Have you ever watched a stock get punished for a threat that never fully materializes, only to see the smartest money in the room sit quietly through the entire episode? That pattern keeps showing up in markets, and it feels almost primal. Our brains evolved to spot danger in the grass, yet they also invent tigers that turn out to be nothing more than wind. Investors do the same thing every time a new technology narrative hits. Lately the narrative has been artificial intelligence swallowing software and data businesses whole. One name that flinched hard earlier this year is now pressing against the top of a long sideways range, and a certain legendary holding company never sold a single share.
Why This Long-Held Berkshire Position Still Matters Today
The company in question is Moodys. For more than two decades it has sat inside the Berkshire Hathaway portfolio as a core position. The latest available filings show roughly 24.7 million shares, a stake valued near eleven billion dollars, with the share count left unchanged through the recent volatility. That kind of patience is rare. Most market participants react first and analyze later. Berkshire simply held the line.
I find that quiet confidence more interesting than any short-term price action. When a firm with that track record refuses to sell into an AI-driven sell-off, it forces you to look past the noise and examine the underlying business. What follows is a closer look at why the data Moodys controls may actually become more valuable, not less, in an AI-dominated world, and why the current technical setup is worth watching for patient investors.
The Evolutionary Bias That Drives Market Reactions
Humans are prediction machines. Dark clouds once meant rain. Certain berries meant illness. Rustling grass might have been a predator. Missing a real threat could end a bloodline. Seeing a false one cost only a jolt of adrenaline. Evolution therefore tuned us to over-detect patterns. Markets amplify that tendency. When a new threat appears, capital flees first and asks questions later. Sometimes the tiger is real. Sometimes the grass was just moving in the wind.
Earlier this year the AI narrative hit data providers and ratings agencies hard. Moodys dropped more than nineteen percent by mid-February and eventually touched a low near four hundred two dollars. The stock then spent most of the year locked in a wide trading range. It has since recovered and is now testing the upper boundary of that range near the five-hundred-fifteen level. The January high around five hundred forty-seven dollars remains the key level to clear for a more decisive move higher.
In my view the sell-off was classic pattern recognition gone slightly wrong. Investors saw software and information services under pressure and assumed every related name would suffer equally. Moodys is different. Its core product is not generic software that large language models can easily replicate. Much of its data cannot be scraped from public sources. That distinction matters more than the market initially appreciated.
How Proprietary Data Becomes More Valuable With AI
Large language models are only as good as the information they can access. Public data is already being digested at scale. Proprietary, high-quality datasets that sit behind paywalls or regulatory barriers become scarce resources. Moodys sits squarely in that category. Management has noted that customers are already feeding Moodys research directly into their own AI platforms. Usage of certain integration protocols has more than quadrupled in a short period, and daily content consumption by analysts has risen sharply.
Think of the underlying data as the oil that keeps the AI engine running. Without reliable fuel the most sophisticated model still stalls. Credit ratings, risk analytics, insurance catastrophe models, and compliance datasets are not easily reinvented by a general-purpose algorithm. They require decades of structured history, regulatory relationships, and continuous updating. That is the real moat.
AI is amplifying the value of the trusted data insights rather than replacing them.
That perspective from the company itself feels right. The firms that own the clean, structured, hard-to-replicate information should see demand rise as more institutions build internal AI tools. Moodys Analytics already generates the majority of its revenue on a recurring basis, with annual recurring revenue in the multi-billion-dollar range and growth still solid even during softer issuance periods.
Breaking Down the Two Core Business Segments
Moodys operates through two primary segments that complement each other. The first is the traditional ratings business. It rates corporate and government debt and generates revenue when issuance volumes are healthy. In the most recent quarter rated debt issuance ran near one point seven trillion dollars, delivering over a billion dollars in revenue for that segment alone.
The second segment, Analytics, focuses on credit risk tools, insurance models, and know-your-customer compliance data sold primarily to banks and insurers. This part of the business carries a very high percentage of recurring revenue. That predictability is attractive in any market environment, but it becomes especially valuable when traditional issuance slows.
Recent results showed overall revenue growth of roughly one percent and adjusted earnings per share growth of about six percent. Those numbers came in slightly below consensus expectations, largely because high-yield and leveraged loan issuance pulled back. Yet investment-grade issuance rose solidly, and AI-related issuance activity ran multiples higher than the prior year. First-time rating mandates also jumped meaningfully, marking the strongest quarterly total in years. Management kept full-year guidance intact, which suggests internal confidence remains steady.
Perhaps the most interesting aspect is how the two segments reinforce each other. Ratings activity generates proprietary insight that feeds the analytics products. Analytics customers then become natural users of the ratings franchise. That feedback loop is difficult for pure-play software firms to copy.
Comparing the Setup to a Similar Data Provider
A related data and index provider experienced a similar AI-related drawdown earlier in the year and has since recovered to fresh all-time highs. Moodys has lagged that recovery. Year-to-date performance sits near flat while the peer is up meaningfully. The difference creates an interesting relative-value discussion. One stock is already confirming a new uptrend. The other is still building a base and testing resistance.
I tend to prefer joining confirmed trends when possible, yet there is also merit in identifying setups before the breakout becomes obvious. Moodys currently sits in the latter category. The stock has spent months consolidating after the February low. Multiple attempts to push higher in the summer stalled near the five-hundred-twenty area. A brief poke above that zone earlier this month failed to hold, but the subsequent pullback has been orderly. Momentum readings cooled without price collapsing, which often leaves room for another attempt at the highs.
