穆迪 (MCO) 2026 Q3 法說會逐字稿

Good day, everyone, and welcome to the Moody's Corporation third-quarter 2026 earnings call. At this time, I would like to inform you that this conference is being recorded. (Operator Instructions) This call is scheduled to last approximately one hour.

各位大家好，歡迎參加穆迪公司（Moody's Corporation）2026 年第三季財報電話會議。此刻我想提醒各位，本次會議將被錄音。(接線員指示) 本次電話會議預計約一小時。

I will now turn the call over to Shivani Kak, Head of Investor Relations. Please go ahead.

現在我將把電話交給投資人關係主管 Shivani Kak。請開始。

Thank you, and thank you for joining us today. I'm Shivani Kak, Head of Investor Relations at Moody's. This morning, we reported our third-quarter 2026 results. The press release and today's presentation are posted at ir.moodys.com. We'll reference non-GAAP or adjusted measures. Please see the tables in our earnings release for reconciliations to US GAAP.

謝謝，也感謝各位今天加入我們。我是穆迪投資人關係主管 Shivani Kak。今天早上，我們已公布 2026 年第三季業績。新聞稿與今天的簡報已發布於 ir.moodys.com。我們將引用非 GAAP 或調整後指標。請參閱我們財報新聞稿中的表格，以了解與美國 GAAP 的調節。

Today's remarks may include forward-looking statements under the Private Securities Litigation Reform Act of 1995. Please see the Safe Harbor language in our earnings release and the risk factors and MD&A in our most recent Form 10-K and other SEC filings available on our website and the SEC's website. These factors could cause actual results to differ materially from those expressed or implied. Members of the media may be listening in a listen-only basis.

今天的發言可能包含依據 1995 年《私人證券訴訟改革法》（Private Securities Litigation Reform Act of 1995）所定義的前瞻性陳述。請參閱我們財報新聞稿中的「安全港」聲明，以及我們最新的 Form 10-K 與其他可於本公司網站及美國證券交易委員會（SEC）網站查閱之申報文件中的風險因素與管理階層討論與分析（MD&A）。這些因素可能導致實際結果與所表達或暗示者出現重大差異。媒體成員可能以僅收聽方式在線收聽。

And with that, I'll turn it over to Rob.

接下來，我把時間交給 Rob。

Thanks, Shivani, and hello, everybody. Thanks for joining us today. Moody's delivered another quarter of strong profitability and earnings growth with disciplined execution and healthy cash generation amidst an increasingly complex operating environment.

謝謝，Shivani，各位大家好。感謝各位今天加入我們。在營運環境日益複雜之際，穆迪憑藉嚴謹的執行力與健康的現金創造能力，再度交出強勁的獲利能力與盈餘成長的一季。

I want to start with a few key takeaways. So first, Moody's topped $2 billion in revenue for the third straight quarter, and we built on a record quarter from last year. Overall rated issuance declined 2%, yet we delivered adjusted operating margin of 52.6% and adjusted diluted EPS growth of 6%. And I think this really illustrates the resilience of the Moody's franchise and our ability to perform across a range of market environments even as market conditions became more uneven during the quarter.

我想先從幾個重點結論談起。首先，穆迪連續第三季營收突破 20 億美元，並在去年創紀錄的一季基礎上再接再厲。整體評級發行量下滑 2%，但我們仍達成 52.6% 的調整後營業利益率，以及 6% 的調整後稀釋每股盈餘（EPS）成長。我認為這充分展現了穆迪業務版圖的韌性，以及即使本季市場狀況更為不均衡，我們仍能在各種市場環境下維持表現的能力。

Now second, the investment agenda is intact. And I acknowledge that issuance can be volatile from quarter-to-quarter. And amidst a mixed backdrop, MIS delivered its second strongest third quarter on record. And financing activity tied to AI infrastructure remained very strong in the quarter, while on the other hand, leverage finance activity moderated against a strong prior year comp and more selective issuance environment. And MIS delivered an adjusted operating margin of 64.4%. I think that's proof of our disciplined cost management, preserving the operating leverage in the business.

第二，投資議程仍然完整推進。我也承認，發行量可能在季度之間波動。在多空交雜的背景下，MIS 交出其史上第二強的第三季表現。本季與 AI 基礎設施相關的融資活動依然非常強勁；另一方面，槓桿融資活動則在去年同期高基期以及更為審慎的發行環境下有所降溫。MIS 的調整後營業利益率達 64.4%。我認為這證明了我們嚴謹的成本管理，並保留了業務的營運槓桿。

In MA, ARR grew approximately 8%. That was in line with our high single-digit outlook driven by healthy retention and strong recurring revenue growth, and we continue to see some encouraging momentum across several growth areas. That includes lending, insurance, KYC private credit and AI-enabled solutions. And we expanded MA's adjusted operating margin by 100 basis points versus last year to 35.3%, and we're on track to achieve our full year target.

在 MA 方面，ARR 約成長 8%。這與我們「高個位數」的展望一致，主要由健康的續約率與強勁的經常性收入成長所帶動；我們也持續在多個成長領域看到令人鼓舞的動能。其中包括放貸、保險、KYC、私募信貸以及 AI 賦能解決方案。我們將 MA 的調整後營業利益率較去年提升 100 個基點至 35.3%，並正按計畫達成全年目標。

Finally, the quarter reinforced the measured approach that we've been taking to guidance throughout the year. And following a very strong second quarter, you'll all recall, we chose not to raise guidance in July, recognizing challenging issuance comps and some question marks about the global macro environment in the second half of the year.

最後，本季再次印證了我們今年以來對財測所採取的審慎作法。各位都記得，在第二季表現非常強勁之後，我們在 7 月選擇不調升財測，因為我們意識到發行量比較基期具挑戰性，且對下半年全球總體經濟環境仍存在一些不確定性。

And as the third quarter unfolded, strength in some markets was offset by softness in others. And to that end, we are reaffirming our full year guidance metrics, and we'll talk about the puts and takes to that guidance later in the call.

隨著第三季展開，部分市場的強勁表現被其他市場的疲弱所抵銷。因此，我們重申全年財測指標；稍後在電話會議中，我們將說明影響該財測的利多與利空因素。

So let me move to ratings. While issuance activity moderated during the quarter, my broader view of the funding deep current has not changed. And for several years now, we've discussed this very significant capital investment requirements across the global economy. That's across AI infrastructure and data centers, power generation and transmission, renewable energy, and carbon transition, defense and security, transportation infrastructure, the world continues to require substantial amounts of investment capital.

接下來我談評級業務。雖然本季發行活動有所降溫，但我對資金需求「深層趨勢」的整體看法並未改變。過去幾年，我們一直在討論全球經濟中極為龐大的資本投資需求。這涵蓋 AI 基礎設施與資料中心、發電與輸配電、再生能源與碳轉型、國防與安全、交通運輸基礎設施；世界仍持續需要大量投資資本。

And those long-term funding requirements are creating durable opportunities for debt financing and capital formation. We continue to see evidence of that in many parts of the market, particularly in sectors tied to AI infrastructure and power investment where issuance increased approximately fourfold from the third quarter of last year.

而這些長期資金需求正在為債務融資與資本形成創造持久的機會。我們在市場的許多領域持續看到相關證據，尤其是在與 AI 基礎設施及電力投資相關的產業，本季發行量較去年第三季約增加至四倍。

Meanwhile, certain markets, most notably leveraged finance, were softer than the very strong issuance levels of a year ago and overall issuance reflected that dynamic. But the broader forces driving global capital needs remain in place.

同時，某些市場——最明顯的是槓桿融資——相較於一年前非常強勁的發行水準更為疲弱，整體發行量也反映了這種動態。但推動全球資本需求的更廣泛力量仍然存在。

That said, our results reflect much more than just AI-related transaction. And I'm going to spotlight a couple of examples. First, in private credit, the number of rated deals grew by more than 40% year-over-year and market innovation continues. Good example of that is a deal called AMAP VI. That's a broadly diversified, actively managed multi-asset credit vehicle from Apollo. And that deal allocates across five evergreen strategies, and we rated -- we rated billions of dollars of debt out of this vehicle, showcasing our analytical depth and complex private credit-related structures.

