Moody’s: How Ratings Create Value
In 2012, Heineken NV decided to start borrowing in capital markets. The company had operated for nearly 150 years without a formal credit rating, funding itself through bank loans and retained earnings. When Moody’s finally assigned it an investment-grade assessment, something peculiar happened: Heineken’s cost of debt dropped by thirty to fifty basis points immediately. The rating itself, an alphanumeric symbol, a theoretical opinion about future repayment probability, was worth real money. Yet Moody’s charged only seven to eight basis points for this service. The asymmetry is the entire business. Moody’s does not sell information; it sells permission to speak the language of global finance. For eight basis points, a borrower buys the right to say “I am A2” instead of “trust me,” and in doing so, saves forty. It is a tollbooth masquerading as an opinion shop, and it generates two and a half billion dollars in operating income on roughly one billion dollars of tangible capital, yielding returns on tangible assets that exceed two hundred percent. The economics are so favorable that they seem to describe a monopoly, yet three competitors share the market. The mystery is not why Moody’s is profitable, but why it has survived for more than a century as a private gatekeeper to public capital, and why regulators, who have tried repeatedly to destroy its business model, have always failed.
The Grammar of Global Debt
The most durable advantage Moody’s possesses is not its data, though the data is formidable. It is not its brand, though the brand is century-old. It is that Moody’s ratings function as a protocol: grammatical standard for describing creditworthiness that has achieved the status of infrastructure. When a treasurer issues a bond, they do not merely want a rating; they need a Moody’s rating specifically, because the institutional buyers on the other side of the trade have built their mandates, their compliance systems, and their liability shields around the particular taxonomy that Moody’s and its two major competitors provide. An “Aaa” or a “Baa1” is not simply an assessment; it is a coordinate in a shared linguistic system that allows trillions of dollars to move without translation friction.
This protocol generates exponential switching costs. To abandon Moody’s for a competitor would require not merely that one issuer learn a new rating scale, but that every investor, every index provider, every pension fund consultant, and every regulator simultaneously agree to interpret a new vocabulary. The coordination problem is insurmountable because the penalty for non-adoption is immediate and quantifiable: bonds lacking ratings from at least two of the “big three” trade at a thirty to fifty basis point penalty, if they trade at all. Unlike switching from English to Mandarin, which is merely difficult, switching from Moody’s to an alternative is economically irrational unless everyone switches at once. No individual market participant can justify bearing the transition cost for the collective benefit. Consequently, the market has remained an oligopoly of three for over a century, with Moody’s and S&P Global controlling roughly forty percent each, and Fitch holding fifteen. Despite post-2008 regulatory reforms that expanded the number of Nationally Recognized Statistical Rating Organizations from three to more than ten, no competitor has gained meaningful share. The language perpetuates itself.
Beneath this protocol lies a data moat that widens with each passing year. Moody’s possesses more than a century of accumulated default histories, recovery rates, and financial trajectories that cannot be replicated by backtesting or algorithmic inference. A competitor can build a model; it cannot create 1920s railroad bond defaults or 1990s Asian sovereign crises to train it. This historical archive underpins both the ratings business, providing credibility to the symbols, and the analytics division, which monetizes the data through subscriptions. The combination creates a reinforcing loop: the protocol generates the cash flows that fund the data acquisition, and the data deepens the protocol’s authority. The result is a business that requires virtually no tangible capital, no factories, no inventory, no significant property or equipment, yet extracts a levy on seventy-three trillion dollars of outstanding debt.
The Eight Basis Point Asymmetry
The unit economics of a rating transaction reveal the true nature of Moody’s power. The company charges seven to eight basis points on new issuance, plus annual surveillance fees, which constitute a trivial fraction of the total cost of issuing debt. A typical investment-grade bond issuance incurs total costs of seventy-five to one hundred basis points in underwriting, legal, and administrative fees. Moody’s slice is less than ten percent of that total, yet it delivers value, through interest rate savings, that amounts to four or five times its fee. This value-to-price gap is the source of the company’s pricing power. Because the cost of the rating is invisible relative to the value it creates, and because the alternative is a thirty to fifty basis point penalty for remaining unrated, Moody’s has been able to raise prices three to four percent annually for decades without demand destruction. Issuers grumble, but they pay.
