have Aaa rating and faces default risk. In that case, you will need to adjust the government bond rate (just as I did with the US dollar) for the default spread, to prevent double counting risk.   Staying with the example of an Indian investment, the expected return on an average-risk investment in Indian rupees would be computed as follows: Indian government bond rate on July 1, 2025 = 6.32% Default spread for India, based on rating on July 1, 2025 = 2.16% Indian rupee risk free rate on July 1, 2025 = 6.32% - 2.16% = 4.16% ERP for India on July 1, 2025 = 7.46% Expected return on average Indian equity in rupees on July 1, 2025 = 4.16% + 7..46% = 11.62% Note also that if using the Indian government bond rate as the riskfree rate in rupees, you would effectively be double counting Indian country risk, once in the government bond rate and once again in the equity risk premium.     I know that the ERP is in dollar terms, and adding it to a rupee riskfree rate may seem inconsistent, but it will work well for riskfree rates that are reasonably close to the US dollar risk free rate. For currencies, like the Brazilian real or Turkish lira, it is more prudent to do your calculations entirely in US dollars, and convert using the differential inflation rate: US dollar riskfree rate on July 1, 2025 = 3.97% ERP for Turkey on July 1, 2025 = 10.87% Expected return on average Turkish equity in US $ on July 1, 2025 = 3.97% + 10.87% = 14.84% Expected inflation rate in US dollars = 2.5%; Expected inflation rate in Turkish lira = 20% Expected return on average Turkish equity Turkish lira on July 1, 2025 = 1.1484 *(1.20/1.025) -1 = 34.45% Note that this process scales up the equity risk premium to a higher number for high-inflation currencies. Step 3: Estimate the equity risk premium or premiums that come into play based on operations    Many analysts use the equity risk premiums for a country when valuing companies that are incorporated in that country, but I think that is too narrow a perspective. In my view, the exposure to country risk comes from where a company operates, not where it is incorporated, opening the door for bringing in country risk from emerging markets into the cost of equity for multinationals that may be incorporated in mature markets. I use revenue weights, based on geography, for most companies, but I am open to using production weights, for natural resource companies, and even a mix of the two.  In corporate finance, where you need equity risk premiums to estimate costs of equity and capital in project assessment, the location of the project will determine which country’s equity risk premiums come into play. When Amazon decides to invest in a Brazilian online retail project, it is the equity risk premium for Brazil that should be incorporated, with the choice of currency for analysis determining the riskfree rate.  Step 4: Estimate project-specific or company-specific risk measures and costs     The riskfree rate and equity-risk premiums are market-wide numbers, driven by macro forces. To complete this process, you need two company-specific numbers: Not all companies or projects are average risk, for equity investors in them, and for companies that are riskier or safer than average, you need a measure of this relative risk. At the risk of provoking those who may be triggered by portfolio theory or the CAPM, the beta is one such measure, but as I have argued elsewhere, I am completely at home with alternative measures of relative equity risk. The cost of equity is calculated as follows:  Cost of equity = Riskfree rate + Beta × Equity Risk Premium The beta (relative risk measure) measures the risk of the business that the company/project is in, and for a diversified investor, captures only risk that cannot be diversified away. While we are often taught to use regressions against market indices to get these betas, using industry-average or bottom-up betas yields much better estimates for projects and companies. For the cost of debt, you need to estimate the default spread that the company will face. If the company has a bond rating, you can use this rating to estimate the default spread, and if it is not, you can use the company's financials to assess a synthetic rating. Cost of debt =Riskfree Rate + Default spread Harking back to the discussion of riskfree rates, a company in a country with sovereign default risk will often bear a double burden, carrying default spreads for both itself and the country. The currency choice made in step two will hold, with the riskfree rate in both the cost of equity and debt being the long-term default free rate in that currency (and not always the government bond rate). Step 5: Ensure that your cash flows are currency consistent      The currency choice made in step 2 determines not only the discount rates that you will be using but also the expected cash flows, with expected inflation driving both inputs. Thus, if you analyze a Turkish project in lira, where the expected inflation rate is 20%, you should expect to see costs of equity and capital that exceed 25%, but you should also see growth rates in the cash flows to be inflated the same expected inflation. If you assess the same project in Euros, where the expected inflation is 2%, you should expect to see much lower discount rates, high county risk notwithstanding, but the expected growth in cash flows will also be muted, because of the low inflation.     There is nothing in this process that is original or path-breaking, but it does yield a systematic and consistent process for estimating discount rates, the D in DCF. It works for me, because I am a pragmatist, with a valuation mission to complete, but you should feel free to adapt and modify it to meet your concerns.  YouTube Video Paper Country Risk Determinants: Determinants, Measures and Implications - The 2025 Edition Datasets Equity Risk Premiums, by country - July 2025 Country Risk Links EIU Democracy Index Global Peace Index (Exposure to Violence) Corruption Index International Property Rights Index Moody's Sovereign Ratings Political Risk Services (PRS) Country Risk Scores Spreadsheets Implied Equity Risk Premium for S&P 500 on July 1, 2025

Escape newsletter inbox chaos and algorithmic surveillance by building your own enshittification-proof newspaper from the writers you already read