話雖如此，我們的業績反映的遠不僅僅是 AI 相關交易。我想聚焦幾個例子。第一，在私募信貸方面，評級交易筆數年增超過 40%，市場創新仍在持續。一個很好的例子是名為 AMAP VI 的交易。這是 Apollo 推出的廣泛分散、主動管理的多資產信貸工具。該交易在五個常青（evergreen）策略間進行配置，而我們對此工具所發行的數十億美元債務進行了評級，展現我們在分析深度以及複雜私募信貸結構方面的能力。

Second, in emerging markets, we're acquiring a minority stake in PhilRatings. That's our sixth domestic credit rating agency affiliate in Asia Pac. It gives us a leading position in the Philippines, where more than $100 billion of infrastructure investment is planned by 2030. And Southeast Asia's domestic bonds outstanding are more than twice the size of cross-border holdings and growing, and this investment enhances our very strong position in domestic debt markets around the world.

第二，在新興市場方面，我們將收購 PhilRatings 的少數股權。這將成為我們在亞太地區的第六家本土信用評級機構聯盟夥伴。此舉使我們在菲律賓取得領先地位；該國至 2030 年規劃的基礎建設投資超過 1,000 億美元。此外，東南亞的在外流通本土債券規模是跨境持有規模的兩倍以上且仍在成長；這項投資也強化了我們在全球各地本土債務市場的既有強勢地位。

Now as markets evolve, our analytical frameworks are evolving with them, and our new cross-sector digital infrastructure methodological framework integrates corporate and project finance analytical approaches to assess an asset class characterized by distinct business models, financing structures and credit profiles. And this is critical to how Moody's ensures analytical rigor and consistency across AI infrastructure buildouts.

隨著市場演進，我們的分析框架也同步演進；我們新的跨產業數位基礎設施方法論框架，整合公司金融與專案融資的分析方法，以評估一類具有獨特商業模式、融資結構與信用特徵的資產類別。這對穆迪如何在 AI 基礎設施建置中確保分析嚴謹性與一致性至關重要。

So taken together, these deep currents are translating into strong growth in new rating relationships. First-time mandates increased 40% in the third quarter. That's the highest quarterly total in five years. And this is notable because it really demonstrates the continued importance, relevance and value of a Moody's rating in a rapidly changing world.

綜合來看，這些深層趨勢正在轉化為新評級關係的強勁成長。第三季首次委任（first-time mandates）增加 40%。這是五年來最高的單季總量。這一點值得注意，因為它確實顯示在快速變動的世界中，穆迪評級的重要性、相關性與價值仍在延續。

Now this leads me to our annual refinancing study, which highlights the -- really the sheer magnitude of debt coming due over the next several years. And that outstanding debt must ultimately be refinanced and that creates substantial and highly visible foundation for future issuance activity. And while markets will inevitably experience periods of volatility, this refinancing cycle remains an important structural driver of long-term transaction volumes.

接著談到我們的年度再融資研究，該研究凸顯了——確實是——未來數年到期債務的龐大規模。而這些未償債務最終必須再融資，這為未來發行活動提供了可觀且高度可見的基礎。雖然市場不可避免會經歷波動期，但這一輪再融資循環仍是長期交易量的重要結構性驅動因素。

Over the next four years, refinancing needs across the US and EMEA topped $5 trillion for the second straight year. And while there is a modest decline from last year, it reflects a combination of FX and maturity extensions primarily in leveraged finance. However, investment-grade maturities grew about 5%, with the US leading the way at approximately 11% growth and pull-forward activity generally stayed within historical norms.

在接下來四年內，美國與歐洲、中東及非洲（EMEA）地區的再融資需求連續第二年超過5兆美元。雖然較去年小幅下降，但主要反映匯率（FX）因素與到期日展延的綜合影響，且主要集中在槓桿融資領域。然而，投資級到期規模約成長5%，其中美國領先、約成長11%，而提前發行（pull-forward）活動整體仍維持在歷史常態範圍內。

So recent issuance also carries tenors longer than four years. So the strength of the last several years will show up beyond the standard four year window. In fact, the US five year view provides additional useful context and refinancing needs grew approximately 4% to about $3.5 trillion, and the former peak maturity year in 2028, that's now 2031. So reach out to our Investor Relations team if you're interested in a copy of these reports.

因此，近期的發行也包含期限長於四年的券期。因此，過去幾年的強勁發行將在標準四年視窗之外顯現。事實上，從美國五年期視角可提供更多有用的背景：再融資需求約成長4%至約3.5兆美元，而原先到期高峰年在2028年，現在已移至2031年。因此，若您有興趣取得這些報告的副本，請聯繫我們的投資人關係團隊。

And as you've likely heard me say, but I want to say it again, analytical rigor, robust methodologies, and ratings quality are and will remain our North Star. Moody's ratings are the benchmark standard for pricing credit risk and we hold the line ratings integrity even on marquee transactions. And we've seen several recent transactions priced closer to our assessment when our rating diverges from the market or isn't present, reinforcing the credibility of our methodologies and analytical approach with investors. So for Moody's ratings, I think the through line is clear. Markets will ebb and flow, but the multiyear funding needs are intact and Moody's is the agency of choice for issuers and investors alike.

而且正如你們可能聽我說過的，我想再強調一次：嚴謹的分析、健全的方法論與評等品質，是且將持續是我們的北極星。穆迪評等是信用風險定價的基準標準，即使在指標性交易上，我們也堅守評等的完整性。我們也看到近期有數筆交易在我們的評等與市場看法出現分歧或市場未提供評等時，其定價更貼近我們的評估，進一步強化投資人對我們方法論與分析取徑的信任。因此，就穆迪評等而言，我認為主軸非常清楚。

All right. So now let me turn to Moody's Analytics. And we delivered another solid quarter, driven by continued momentum across several of our core growth areas that includes lending, insurance, KYC, private markets and AI-enabled solutions. And we were pleased to be named number one in the Chartis RiskTech100 for the fifth consecutive year with a record 18 category wins across credit, insurance data and AI.

好的。接下來我轉到穆迪分析（Moody's Analytics）。我們又交出一個穩健的季度表現，動能來自多個核心成長領域的持續推進，包括放貸、保險、KYC、私募市場以及AI賦能解決方案。我們也很高興連續第五年在Chartis RiskTech100中名列第一，並在信用、保險數據與AI等領域創下紀錄、共拿下18個類別冠軍。

And while I get that awards are not the objective, we do view this as valuable external validation of the strength of our capabilities. And I want to spotlight a few of our fastest-growing areas. I'm going to start with insurance where ARR grew 10% this past quarter.

雖然我理解獎項並非目標，但我們確實將其視為對我們能力強度的寶貴外部驗證。我想特別聚焦幾個成長最快的領域。我先從保險開始，本季ARR成長10%。

So Cristina and I recently attended Rendez-vous in Europe. And our conversations increasingly centered on how Moody's can help address the industry's biggest challenges, including reducing the protection gap, bringing more alternative capital into the market and delivering better real-time insights and productivity to the industry. And this led to some great discussions about our new offerings in casualty, in CAT bonds and our AI-enabled intelligent risk platform and more.

Cristina和我最近參加了在歐洲舉行的Rendez-vous。我們的對話愈來愈聚焦在穆迪如何協助解決產業最大的挑戰，包括縮小保障缺口、引入更多替代資本進入市場，以及為產業提供更好的即時洞察與生產力。這也促成了許多很好的討論，涵蓋我們在意外險、巨災債（CAT bonds）、AI賦能的智慧風險平台等新產品與更多內容。

The fastest growth areas in insurance this past quarter were cat risk and casualty solutions. In CAT modeling, we continue to drive adoption of our IRP our high-definition model suite and our risk data lake, all of which offer the ability to deliver upsell with existing customers as well as bring in new customers to the platform.