The ratings business consequently generates operating margins exceeding fifty percent. On roughly three billion dollars of revenue, the division produces one and a half billion in operating income. The analytics business, by contrast, generates twenty to thirty percent margins on a similar revenue base, reflecting the reality that data subscriptions compete in a marketplace with specialized software vendors and information providers. The contrast is stark: the ratings division is a tollbooth on a bridge that cannot be circumvented; the analytics division is a competitive software business. Together they produce normalized operating income of two and a half billion dollars on tangible capital of roughly one billion, yielding a return on tangible assets of two hundred and fifty percent. The cash conversion is nearly one hundred percent; there are no receivables cycles, no working capital needs, no capital expenditures to maintain the physical plant. The business is a pure conversion of reputation into cash.
This economic engine depends on a specific alignment of incentives that is often misunderstood. Investors, not issuers, ultimately consume the ratings, yet issuers pay the fees. This principal-agent misalignment is not a bug to be eliminated but a feature that sustains the oligopoly. Because investors, often agents managing other people’s money, demand ratings from established brands to shield themselves from liability, they enforce the usage of Moody’s symbols without bearing the cost. Meanwhile, issuers pay willingly because the alternative is exclusion from institutional capital pools. The system survives because both sides are captive: investors cannot abandon the standard without exposing themselves to fiduciary risk, and issuers cannot access the market without purchasing the standard’s seal. The eight basis point fee is the equilibrium point where both captivity and value extraction coexist.
The Regulatory Sword
Yet this edifice rests on a structural constraint that cuts both ways. Moody’s exists as a private toll collector only because regulators have deputized it. Securities and Exchange Commission rules mandate that institutional investors hold NRSRO-rated securities to satisfy capital requirements; Basel banking accords embed these ratings into global financial regulation. Without this regulatory entrenchment, the thirty to fifty basis point penalty for unrated bonds might not exist, and Moody’s linguistic monopoly would dissolve into a competitive market for credit opinions. The company is, in essence, a government-licensed tax collector.
The fragility lies in what the government giveth, the government can rescind, r restructure. Following the 2008 financial crisis, when Moody’s downgraded ninety-one percent of single-family CDOs and thirty-seven percent of residential mortgage-backed securities, revealing that its previous “AAA” ratings had been catastrophically wrong, regulators responded by expanding NRSRO status to more than ten competitors and mandating transparency in rating methodologies. The assumption was that competition would discipline quality. It did not. Market share remained unchanged because the protocol effect proved stronger than regulatory intent. Investors continued to demand the big three; issuers continued to pay them.
But survival is not the same as safety. The regulatory forbearance that allows Moody’s to operate an issuer-pays model, where the rated entity funds the rater, exists on a contingent basis. The model survived 2008 not because it was vindicated, but because the alternative, building a new regulatory infrastructure, was administratively complex and politically costly. The next crisis may not offer the same reprieve. If a future ratings failure triggers a systemic cascade, regulators could mandate a subscription model where investors pay for ratings rather than issuers, or they could establish a government-provided rating service for certain asset classes. Either change would collapse the tollbooth. Unsolicited ratings, where agencies rate bonds without being paid by the issuer, would become viable if the regulatory requirement for issuer-paid ratings disappeared, allowing low-cost or free competitors to gain traction. The eight basis point fee persists only because regulators continue to tolerate a conflict of interest that would be illegal in most other domains of financial intermediation.
The Credibility Tightrope
This brings us to the central tension of the business model. Moody’s must maintain sufficient credibility to preserve the value of its linguistic protocol, yet it is paid by the very entities that desire favorable scores. The incentive for ratings inflation is structural and constant. In the years preceding 2008, this incentive yielded predictably distorted results: structured finance products received investment-grade ratings that implied negligible default risk, only to be mass-downgraded to junk status when the housing market collapsed. The fact that Moody’s market share remained intact after this failure suggests that the protocol effect is stronger than the accuracy effect. Investors continued to use Moody’s ratings not because they trusted them, but because they could not function without them.