Traders watching the chart often focus on the five-hundred-forty-seven level as the clear breakout point. A sustained close above that January peak would change the technical picture. Until then the range remains intact. Risk levels are relatively clean. The late-October low near four hundred seventy-five aligns closely with the rising fifty-day moving average. A close back below that zone would signal that the post-earnings advance has lost momentum. Longer-term investors sometimes look further down to the earnings-day low near four hundred fifty-five, which also sits close to the two-hundred-day moving average.
Why Berkshire Never Flinched During the Sell-Off
The most telling signal may not be the chart at all. Berkshire Hathaway has owned the shares since the year two thousand. Through the entire AI scare the position size stayed constant. No selling. No adding. Just holding. That behavior says something about how the firm views the durability of the business model.
In my experience the best long-term holdings often look boring during periods of narrative-driven volatility. They do not need constant reinvention. They simply keep delivering a product that institutions cannot easily replace. Credit ratings remain essential for capital markets. Risk analytics and compliance tools remain essential for regulated financial firms. AI may change how those tools are delivered and consumed, but it does not eliminate the need for the underlying data.
The fact that customers are already integrating Moodys content into their own AI systems reinforces that view. Rather than being disrupted, the company appears to be becoming a preferred data source for the new generation of tools. That shift does not show up overnight in quarterly numbers, yet it can compound over years.
Practical Considerations for Different Types of Investors
Not every investor approaches a name like this the same way. Short-term traders may wait for a confirmed breakout above the prior high before committing capital. They can define risk tightly around the recent swing low and the fifty-day average. Position sizing stays modest until the range resolves higher.
Intermediate-term investors might begin building a position inside the current range, accepting that the stock could still retest lower levels. The earnings-day low provides a logical invalidation point for the recent advance. If that level holds on a weekly basis, the base remains constructive.
Long-term holders already comfortable with the business model often ignore the short-term technical noise entirely. They focus on the recurring revenue characteristics of the analytics segment, the durability of the ratings franchise, and the expanding role of proprietary data in an AI world. For them the current price simply represents another point on a multi-decade ownership timeline.
- Short-term approach: wait for a clean close above the prior high and use the late-October low as a stop reference
- Intermediate approach: scale in inside the range while watching the earnings-day low for confirmation of support
- Long-term approach: treat the business quality and data moat as the primary decision factors rather than weekly chart patterns
None of these approaches is inherently superior. They simply match different time horizons and risk tolerances. What matters is aligning the method with the actual thesis. If the thesis rests on the long-term value of the data franchise, then short-term underperformance relative to a peer becomes less important.
Key Risks That Still Deserve Attention
No investment thesis is complete without acknowledging the risks. Issuance volumes remain cyclical. When high-yield markets freeze or leveraged loan activity dries up, the ratings segment feels the pressure quickly. Regulatory scrutiny of the ratings industry never fully disappears. Competition in analytics is real, even if the moat around certain datasets remains wide.
There is also the possibility that the current range resolves lower instead of higher. Markets can stay range-bound longer than expected, or they can break in the direction that frustrates the most participants. A sustained move back below the key moving averages would force a reassessment of the near-term technical picture.
AI itself carries both opportunity and uncertainty. While current evidence suggests the data becomes more valuable, future developments could still surprise. New models might eventually extract more insight from public sources than currently seems possible. That risk exists, though it appears lower for specialized credit and risk datasets than for general software applications.
I’ve found that the most durable positions tend to survive these risks because the underlying customer need does not vanish. Banks still need to manage credit risk. Insurers still need catastrophe models. Capital markets still need independent ratings. The delivery mechanism may evolve. The core requirement persists.
Putting the Current Setup Into Context
At roughly five hundred fifteen dollars the stock trades about six percent below its fifty-two-week high. Momentum has cooled from overbought readings without price giving back large ground. The fifty-day and two-hundred-day moving averages sit comfortably below current levels. Volume patterns around the recent earnings report showed buyers stepping in when sellers tried to push the stock lower. Those are constructive details inside an otherwise sideways market.
The bigger picture remains the multi-year ownership by one of the most disciplined capital allocators in history. That ownership has survived multiple market cycles, regulatory changes, and technology shifts. The latest AI episode is simply the newest test. So far the position has passed it by remaining unchanged.
Whether the stock finally breaks out of its year-long range in the coming weeks or months is uncertain. What feels more certain is that the business continues to occupy a useful place in the financial ecosystem. Proprietary data that cannot be easily replicated tends to retain pricing power over time. Recurring revenue streams provide a cushion when transactional volumes slow. And a long-term shareholder base that refuses to panic creates a different kind of stability.
Final Thoughts on Pattern Recognition and Patience
Markets will keep producing new narratives. Some will prove accurate. Many will exaggerate the threat. Our evolutionary wiring will continue to make us flinch at the first rustle in the grass. The investors who occasionally pause and examine whether the tiger is real or imagined often end up holding the names that compound quietly for decades.
Moodys currently sits at an interesting intersection of technical setup and fundamental durability. The stock has digested a sharp AI-driven decline and spent months rebuilding. The underlying data franchise appears more relevant, not less, as institutions adopt new tools. And the longest-standing major shareholder has shown no desire to exit.
That combination does not guarantee immediate upside. Setups can fail. Ranges can persist. Yet for investors willing to look past the latest narrative and focus on the actual product customers pay for, the name remains worth understanding. Sometimes the best opportunities are the ones that never needed to reinvent themselves in the first place. They simply kept doing the work that capital markets and risk managers require, year after year, while the rest of the market chased the next shiny story.
In the end the real question is not whether AI will change how financial data is consumed. It already is. The more useful question is which firms own the data that the new systems will need most. On that score Moodys continues to look well positioned, just as it has for the better part of two decades inside one of the most carefully constructed portfolios in the market.