本季保險領域成長最快的是巨災風險與意外險解決方案。在巨災建模方面，我們持續推動IRP（Intelligent Risk Platform，智慧風險平台）、高解析度模型套件，以及風險資料湖的採用；這些能力不僅能對既有客戶進行加值銷售（upsell），也能為平台帶來新客戶。

And a great example of this is with a top three US auto and property insurer. They wanted to unify their hurricane, earthquake, and wildfire risk workflows into a single environment. And so they adopted our IRP, which integrates third-party and in-house models along with our high-definition models, including wildfire allowing us to more than double the ARR with this customer.

一個很好的例子是與美國前三大的汽車與財產保險公司之一合作。他們希望將颶風、地震與野火風險的工作流程整合到單一環境中。因此他們採用了我們的IRP，該平台可整合第三方與自建模型，並結合我們的高解析度模型（包含野火模型），使我們在該客戶的ARR增加到原來的兩倍以上。

And perhaps more importantly, it represents further adoption by one of the leading primary insurers paving the way for greater penetration into the primary market beyond reinsurers and brokers. Another great win in the quarter was with the top-tier global reinsurance broker who signed a new five year agreement that included adoption of our full global model suite. That drove nearly 14% ARR growth with this customer and demonstrated the opportunities for substantial growth with existing customers through broader adoption of our platform capabilities and our models.

更重要的是，這代表一家領先的原保險公司進一步採用我們的方案，為我們在再保險公司與經紀商之外、向原保險市場更深滲透鋪路。本季另一項重大斬獲是與頂級全球再保險經紀商簽署新的五年合約，其中包含採用我們完整的全球模型套件。這使該客戶的ARR成長接近14%，並展現透過更廣泛採用我們的平台能力與模型，對既有客戶仍有可觀的成長空間。

In banking, growth was driven by the lending suite, which grew ARR by 12% and remains one of our most compelling growth opportunities. One example was a major new bank logo in Mexico, rounding out our footprint across the country's top five banks. The bank is deploying Moody's lending suite as it automates the entire end-to-end credit process with our origination to monitoring platform as the foundation and giving us a flagship reference account for further Latin American expansion.

在銀行領域，成長主要由放貸套件帶動；其ARR成長12%，仍是我們最具吸引力的成長機會之一。其中一個例子是我們在墨西哥新增一家大型銀行客戶，讓我們在該國前五大銀行的布局更為完整。該銀行正在部署穆迪放貸套件，並以我們從授信發起到監控（origination-to-monitoring）的平台作為基礎，自動化整個端到端的信用流程，同時也為我們進一步拓展拉丁美洲提供旗艦級的參考客戶。

Now turning to KYC and compliance. We continue to see good momentum, particularly in the public sector. In this past quarter, we expanded our relationship with a fast-growing supply chain analytics company signing a five year reseller agreement that will extend Orbis penetration into multiple government agencies.

接著談KYC與合規。我們持續看到良好動能，尤其是在公部門。本季我們擴大與一家快速成長的供應鏈分析公司的合作，簽署一份五年期轉售商協議，將把Orbis的滲透延伸至多個政府機構。

Another example: a new federal agency in Germany is deploying our data and compliance solutions to power investigative and intelligence workflows. We also renewed an agreement with the European Commission to support the next generation of ARACHNE, its risk scoring and data mining platform designed to help protect EU funds against fraud and financial risk.

另一個例子：德國一家新的聯邦機構正在部署我們的數據與合規解決方案，以支援調查與情報工作流程。我們也與歐盟執委會續約，以支援下一代ARACHNE——其風險評分與資料探勘平台，旨在協助保護歐盟資金免於詐欺與金融風險。

And collectively, these examples show how we're monetizing our data and analytics into use cases like national security, defense, fraud and beyond. And we see further opportunities across the public sector, both in the US and Europe.

整體而言，這些案例顯示我們正將數據與分析能力變現到國家安全、國防、反詐欺等多元使用情境，並延伸至更多領域。我們也在美國與歐洲的公部門看到更多機會。

Now turning to AI. Our AI-enabled screening capabilities enabled us to displace an incumbent with a large and sophisticated bank. Screening is a prime space for agentic tools coupled with Moody's decision grade data. For example, a top five US commercial bank adopted our compliant solution to reduce false positives and improve screening effectiveness and efficiency.

接著談AI。我們的AI賦能篩檢能力，讓我們在一家大型且成熟的銀行中取代了既有供應商。篩檢是代理式（agentic）工具結合穆迪決策級（decision-grade）數據的理想應用場景。例如，美國前五大商業銀行之一採用了我們的合規解決方案，以降低誤判（false positives），並提升篩檢的有效性與效率。

Now zooming out customers are adopting AI across every way they access Moody's through MCP-enabled interfaces and APIs into their own AI platforms, through AI built into our applications and through partner platforms where they already work.

從更宏觀的角度看，客戶正透過各種方式採用AI來使用穆迪：包括透過MCP啟用的介面與API接入其自有AI平台、在我們的應用程式中內建AI，以及透過他們既有的合作夥伴平台在原工作環境中使用。

And some of our largest and most sophisticated customers are bringing Moody's data and research directly into their own AI platforms through MCPs, and adoption is accelerating. So it is early days. But in just the last three months, users have more than quadrupled and usage is up more than tenfold. And particularly important, analysts using the MCP are reaching 3 times more content on a working day while still accessing it in our own platforms. Those are our research MCPs.

我們一些規模最大、最成熟的客戶，正透過MCP將穆迪的數據與研究內容直接導入其自有AI平台，且採用速度正在加快。因此，這仍處於早期階段。但僅在過去三個月內，使用者數量已超過四倍成長，使用量也提升逾十倍。尤其重要的是，使用MCP的分析師在工作日可觸及的內容量增加到原來的3倍，同時仍可在我們自有平台中存取。這些是我們的研究MCP。

In addition to MCPs though, we're making AI available in our own applications. So in insurance, for example, we're embedding AI into underwriting exposure management and portfolio workflows in the IRP. Our risk data refinery turns unstructured submissions into analytics-ready data with around 95% first pass accuracy. Customers can cut slip interpretation time by as much as 80% to 90% and statement of values preparation by up to 75%, and that means less time wrangling data and faster underwriting decisions.

除了MCP之外，我們也在自有應用程式中提供AI能力。例如在保險領域，我們正將AI嵌入IRP中的核保曝險管理與投資組合工作流程。我們的風險資料精煉（risk data refinery）可將非結構化的投保資料轉換為可供分析的資料，首次處理正確率約95%。客戶可將承保條款（slip）解讀時間縮短多達80%至90%，並將保額明細表（statement of values）準備時間最多縮短75%；這意味著更少時間用於整理資料，並能更快做出核保決策。

And building on this, we're also investing in the demand in growing markets and deepening our bench of strategic partnerships. And I want to just give you two examples from the quarter. So first, in private credit, our Allvue partnership pairs Moody's forward-looking models with Allvue's loan-level private credit data. And we believe that this gives us the best view of credit risk in the private markets. And in turn, that gives lenders and investors an independent view of emerging borrowers stress, helping support underwriting, portfolio monitoring valuation, and risk oversight.

在此基礎上，我們也正在加大對成長型市場需求的投資，並深化我們的策略合作夥伴陣容。我想就本季提供兩個例子。首先，在私募信貸方面，我們與 Allvue 的合作將穆迪（Moody's）具前瞻性的模型與 Allvue 的逐筆貸款層級私募信貸資料相結合。我們相信，這讓我們能對私募市場的信用風險取得最佳視角。進而，這也讓放款人與投資人能以獨立角度觀察新興借款人的壓力情況，協助承作審核、投資組合監控與估值，以及風險監督。

And second, in casualty insurance, Moody's and Lloyd's of London are partnering to combine our data and models with Lloyd's historical market data to pinpoint which signals predict casualty and financial lines risk where exposure is concentrated and when underwriting cycles are turning.