Nevertheless, the equilibrium is delicate. If ratings inflation becomes too obvious, if, for example, a highly rated sovereign or corporate issuer defaults without warning, the credibility loss could trigger regulatory intervention that overrides the protocol’s network effects. Conversely, if Moody’s becomes too rigorous in its assessments, issuers will defect to competitors or push for regulatory acceptance of alternative standards. The company must perform a continuous balancing act, producing ratings that are inflated enough to keep issuers paying, but credible enough to keep investors using the language. This is not a moral failure but an economic necessity. The moment the ratings become either too accurate (driving away issuers) or too lenient (driving away investors), the tollbooth crumbles.
The post-2008 period has seen Moody’s navigate this tightrope by diversifying away from the pure ratings business, not by fixing its conflict of interest, but by rendering it less consequential to the corporate income statement.
Buying Stability at the Cost of Simplicity
Between 2017 and 2021, Moody’s spent more than six billion dollars acquiring analytics businesses: Bureau van Dijk, a corporate data provider, for $3.5 billion; RMS, a risk modeling firm for property and casualty markets, for $2 billion; and a KYC/AML compliance suite for $700 million. The strategic logic is cyclical smoothing. The ratings business is transactional; revenues rise and fall with debt issuance volumes, which correlate with interest rate environments and economic optimism. The analytics business is subscription-based, with ninety-four percent recurring revenue and retention rates exceeding ninety percent. By building a data and software division that now represents more than half of total revenue, management seeks to transform a cyclical toll collector into a steady-state information utility.
This strategy treats organizational capability as an economic variable. The ratings business is simple: hire analysts, maintain methodologies, extract fees. The analytics business is complex: it requires software engineering talent, product management, customer success functions, and continuous integration of disparate datasets. Each acquisition introduces new technical debt, cultural friction, and competitive exposure. Bureau van Dijk competes with Bloomberg and Refinitiv; RMS competes with specialized catastrophe modelers; the KYC suite competes with identity verification startups. These are arenas where fifty percent margins are impossible and where technological obsolescence is a real threat.
The risk is that in pursuing earnings stability, Moody’s dilutes the economic characteristics that made it valuable in the first place. The analytics division generates twenty to thirty percent margins, dmirable for a software business, but a drag on the portfolio compared to the ratings division’s fifty percent. If the acquisitions fail to integrate, or if specialized competitors disrupt individual verticals, Moody’s will have traded its high-return simplicity for a mediocre conglomerate. The capital allocation decisions of the past five years represent a bet that organizational complexity can be managed without sacrificing the tollbooth’s economics. It is a bet on managerial competence in a business that previously required none.
The Assumption We Dare Not Test
All of this, the protocol effect, the regulatory license, the pricing power, the analytics diversification, rests on a single fragile assumption: that regulators will never succeed in breaking the issuer-pays model. The entire two-hundred-fifty percent return on tangible assets depends on the continuation of a specific regulatory truce. Governments tolerate Moody’s because the cost of replacing it appears higher than the cost of monitoring it, and because no crisis since 2008 has been severe enough to force structural overhaul. But the assumption is contingent, not permanent.
If a future financial crisis involves a cascade of defaults among currently highly-rated entities, sovereigns, for instance, or systemically important financial institutions, the political will to dismantle the issuer-pays oligopoly may prove irresistible. Regulators could mandate that investors subscribe to ratings directly, eliminating the toll on issuers and opening the market to low-cost alternatives. They could establish a public ratings utility for certain debt classes. They could require that all ratings be unsolicited and publicly available, funded by regulatory fees rather than issuer payments. Any of these changes would destroy the network effect, because the switching cost barrier would vanish once the regulatory mandate for issuer-paid opinions disappeared.
Moody’s has survived for more than a century by positioning itself as the grammar of debt markets: language too costly to unlearn. But grammar requires consent. The company assumes that global capital markets will continue to speak Moody’s for the indefinite future, not because it is accurate, not because it is just, but because the cost of learning a new language exceeds the pain of paying the toll. This assumption has held through wars, depressions, and the 2008 financial crisis. Whether it survives the next one remains the only question that matters for the economics of the business.