第二，在意外險方面，穆迪與倫敦勞合社（Lloyd's of London）合作，將我們的資料與模型與勞合社的歷史市場資料結合，以精準辨識哪些訊號可預測意外險與金融線（financial lines）的風險、曝險集中於何處，以及承保循環何時轉向。

And this is really the same playbook that we've run in other insurance lines, but now applied to a large underpenetrated casualty market and with one of the most important players in the global insurance industry.

這其實是我們在其他保險險種中一貫採用的相同作法，只是現在把它應用到一個滲透率偏低、規模龐大的意外險市場，並且與全球保險產業中最重要的參與者之一合作。

So before I turn it over to Noemie, Cristina just marked her first 100 days leading MA, and she's brought a real sense of urgency and ambition. She spent a lot of time on the road with customers. She's challenging us to think differently and moving quickly on the foundational changes that I believe can unlock the next phase of growth.

在我把時間交給 Noemie 之前，Cristina 剛完成她領導 MA 的前 100 天，她帶來了強烈的緊迫感與企圖心。她花了很多時間在外拜訪客戶。她挑戰我們以不同方式思考，並快速推動我相信能解鎖下一階段成長的基礎性變革。

So big picture, even as the operating environment we face keeps evolving. The examples that I shared reinforce our confidence in the opportunities ahead. Substantial financing needs underpin ratings, demand for our decision grade intelligence underpins Moody's Analytics and AI is amplifying the value of the trusted data, insight, and analytical context that's at the heart of everything we do.

從大方向來看，即使我們所面對的營運環境持續演變。我分享的這些例子都強化了我們對未來機會的信心。龐大的融資需求支撐評級業務；對我們「決策級情報」（decision grade intelligence）的需求支撐穆迪分析（Moody's Analytics）；而 AI 正在放大我們可信賴的資料、洞見與分析脈絡的價值——這些正是我們一切工作的核心。

So Noemie, over to you.

Noemie，交給你。

Thank you, Rob, and hello, everyone. Let me start with Moody's Analytics. Moody's Analytics delivered 3% reported revenue growth versus last year, and that's 8% on an organic constant currency basis. Recurring organic constant currency revenue grew 9%, and that now represents about 99% of total MA revenue. Transactional revenue declined 76% year-over-year, primarily reflecting the Learning Solutions divestiture.

謝謝你，Rob，各位大家好。我先從穆迪分析（Moody's Analytics）談起。穆迪分析本季報告口徑營收較去年成長 3%，若以有機、固定匯率基礎則成長 8%。經常性（Recurring）有機固定匯率營收成長 9%，目前約占 MA 總營收的 99%。交易型（Transactional）營收年減 76%，主要反映我們處分 Learning Solutions 業務。

We ended the quarter with $3.7 billion in annualized recurring revenue, ARR and that's up approximately 8% year-over-year, and we are on track to meet our full year guidance. Retention was very strong at 95% on a trailing 12 month basis and that reflects the mission-critical nature of our solutions and the durability of our customer relationships, underpinned by trust and delivery of outcomes.

本季結束時，我們的年化經常性收入（ARR）為 37 億美元，年增約 8%，並且我們正按計畫達成全年財測指引。以過去 12 個月（trailing 12 month）計算的留存率（Retention）非常強勁，達 95%，這反映我們解決方案的關鍵任務屬性（mission-critical）以及客戶關係的韌性，而這些都建立在信任與成果交付之上。

Decision Solutions ARR grew 10% and that's very consistent with the prior quarter, and that represents about 44% of MA ARR. Within Decision Solutions, KYC ARR grew 12% and that's driven by healthy demand across KYC data use cases, as you just heard from Rob.

Decision Solutions 的 ARR 成長 10%，與前一季非常一致，約占 MA ARR 的 44%。在 Decision Solutions 內，KYC 的 ARR 成長 12%，這是由於 KYC 資料各類使用情境的需求健康成長所帶動，正如你剛才從 Rob 那裡聽到的。

Insurance ARR grew nearly 10%, driven by strong demand for catastrophe data and modeling solutions and particularly strong retention across the portfolio. Banking ARR grew 8% and with strategic growth areas such as lending, up 12% versus last year and leading performance. While more mature parts of the portfolio, we're growing at a more measured pace, which is consistent with our prior commentary.

保險業務 ARR 成長近 10%，主要由巨災資料與建模解決方案的強勁需求所帶動，且整體產品組合的留存率特別強。銀行業務 ARR 成長 8%，其中放款等策略成長領域較去年成長 12%，表現領先。至於產品組合中較成熟的部分，我們以較為穩健的步調成長，這也與我們先前的說法一致。

Research & Insights reported revenue growth of 3% reflects a tough Q3 comparison, as the prior year period benefited from revenue recognition on large on-prem contract renewals. ARR grew about 6%, supported by Moody's OneView, which brings our research, data and AI-powered research assistant together in a single platform.

Research & Insights 報告口徑營收成長 3%，反映第三季比較基期偏高，因去年同期受惠於大型地端（on-prem）合約續約的收入認列。ARR 成長約 6%，主要由 Moody's OneView 所支撐；OneView 將我們的研究、資料與 AI 驅動的研究助理整合於單一平台。

CreditView customers moving to OneView are getting a richer, more integrated solution, and that added value is translating into an uplift at renewal. And OneView is also expanding our research to customers who haven't used it before.

CreditView 客戶轉移至 OneView 後，獲得更豐富、整合度更高的解決方案，而這項附加價值也在續約時轉化為更高的提升幅度（uplift）。同時，OneView 也正把我們的研究擴展到過去未使用研究服務的客戶。

In Q3, a large quantitative trading firm with more than $1 trillion in AUM, already a Moody's data customer converted a research trial into a multiyear agreement to access OneView research by API. It's a strong validation of our proof-of-value approach and it shows the opportunity to bring our research into quantitative investment workflows.

在第三季，一家管理資產規模（AUM）超過 1 兆美元的大型量化交易公司，原本已是穆迪資料客戶，將研究試用轉換為一份多年期協議，透過 API 存取 OneView 研究內容。這強力驗證了我們的價值驗證（proof-of-value）方法，也顯示將我們的研究導入量化投資工作流程的機會。

Data & Information ARR grew 7%, reflecting demand for ratings data feeds and the data estate, including Orbis. We also signed a new deal with SAP to launch Moody's Supplier Intelligence for Ariba, embedding our data directly into supplier onboarding, assessment and monitoring within SAP. This win displaces a competitor and establishes Moody's as a strategic intelligence partner to SAP with potential to expand across its global customer base.

Data & Information 的 ARR 成長 7%，反映對評級資料供給（ratings data feeds）與資料資產（data estate，包括 Orbis）的需求。我們也與 SAP 簽署新合作，推出適用於 Ariba 的「穆迪供應商情報」（Moody's Supplier Intelligence for Ariba），將我們的資料直接嵌入 SAP 內部的供應商導入、評估與監控流程。這項勝出取代了一家競爭對手，並使穆迪成為 SAP 的策略情報合作夥伴，未來有機會擴展至其全球客戶基礎。

Along the same lines, we signed a multiyear private market data distribution partnership with one of the world's largest asset managers embedding Moody's Company intelligence, including Orbis into its next-generation private market platform. This expands the reach of our data across the private market ecosystem and create opportunities for future growth.

同樣地，我們也與全球最大資產管理公司之一簽署多年期私募市場資料分銷合作夥伴關係，將穆迪的公司情報（Moody's Company intelligence，包括 Orbis）嵌入其下一代私募市場平台。這擴大了我們資料在私募市場生態系中的觸及範圍，並創造未來成長機會。

And we also expanded our relationship with a global technology customer, deepening its use of Orbis for lead identification and arrangement. The relationship has grown 2.5 times in under two years, with Moody's data now embedded in its global sales workflow.

此外，我們也擴大了與一家全球科技客戶的合作關係，深化其在潛在客戶辨識與安排（lead identification and arrangement）方面對 Orbis 的使用。在不到兩年的時間裡，這段合作關係成長了 2.5 倍，穆迪資料如今已嵌入其全球銷售工作流程。

So across these examples, the pattern we are observing is very consistent. Customers are increasingly embedding Moody's data and analytics into their core workflows that strengthen our position in day-to-day decision-making and creates opportunities to deepen and expand those relationships over time.

綜合這些例子，我們觀察到的模式非常一致。客戶正日益將穆迪的資料與分析嵌入其核心工作流程，這強化了我們在日常決策中的地位，並隨時間推進創造深化與擴展合作關係的機會。

Turning to profitability. MA adjusted operating margin was 35.3% and that's up 100 basis points year-over-year. We're seeing the benefits of simplifying the portfolio, consolidating our technology platforms and beginning to capture efficiencies from AI, helping us manage head count growth while continuing to invest in priority areas like our data estate and technology. And importantly, we're doing both funding growth and expanding margins.

接著談獲利能力。MA 調整後營業利益率為 35.3%，年增 100 個基點。我們正看到簡化產品組合、整併技術平台，以及開始從 AI 擷取效率所帶來的效益，這有助於我們在持續投資資料資產與技術等優先領域的同時，管理人力成長。更重要的是，我們同時做到為成長提供資金並擴大利潤率。

Shifting to ratings. Rated issuance moderated in the summer as expected, but remained very strong by historical standards. At approximately $1.7 trillion, issuance was down 2% year-over-year against a very strong comparison while still representing the second highest third quarter level on record. Importantly, and as you heard from Rob, this headline figure masks considerable variation across asset classes, with strength in investment-grade infrastructure financing, offset by softer leveraged finance activity.

接著轉到評級業務。如預期，夏季的評級發行量（rated issuance）有所降溫，但以歷史標準來看仍然非常強勁。發行量約 1.7 兆美元，年減 2%；在比較基期非常強的情況下，仍創下史上第三季第二高的紀錄。重要的是，正如你從 Rob 那裡聽到的，這個總體數字掩蓋了不同資產類別之間的顯著差異：投資級基礎建設融資表現強勁，但槓桿融資活動較為疲弱，形成抵銷。

Markets remain largely constructive in the third quarter, with the issuers moving quickly when windows were open. Resilient investor demand, continued AI and hyperscale related financing and ongoing infrastructure needs supported issuance. Credit spreads remained relatively stable for most of the quarter before widening by approximately 10% towards the end of September amid rising macro and policy uncertainty. But even with that widening, spreads remain near historical tight.

第三季市場整體仍大致偏向建設性，發行人在窗口開啟時迅速把握機會。具韌性的投資人需求、持續的 AI 與超大規模（hyperscale）相關融資，以及持續的基礎建設需求，共同支撐了發行量。本季大部分時間信用利差相對穩定，但在 9 月底因總體與政策不確定性上升而擴大約 10%。即便如此，利差仍接近歷史低檔（tight）水準。

MIS revenue was $1.1 billion, down 1% in the third quarter -- against this tough compare in the third quarter last year. Importantly, the relationship between issuance and transaction revenue this quarter reflects the mix. Issuance declined 2%, while transaction revenue declined 5%. Third quarter issuance was driven by lower-yielding jumbo transactions and frequent issuers, while at the same time, higher yielding leverage finance issuers significantly pulled back their issuance activities.

MIS 營收為 11 億美元，第三季年減 1%——主要是因為去年第三季的比較基期偏高。重要的是，本季發行（issuance）與交易營收之間的關係反映了業務組合。發行量下滑 2%，而交易營收下滑 5%。第三季的發行主要由收益率較低的大額（jumbo）交易與高頻發行人所帶動；同時，收益率較高的槓桿融資發行人則明顯縮減其發行活動。

And you can see that clearly within Corporate Finance. High yield and bank loan revenue declined 40% and 27%, respectively, while investment-grade revenue increased 18% on 16% issuance growth. Investment-grade issuance reached a third quarter record, supported by AI infrastructure and hyperscaler financing. And year-to-date, jumbo corporate issuance has already surpassed the full year 2025 level.

而你可以在企業融資（Corporate Finance）中清楚看到這一點。高收益與銀行貸款營收分別下滑 40% 與 27%，而投資級營收在發行量成長 16% 的帶動下增加 18%。投資級發行量創下第三季新高，受 AI 基礎設施與超大規模雲端業者（hyperscaler）融資支撐。而截至目前（year-to-date），大額企業發行量已經超過 2025 全年水準。

There was a similar mix dynamics in the financial institutions group where issuance came predominantly from banking issuers on frequent issuer pricing contracts. PPIF grew across projects, infrastructure and US public finance, including data center-related activity. And structured finance benefited from ABS, RMBS and CMBS that was partially offset by lower CLO activity.

在金融機構組（financial institutions group）也出現類似的組合動態，發行主要來自銀行類發行人，且多為高頻發行人定價合約。PPIF 在專案、基礎設施與美國公共融資（US public finance）等領域皆有成長，包括與資料中心相關的活動。結構性融資受惠於 ABS、RMBS 與 CMBS，但部分被 CLO 活動較弱所抵銷。

Recurring revenue grew 8% on new mandates, value-based pricing and growth in monitored credits. First-time mandates support future recurring revenue, like what you see this quarter. And given our performance to date and visibility into our pipeline, we're raising our full year First Time Mandate outlook to a range of 900 to 950. MIS adjusted operating margin remained strong at 64.4%, down approximately 80 basis points year-over-year, but the margin reflects continued disciplined resource allocation and technology investment while preserving investment capacity to support future growth.

經常性營收在新委任案（new mandates）、價值導向定價（value-based pricing）以及受監控信貸（monitored credits）成長的帶動下成長 8%。首次委任案（First-time mandates）可支撐未來的經常性營收，就像你在本季所看到的。基於我們迄今的表現以及對管線的可視性，我們將全年首次委任案展望上調至 900 至 950 的區間。MIS 調整後營業利益率仍維持強勁，為 64.4%，年減約 80 個基點；但該利益率反映了我們在維持嚴謹資源配置與科技投資的同時，仍保留投資能力以支持未來成長。

Moving on to insurance outlook. Markets remain open and the pipeline is robust, supported by continued refinancing needs M&A activity and significant financing requirements around AI, data centers, and broader infrastructure. Against that backdrop, our overall issuance outlook remains unchanged at mid-single-digit percent growth with some shifts in the underlying mix.

接著談保險展望。市場仍然開放，且管線強勁，受持續的再融資需求、併購（M&A）活動，以及圍繞 AI、資料中心與更廣泛基礎設施的龐大融資需求所支撐。在此背景下，我們整體發行展望維持不變，仍為中個位數百分比成長，但底層組合將出現一些變化。

As you can see on slide 12, we're tempering our expectation for high yield and leveraged loans given recent weakness, while raising our expectation for structured finance, financial institutions, and PPIF. Our guidance reflects the range of potential outcomes from here, a sustained period of volatility, wider spreads or more constrained access for sub-investment-grade issuers could move us toward the lower end of our revenue and adjusted EPS ranges. Conversely, the pipeline remains healthy, and stronger M&A execution, continued AI and infrastructure financing or an improvement in leverage finance activity could provide additional upside.

如投影片 12 所示，鑑於近期疲弱，我們下調對高收益與槓桿貸款的預期，同時上調對結構性融資、金融機構以及 PPIF 的預期。我們的指引反映了從現在起可能出現的結果範圍：若波動持續一段時間、利差擴大，或非投資級發行人的市場進入更受限，可能使我們的營收與調整後 EPS 區間偏向較低端。相反地，管線仍然健康，若 M&A 執行更強、AI 與基礎設施融資持續，或槓桿融資活動改善，則可能帶來額外上行空間。

Turning to guidance. We are maintaining our full year outlook for both businesses. For MIS, we continue to expect full year revenue growth in the high single-digit percent range, supported by continued AI, data center and GPU-related financing, with issuance trends varying across lines of businesses.