The analytics acquisitions reveal a profound transformation in the nature of the enterprise. When Moody’s purchased Bureau van Dijk in 2017 for $3.5 billion, more than three times the company’s tangible capital base at the time, it was not merely diversifying revenue streams. It was attempting to convert a linguistic monopoly into a data utility, trading the ephemeral authority of credit opinions for the concrete recurring revenue of software subscriptions. Bureau van Dijk’s Orbis database contains information on over four hundred million private companies, a vast repository of corporate structures and beneficial ownership data that complements Moody’s century of public debt ratings. The strategic logic holds that by owning the underlying data layer of corporate identity, Moody’s can embed itself deeper into the workflows of compliance officers, procurement departments, and risk managers who need to know not just how likely a company is to default, but who actually owns it, where it operates, and whom it pays.
Yet this logic assumes that Moody’s can operate a software business with the same facility it operates a ratings franchise. The two require different organizational capabilities. The ratings business is a priesthood: it recruits academically trained credit analysts, socializes them into methodological orthodoxy, and empowers them to issue opinions that carry the force of market infrastructure. Success depends on maintaining the dignity of the temple, nsuring that the process appears rigorous, impartial, and above the fray of commerce. The analytics business, by contrast, is a bazaar. It requires product managers who iterate on user interfaces, sales teams who negotiate annual subscription renewals, and engineers who maintain API uptime. The cultural tension between these two modes is not merely aesthetic; it is economic. When Moody’s acquired RMS, the catastrophe modeling firm, for $2 billion in 2021, it inherited a workforce of climate scientists and actuaries who speak the language of insurance risk rather than credit risk. Integrating these teams into the Moody’s corporate structure without diluting their expertise, or alienating the credit analysts who generate the high-margin ratings revenue, requires managerial attention that was previously unnecessary. The tollbooth ran itself; the software company must be managed.
The risk extends beyond culture to competitive position. In the ratings business, Moody’s faces no serious competitive threat because the protocol effect protects it. In analytics, it faces Bloomberg, Refinitiv, specialized climate risk startups, and increasingly, artificial intelligence systems that can parse corporate filings without human intervention. The analytics division’s twenty to thirty percent margins are not protected by regulatory moats or linguistic lock-in; they are subject to the normal pressures of software economics, where features commoditize and prices erode. If Moody’s finds itself unable to cross-sell analytics products to its ratings customers, if the sales teams cannot convert the captive issuer relationships into software subscriptions, the company will be left holding $6 billion of assets that generate returns inferior to the capital they consume. The smoothing of cyclicality will have come at the cost of permanent impairment to the return on capital.
This organizational shift occurs against a backdrop of regulatory scrutiny that has grown more sophisticated since 2008. The initial post-crisis reforms, the expansion of NRSRO status to new competitors, the mandating of methodological transparency, the prohibition of gifts from issuers to analysts, were designed to increase competition and reduce conflicts of interest. They failed because they misunderstood the nature of Moody’s power. Regulators treated ratings as a market for opinions, subject to supply and demand like any other information product. They failed to recognize that Moody’s operates as a protocol layer, similar to the SWIFT system for payments or the TCP/IP standards for internet communication. Protocols do not compete on quality; they compete on adoption. A technically superior rating methodology is worthless if institutional investors cannot use it to satisfy their compliance requirements.
But regulators are learning. The European Union’s Credit Rating Agencies Regulation, as amended in 2019, began to chip away at the issuer-pays model by requiring rotation, orcing issuers to switch agencies periodically to prevent cozy relationships. More threatening is the recurring legislative proposal, periodically revived in the United States Senate, to establish a government-created platform where ratings would be assigned by random rotation among qualified agencies, with investors rather than issuers funding the system through transaction fees. Such a mechanism would not need to ban Moody’s to destroy its economics; it would simply need to make the eight basis point toll optional rather than mandatory. If issuers could obtain a regulatory-compliant rating from a rotating pool of competitors at lower cost, the protocol effect would fracture. Investors might still prefer Moody’s symbols for private benchmarking, but the regulatory mandate, the source of the thirty to fifty basis point penalty for unrated bonds, would dissolve.
The bear case, then, is not that Moody’s makes a catastrophic analytical error, though that is possible. It is that a future crisis creates the political conditions for structural separation between the regulatory function of ratings and the commercial function of credit assessment. If a major sovereign default occurs, if, for example, a G7 nation currently rated Aaa experiences a disorderly debt restructuring, the retrospective hunt for culpability will focus on the rating agencies that maintained investment-grade assessments until the moment of collapse. The political logic of such a crisis would demand not merely fines or procedural reforms, but the elimination of the perceived conflict of interest that allows private entities to profit from public regulatory functions.