接著談指引。我們維持兩大業務的全年展望不變。就 MIS 而言，我們仍預期全年營收成長落在高個位數百分比區間，受持續的 AI、資料中心與 GPU 相關融資支撐，但各業務線的發行趨勢將有所差異。

What that means for the fourth quarter is we expect issuance to decline in the mid-teens percent versus the prior year, broadly consistent with the second half cadence we discussed last quarter, following very strong year-to-date activity and reflecting more tempered expectations for leveraged finance.

這對第四季的意涵是：我們預期發行量相較去年同期將下滑約十幾個百分點（mid-teens percent），大致符合我們上季所討論的下半年節奏；此前年初至今活動非常強勁，而本預期也反映我們對槓桿融資較為保守的看法。

We expect MIS revenue to decline sequentially versus the third quarter, very consistent with typical historical seasonality, while growing in the single-digit percent range versus the prior year and margins will follow a similar seasonal pattern as last year.

我們預期 MIS 營收相較第三季將呈現季減，這與典型的歷史季節性非常一致；但相較去年同期仍將維持個位數百分比成長，而利潤率也將呈現與去年相似的季節性型態。

For MA, we continue to expect ARR growth in the high single-digit percent range and adjusted operating margin of 34% to 35%. The fourth quarter remains the largest sales and renewal period of the year, and our outlook assumes continued execution against the pipeline.

就 MA 而言，我們仍預期 ARR 成長落在高個位數百分比區間，調整後營業利益率為 34% 至 35%。第四季仍是一年中最大的銷售與續約期間，而我們的展望假設將持續有效執行管線。

At the MCO level, we're maintaining our guidance for the full year with revenue growth in the high single-digit percent range, adjusted operating margin of 52% to 53% and adjusted diluted EPS of $16.50 to $17, which represents 12% growth at the midpoint.

在 MCO 層級，我們維持全年指引不變：營收成長為高個位數百分比區間、調整後營業利益率為 52% 至 53%，以及調整後稀釋 EPS 為 16.50 至 17 美元，以中位數計算代表成長 12%。

Third quarter adjusted expenses increased 2% year-over-year, reflecting disciplined cost management, thoughtful capital allocation as we continue to invest behind our growth priorities. And we continue to expect mid-single-digit percent expense growth for the full year. Our capital allocation priorities remain unchanged, investing to support growth, driving sustainable margin expansion, preserving balance sheet flexibility and returning excess capital to shareholders.

第三季調整後費用年增 2%，反映我們嚴謹的成本管理，以及在持續投資成長優先事項的同時，審慎進行資本配置。我們仍預期全年費用成長為中個位數百分比。我們的資本配置優先順序維持不變：投資以支持成長、推動可持續的利潤率擴張、維持資產負債表彈性，並將多餘資本回饋股東。

Year-to-date, we have repurchased approximately $2.6 billion worth of shares and continue to expect up to $3 billion of repurchases for the full year. Free cash flow was $2.2 billion year-to-date. That's up 22% year-over-year, and we remain on track to return more than 130% of free cash flow to shareholders this year.

年初至今，我們已回購約 26 億美元的股票，且仍預期全年回購金額最高可達 30 億美元。年初至今自由現金流為 22 億美元。年增 22%，我們仍按計畫在今年將超過 130% 的自由現金流回饋給股東。

Now stepping back, the third quarter demonstrates the strength and consistency of our model, durable recurring growth in analytics, strong profitability and ratings, disciplined investments behind our growth priorities and significant cash generation. That financial strength gives us the flexibility to continue investing in the business while returning meaningful capital to shareholders.

回頭來看，第三季展現了我們商業模式的強度與一致性：分析業務具韌性的經常性成長、強勁的獲利能力與評級業務、在成長優先事項上的紀律性投資，以及顯著的現金創造能力。這樣的財務實力讓我們在持續投資業務的同時，也能向股東回饋具意義的資本。

Thank you, and operator, we'll now take the questions.

謝謝，接線員，我們現在開始提問。

Thank you. (Operator Instructions) George Tong, Goldman Sachs.

謝謝。（接線員指示）George Tong，高盛。

You highlighted a more than 4 times increase in MCP users and a more than 10 times increase in AI usage over the last three months. Can you talk about where you're seeing the best evidence that this adoption is generating incremental revenue and incremental ARR rather than primarily supporting retention and engagement. And what needs to happen for AI to become a more meaningful contributor to MA's overall growth rate?

你提到過去三個月 MCP 使用者增加超過 4 倍，AI 使用量增加超過 10 倍。你能否談談你在哪些方面看到最有力的證據，顯示這種採用正在帶來增量營收與增量 ARR，而不只是主要用於支撐留存與互動。另外，要讓 AI 成為 MA 整體成長率更有意義的貢獻者，需要發生哪些事情？

Yes, George. Thanks for the question. I'm going to zoom out here and help us all think about the ways that AI is being monetized across the MA product suite because it's not really just one line item. And I know people want that. But it's increasingly embedded across everything that we do.

好的，George。謝謝你的問題。我先拉高一點視角，幫助大家思考 AI 在 MA 產品組合中是如何變現的，因為它其實不只是單一的一個項目。我知道大家會想要那樣的拆分。但 AI 正越來越深度地嵌入我們所做的一切。

And it shows up in customer adoption, product engagement, new sales, win loss rates, renewals, everything, right? In many cases, this stuff is table stakes. And I would say that we see customers consuming our AI capabilities in three ways. So it's through AI that's embedded in our own applications, it's through AI agents and workflow that we provide, and it's through direct access to Moody's intelligence via MCPs and APIs into the environments that either customers are building or platforms that they're working on.

它會體現在客戶採用、產品互動、新增銷售、贏單/輸單比率（win-loss rates）、續約等各方面，對吧？在很多情況下，這些已經是基本門檻（table stakes）。我會說，我們看到客戶以三種方式使用我們的 AI 能力。第一是嵌入在我們自有應用程式中的 AI；第二是我們提供的 AI 代理（agents）與工作流程（workflow）；第三是透過 MCP 與 API，直接在客戶正在建置的環境或其使用的平台中，存取穆迪（Moody's）的情報（intelligence）。

So on the MCP side, look, I think I've talked -- been talking with investors, we got to start by looking at adoption and usage. And over time, we will see more and more revenue. So the number of customers paying for MCP access has more than doubled in this last quarter. It's still off of a small base. So we've made a lot of this available, but not all of the customers are able to consume the content via MCP. So a lot of customers are working on being able to take MCPs into their environment.

就 MCP 這一端來說，你看，我想我一直在跟投資人談，我們得先從採用與使用情況來看。隨著時間推進，我們會看到越來越多的營收。在上一季，付費購買 MCP 存取權的客戶數已經增加一倍以上。但目前仍是從很小的基礎起步。因此我們已經提供了很多內容，但並非所有客戶都能透過 MCP 來使用內容；許多客戶正在努力把 MCP 導入他們的環境中以便使用。

I can tell you that the open opportunities are up very substantially since the end of the second quarter, and about half of our pipeline by value has been created since the end of the second quarter. So we're seeing this start to come into the pipeline. And a lot of that is newly created demand. So you got to understand, there are new buying motions. Many customers want to start with a proof of value before they're committing.

我可以告訴你，自第二季末以來，未結商機大幅增加，而按價值計算，我們約有一半的銷售管線是在第二季末之後新建立的。因此，我們正看到這些開始進入管線。而其中很大一部分是新創造的需求。所以你必須理解，現在出現了新的採購模式。許多客戶希望在承諾之前，先從價值驗證（proof of value）開始。

So I think we're going to see adoption and some of these revenue metrics progressively build. I mentioned a cohort of 1,000 trial research users. And again, this is really -- I think you want to look at the usage because we'll understand how we're going to price behind that. I talked about usage is growing more than 10 times in just three months. And we're seeing the users consume a lot more of our content, 3.5 times more content, 5 times more companies.