Consider the mechanics of how such an intervention might unfold. The Securities and Exchange Commission could promulgate a rule requiring that all NRSRO ratings used for regulatory capital purposes be funded by a collective utility rather than by individual issuers. The Federal Reserve could announce that bank capital calculations would accept only ratings from a newly created public credit assessment facility, or from agencies that operate on a subscription basis with investors. These changes would not require legislation; they could be implemented through the administrative state’s existing authority over financial stability. Moody’s would retain its data and its brand, but it would lose the tollbooth. The eight basis point fee would become unsustainable because competitors would offer unsolicited ratings for free, funded by advertising or investor subscriptions, and issuers would have no regulatory incentive to pay Moody’s premium.
Even absent regulatory revolution, the analytics strategy carries hidden fragilities that compound over time. Moody’s has effectively borrowed against the cyclical stability of the ratings business to fund the acquisition of lower-margin software companies. The debt incurred for these purchases, while serviceable during normal interest rate environments, creates fixed obligations that persist even when debt issuance volumes collapse. In a severe recession where corporate bond issuance drops by forty percent for two consecutive years, the ratings business would see its revenues fall proportionally, while the analytics business, despite its recurring revenue streams, would face subscription cancellations as banks and insurers cut costs. The diversified revenue model might prove less resilient than the pure-play tollbooth it replaced. The company would find itself owning complex, difficult-to-manage software assets that no longer provide the earnings smoothing they were purchased to ensure, while the core ratings franchise has atrophied from lack of investment.
The historical data moat, meanwhile, is not as unassailable as it appears. While Moody’s possesses a century of default histories, the relevant timeframe for modern credit modeling is shortening. The financial structures of the 1920s, railroad bonds, municipal infrastructure debt, bear little resemblance to contemporary structured products, cryptocurrency lending protocols, or climate-transition finance. A competitor that begins accumulating data today on green bonds, decentralized finance collateral, or carbon-intensive corporate transitions might possess a more relevant dataset than Moody’s within a decade. The hundred-year archive becomes a liability if it saddles the company with legacy methodologies ill-suited to new asset classes. The Bureau van Dijk acquisition was an attempt to solve this by capturing private company data, but private companies do not default in ways that are always visible to credit analysts, and the integration of this data with traditional credit ratings remains technically and culturally incomplete.
The issuer-pays conflict, meanwhile, grows more acute as the market structure evolves. The rise of private credit, direct lending by non-bank institutions, has created a shadow debt market where ratings are optional. In this realm, Moody’s has no toll to collect; lenders conduct their own due diligence. If private credit continues to capture share from public bond markets, Moody’s addressable market shrinks regardless of regulatory stability. The analytics acquisitions were partly a hedge against this trend, f debt moves off public exchanges into private negotiations, Moody’s can still sell data about borrowers to the private lenders. But this requires competing in the due diligence support market against law firms, consultancies, and specialist data providers, a far more contested space than the regulated bond rating business.
The cultural implications of this strategic pivot are subtle but consequential. The credit analyst who joined Moody’s in 1995 did so to exercise judgment, to become an authority on the creditworthiness of sovereigns and corporations, to speak the language of finance with the gravitas of a standard-setter. The software product manager who joins today does so to optimize conversion rates and reduce churn in enterprise SaaS accounts. The first role commands deference; the second commands a salary. As the analytics division grows to represent not just half of revenue but potentially two-thirds of headcount, the center of gravity shifts. The company begins to optimize for retention metrics rather than methodological rigor. It invests in user interface design rather than analyst training. This is rational economic behavior, software scales, while judgment does not, but it erodes the very credibility that makes the ratings valuable. A rating is valuable because it represents a moment of expert judgment; if the organization that issues it becomes indistinguishable from a data vendor, the rating becomes merely an opinion among many, and the eight basis point toll becomes indefensible.