所以我認為，我們將會看到採用率以及其中一些營收指標逐步建立。我提到過一個由1,000名試用研究用戶組成的同質群（cohort）。而且再說一次，這其實——我認為你應該看使用量，因為我們會據此理解後續如何定價。我談到使用量在短短三個月內成長超過10倍。而我們也看到用戶消耗了更多我們的內容：內容量多了3.5倍、涵蓋公司數多了5倍。

But then customers are also adopting AI inside of our workflows. And I want to talk about insurance for a moment. So I gave some examples in my remarks about that risk data refinery, right? That's another application that sits on the IRP. It contributes to customers wanting to migrate to the IRP and contributes to new customers wanting to be on the IRP. That's where they can also access APIs, MCP interface. We have more than 700 APIs that allow customers to embed Moody's intelligence from the IRP directly into their workflows. So when you look at that 10% growth rate in insurance, you have to understand that the AI capabilities in the IRP are part and parcel of what is driving that 10% growth.

但接著，客戶也在我們的工作流程內採用AI。我想先談一下保險。我在發言中舉了一些關於風險資料精煉廠（risk data refinery）的例子，對吧？那是另一個建置在IRP上的應用。它促使客戶想要遷移到IRP，也促使新客戶希望使用IRP。那也是他們可以存取API、MCP介面的地方。我們有超過700個API，讓客戶能把來自IRP的穆迪（Moody’s）智慧直接嵌入他們的工作流程。所以當你看到保險業務10%的成長率時，你必須理解，IRP中的AI能力正是推動這10%成長的重要組成部分。

And let me shift to banking for a moment. So we have something called Banking Decision Intelligence. It's really like a connected brain that unifies our risk and finance capabilities across our application suite and creates almost like an auditable digital worker for banks, right? So this is sitting across our application stack in banking. We've already got three live customers. We've got a fourth that has done a completed proof of concept. We've got a number that are in trial.

接著我把話題轉到銀行業。我們有一個稱為「銀行決策智慧」（Banking Decision Intelligence）的產品。它很像一個互聯的大腦，把我們在整個應用套件中的風險與財務能力整合起來，並為銀行打造幾乎像可稽核的數位員工，對吧？所以這是部署在我們銀行業務的應用堆疊之上。我們已經有三個上線客戶。第四個客戶已完成概念驗證（proof of concept）。還有多個客戶正在試用中。

We're building pipeline. And we're hearing from customers, literally, they're saying to us, I had a team of analysts doing a certain kind of work, bringing information -- data and information across multiple disconnected applications that would take weeks and it's being done in minutes and hours.

我們正在建立銷售管線。而且我們從客戶那裡聽到的回饋是，真的，他們對我們說：我原本有一個分析師團隊在做某一類工作，把資料與資訊從多個彼此不相連的應用程式中彙整起來，原本要花好幾週，現在幾分鐘到幾小時就能完成。

So not surprisingly, there's a lot of interest from our banking customers, and we're building pipeline. In KYC, obviously, the Agentic opportunity is a very real one and going after both labor productivity and effectiveness. And off of a small base, our KYC agents are growing almost 3 times faster than the KYC business as a whole.

所以不意外地，我們的銀行客戶非常感興趣，而我們也在建立銷售管線。在KYC方面，顯然，代理式（Agentic）的機會非常真實，目標同時涵蓋勞動生產力與有效性。而且在一個相對較小的基礎上，我們的KYC代理（agents）成長速度幾乎是整體KYC業務的3倍。

Again, it's small but it's growing. So I guess, George, that was a long answer, but hopefully, it gives you a sense of the different ways that AI is coming into the numbers, and it's not always just going to be in one line that we can call out.

再說一次，規模還小，但正在成長。所以我想，George，這是一個很長的回答，但希望能讓你感受到AI以不同方式反映在數字上，而且不一定總是會集中在某一條我們可以特別點名的單一項目裡。

Andrew Steinerman, JPMorgan.

Andrew Steinerman，摩根大通（JPMorgan）。

Could you please review the puts and takes on Moody's assumed fourth quarter issuance, particularly talk about AI associated and MA associated financing? And also a quick point on fourth quarter typically being a briefer quarter for issuance seasonally.

你能否請回顧一下穆迪對第四季發行量假設中的利多與利空因素（puts and takes），特別談談與AI相關以及與併購（MA）相關的融資？另外也快速提一下，第四季在季節性上通常是發行較短的一季。

Andrew, let me just start with kind of the market environment. And I would say that right now, we have a very healthy pipeline going into November. Obviously, we've seen treasuries widen out. We've got concerns about sovereign deficits and a very heavy issuance calendar that we've had that I think has put pressure on yields. But issuers with funding needs are still coming to market.

Andrew，我先從市場環境談起。我會說，目前我們進入11月時有非常健康的銷售管線。顯然，我們看到美國公債利差擴大。我們也擔心主權赤字，以及非常密集的發行行事曆；我認為這些因素對殖利率造成了壓力。但有資金需求的發行人仍然在進入市場。

And I would say the more opportunistic issuers, and those are particularly in leveraged finance, have been a little more selective about -- in terms of timing, which is what you'd expect. And so the credit conditions, and we often see this, right, when there's a little bit more focus on risk, we see more differentiation in the investment grade and spec-grade issuers that come to market.

而我會說，較偏機會型的發行人——特別是在槓桿融資領域——在時點選擇上更為挑剔一些，這也符合預期。因此在信用環境方面，我們也常看到這種情況：當市場對風險的關注提高時，進入市場的投資級與投機級發行人之間會出現更明顯的分化。

But I'd say, first of all, the markets are open, they're constructive. And then I would say just in terms of kind of puts and takes, Andrew, for the quarter, so just to remind you, we're assuming a mid-teens year-over-year decline in issuance in the fourth quarter. And that's because in the fourth quarter of last year, we had some very strong activity from bank loan repricing’s, hyperscaler issuance, it's very significant last fourth quarter and banks.

但我想先說，市場是開放的，而且氛圍偏正向、具建設性。接著就本季的利多與利空因素而言，Andrew，提醒你一下，我們假設第四季發行量將年減約十幾個百分點（mid-teens）。原因是去年第四季有一些非常強勁的活動，包括銀行貸款重新定價（bank loan repricing）、超大規模雲端業者（hyperscaler）發行——去年第四季規模非常可觀——以及銀行相關發行。

And some of that was not revenue mix friendly. But we -- so we have assumed a decline, a mid-teens decline. But, and we've assumed that the hyperscalers have done most of their financing for the year. So if we see hyperscaler issuance in any size that's going to be upside. By the way, we still have a healthy pipeline of broader AI and data infrastructure that we think will come to the market.

其中有一部分在營收組合上並不友善。但我們——所以我們假設會下滑，約十幾個百分點的下滑。同時我們也假設，超大規模雲端業者今年大部分的融資已經完成。因此，如果我們看到任何具規模的超大規模雲端業者發行，那將會是上行因素。順帶一提，我們仍然看到更廣泛的AI與資料基礎設施領域有健康的管線，我們認為它們會進入市場。

And to your point about trading days, I think that's a really good point, Andrew, because there are a waning number of trading days left in the year. And so you've got issuers that are starting to think about prefunding 2027 needs on maybe dips in treasuries, getting ahead of a potentially higher rate environment and also thinking about getting it in front of or having a window where there's not as much hyperscaler debt financing in the market.

至於你提到的交易日，我認為這點非常好，Andrew，因為今年剩下的交易日正在減少。因此，發行人開始思考：在公債殖利率回落時，是否要為2027年的資金需求提前預先融資（prefunding），以便在可能更高利率的環境之前先行布局；同時也在思考要搶在市場上超大規模雲端業者債務融資較少的窗口之前完成發行。

Now on the downside, I think this is pretty obvious, but we've got plenty of headline risk. We've got two wars raging around the world. We've got the US midterms could be a wild card, by the way, we may see issuance get pulled in before inauguration in Congress. And a new Congress comes into session, so we're keeping an eye on that.