The fragility of the model, finally, lies in its dependence on the continued expansion of debt issuance itself. Moody’s assumes that global debt-to-GDP ratios will continue their century-long ascent, that corporations and governments will perpetually increase their leverage, requiring ever more ratings. But demographic shifts, deleveraging, or technological disintermediation of traditional banking could stall this growth. If the global economy enters a period of debt reduction, f aging populations in developed markets prioritize repayment over borrowing, if corporations fund themselves through retained earnings rather than bond issuance: he volume of new ratings collapses. The analytics business cannot compensate for a structural decline in the ratings business because the analytics business is largely sold to the same financial institutions that buy ratings. If those institutions shrink, both divisions suffer.
The tollbooth model works only so long as the traffic increases. Moody’s has survived for more than a century not because it has adapted brilliantly to changing conditions, but because the fundamental condition, that debt issuance grows, that regulators demand ratings, that investors need linguistic standards, has remained constant. The analytics acquisitions represent a bet that this constancy is ending, that the company must become something other than a toll collector to survive. But in becoming something else, it risks losing the economic characteristics that made it worth studying in the first place: the pure, minimal-capital extraction of rent from a mandatory network effect. The company is attempting to trade a perpetuity for a going concern, and the exchange may not favor the sellers.
The integration of Bureau van Dijk illustrates the friction inherent in this transformation. Orbis was not merely a database; it was a curated network of corporate relationships built through decades of manual verification, eneficial ownership trails that required local legal expertise in Luxembourg, shell-company registrations in Delaware, and opaque holding structures in the UAE. Maintaining this data requires a different operational cadence than rating bonds. A credit analyst works on quarterly cycles, publishing annual reviews when financial statements arrive. Orbis requires continuous monitoring, aily scraping of corporate registries, weekly updates to ownership graphs, real-time alerts when a director resigns in Singapore. The tempo is different, the talent is different, and the margin structure reflects this: Orbis commands subscription fees that seem substantial until one accounts for the army of data journalists required to verify that a holding company in the Cayman Islands has indeed changed its registered agent.
When Moody’s attempts to cross-sell Orbis to its ratings clients, it encounters a fundamental mismatch. The treasurer of a Fortune 500 company buys a rating to satisfy the board and the bond market; they buy Orbis data to investigate their supply chain or screen acquisition targets. The first purchase is mandatory, the second discretionary. During a recession, the treasurer will cancel the Orbis subscription before they default on the bond, but they cannot cancel the rating without triggering covenant violations. The analytics revenue is therefore less durable than the recurring revenue percentage suggests; it is merely stickier than transactional ratings, not sticky in the way that regulatory mandates are sticky. The attempt to blend these revenue streams in investor presentations, emphasizing the “68% recurring revenue” figure as if it represents a single quality of earnings, obscures this critical distinction. One revenue stream is a tax; the other is a subscription to a publication.
This operational complexity compounds at the leadership level. The executives who built the ratings franchise came from structured finance, from sovereign analysis, from the world of credit committees and methodological committees. The executives now running the analytics divisions come from software, from data aggregation, from the competitive intelligence industry. They speak different languages, iterally. A ratings managing director discusses “default probability” and “loss given default”; an analytics vice president discusses “user engagement” and “data completeness scores.” When capital allocation decisions are made in the same room, the two factions evaluate opportunities through incompatible frameworks. The ratings side sees the analytics acquisitions as dilutive distractions; the analytics side sees the ratings business as a stagnant, regulated utility. Both are correct, yet the corporate center must pretend they are synergistic, must force cross-selling targets that feel unnatural to both salesforces, and must maintain a unified culture that is, in reality, a negotiated truce between two distinct professional tribes.
The danger is not that one side wins, but that the truce collapses. If the analytics acquisitions fail to deliver the promised cyclical smoothing, the ratings organization will have lost years of focus, years during which it could have fortified its methodological defenses against the regulatory threats that never sleep. Meanwhile, the analytics businesses, forced to serve the corporate mandate of “integration,” may lose the entrepreneurial agility that made them acquisition targets in the first place. Moody’s risks becoming not a diversified powerhouse, but a compromised hybrid: too slow to compete with fintech data providers, too distracted to defend its tollbooth against regulatory dissolution. The eight basis point model required only that the world continue to borrow; the new model requires that Moody’s execute a complex organizational transformation while simultaneously managing the politics of its own obsolescence. That is a heavier burden than the balance sheet suggests.