至於下行風險，我想這相當明顯：我們面臨很多頭條風險。全球有兩場戰爭正在進行。美國期中選舉也可能是變數；順帶一提，我們可能會看到發行在就職典禮之前、在國會——新一屆國會開議之前被提前拉動，所以我們也在密切關注。

So hopefully, that gives you a sense. I'm going to pause there because I'm sure we're going to get into the hyperscaler mix on issuance, but let me just pause there, okay?

希望這能讓你有個概念。我先在這裡停一下，因為我確定我們會進一步談到發行中的超大規模雲端業者組合，但我先暫停到這裡，好嗎？

Jason Haas, Wells Fargo.

Jason Haas，富國銀行（Wells Fargo）。

I wanted to ask about some recent high-profile fines for some KYC and anti-money laundering lapses at some large financial institutions that made some pretty big headlines. I’m curious to what extent that’s a potential positive driver for your KYC business going forward.

我想問一下，近期一些大型金融機構因KYC與反洗錢（anti-money laundering）疏失而遭到的高知名度罰款，登上了相當大的新聞版面。我很好奇，這在多大程度上可能成為你們KYC業務未來的正向驅動因素。

Yes, Jason, I think it's a driver for the entire market. And it's interesting because there was a little bit of a narrative that had begun to build in the market where some folks were saying, well, it's been a while since we've seen some of these sizable fines and we're making such huge investments into KYC and financial crime compliance. And some of these things we're seeing recently, I think are just a stark reminder for everybody about the critical importance of getting all of this right.

是的，Jason，我認為這是整個市場的驅動因素。而且很有意思，因為市場上先前開始出現一點敘事，有些人說，嗯，我們已經有一段時間沒看到這類規模可觀的罰款了，而我們也在 KYC 與金融犯罪合規上投入了如此巨大的資金。而我們最近看到的一些事情，我認為只是再次強烈提醒大家，把這一切做到位有多麼關鍵。

And it's -- and it also goes to, Jason, it's not just about efficiency, and you've heard me say this before. The banks say this to us all the time. I've got to be more efficient, and I've got to be more effective. You have to do both of those things.

而且——而且這也回到，Jason，這不只是效率的問題，你以前也聽我說過。銀行一直都這樣跟我們說。我必須更有效率，而且我必須更有效能。你必須兩者都做到。

And so from a KYC perspective, I mentioned that we're investing in our KYC agents. It's still early days there. I would also say that just in terms of our KYC business, and I know this on the minds of some, but we have been working to migrate some customers from our old Orbis Catalyst over to our Maxsight platform.

因此，從 KYC 的角度來看，我提到我們正在投資於我們的 KYC 代理（agents）。目前仍在早期階段。我也想說，就我們的 KYC 業務而言——我知道這也是一些人關心的——我們一直在把部分客戶從舊的 Orbis Catalyst 遷移到我們的 Maxsight 平台。

And so I think we've been a little bit behind in that migration. So we're redoubling our efforts to be able to do that and get more and more customers onto that new Maxsight platform that will offer us more and more monetization opportunities as well as agentic opportunities as well.

因此我認為我們在這項遷移上稍微落後了一些。所以我們正在加倍努力推進，讓越來越多客戶上到新的 Maxsight 平台；這也將為我們帶來更多變現機會，以及更多代理式（agentic）的機會。

Scott Wurtzel, Wolfe Research.

Scott Wurtzel，Wolfe Research。

I just wanted to talk about the banking subsegment within DS for a second. And You talked about maybe some more mature parts of banking growing at a more measured pace. Just wondering if you are thinking about any ways to sort of maybe accelerate that part of the business, whether it is on go-to-market or new product introduction. Any color around that would be great. Thanks.

我想先談一下 DS 裡面銀行子分部的情況。你提到銀行業中一些較成熟的部分可能會以較為溫和的速度成長。我想了解你是否在思考任何方式來加速這一塊業務，不論是在 go-to-market（市場推進）或新產品導入方面。如果能提供一些補充說明會很有幫助。謝謝。

Yes, Scott. Thanks for the question. And you've seen a few of the moves that we've made, cleaning up that banking portfolio, in particular, a few of the things that you've seen around both the learning business and the regulatory reporting business, we're in that banking segment. And that's all part of our drive to really focus in on what we think are the highest growth opportunities, and that is really around lending, which is obviously building on our core strength in credit and then risk and finance.

是的，Scott。謝謝你的問題。你也看到我們做了一些動作來整理那個銀行產品組合；特別是你看到的幾項調整，包含學習業務與監管申報（regulatory reporting）業務，這些都在銀行這個分部裡。這些都是我們推動聚焦於我們認為最高成長機會的一部分，而那主要是在放款（lending）領域——這顯然是建立在我們在信用（credit）方面的核心優勢之上，並延伸到風險與財務。

And so more and more, what we're hearing from companies is they want to be able to connect all of this. And that's why I mentioned, Scott, this decision intelligence layer that we're putting over the application suite that's allowing banks to get really unique new insights much, much faster than they have before. And it also goes to the value proposition of if you have one or two or three or four or five applications from Moody's. Now we're able to connect those applications and give banks unique intelligence that has either been very difficult to get or is actually unique intelligence that they just haven't gotten before.

因此，我們從企業端聽到越來越多的是，他們希望能把這一切串接起來。這也是我提到的原因，Scott：我們正在在應用套件之上建立一個決策智慧（decision intelligence）層，讓銀行能夠比以往更快取得非常獨特的新洞見。這也回到價值主張：如果你使用 Moody’s 的一個、兩個、三個、四個或五個應用。現在我們能把這些應用連結起來，並提供銀行過去很難取得、甚至是他們以前根本沒有得到過的獨特洞察。

So I think Cristina is very, very focused on thinking about how we're bundling our solutions, how we think about pricing and packaging to make this all easier for our customers. The last thing I would say, Scott, is while one of our fastest-growing offerings right now is our AI-enabled lending suite, we're also going to be working with what I'm going to call lighthouse customers. These are going to be early adopters who want to work with us, where we can deploy FDEs to these customers and actually start co-building and agentic solution.

所以我認為 Cristina 非常、非常專注在思考我們如何把解決方案打包（bundling），以及如何在定價與包裝（pricing and packaging）上設計，讓客戶更容易採用。最後我想說的是，Scott，雖然我們目前成長最快的產品之一是 AI 賦能的放款套件（AI-enabled lending suite），我們也將與我所稱的「燈塔客戶」（lighthouse customers）合作。這些會是願意與我們一起合作的早期採用者，我們可以把 FDEs 部署到這些客戶端，並實際開始共同打造一個代理式（agentic）解決方案。

So this may be a next-generation agentic lending workflow where we co-build with them. And we've got a number of customers that we're in discussions with to be those lighthouse customers.

因此，這可能會是一個下一代的代理式放款工作流程，我們會與他們共同開發。我們目前正與多家客戶洽談，希望他們成為這些燈塔客戶。

Craig Huber, Huber Research Partners.

Craig Huber，Huber Research Partners。

I want to ask you a question, Rob, about hyperscalers. It is a question that comes up a lot from investors. However you want to define that, but what is your estimate for the rated debt issuance for them in 2026 in aggregate? And what percentage of global debt issuance do you think that is going to be this year?

Rob，我想問你一個關於超大規模雲端業者（hyperscalers）的问题。這是投資人常常會問到的問題。不管你怎麼定義它們，你對於 2026 年它們合計的評級債務發行量（rated debt issuance）有什麼估計？你認為這會占今年全球債務發行量的百分之多少？

Then my follow-on question is, I know it is early here, but for 2027, what is your thoughts here on the total debt issuance that may come out of the hyperscalers in aggregate? I mean, I think -- do you have a sense of what you are hearing out there?

接著我的追問是，我知道現在談還太早，但對於 2027 年，你對超大規模雲端業者合計可能會有多少總債務發行量有什麼看法？我的意思是——你是否能從你在外面聽到的訊息中，掌握一些方向？

Yes. So Craig, thanks for the question. And Andrew, this will be responsive to the second part of your question as well. And I'm going to give you most of this answer, Craig, and you're going to be able to connect the dots here.

是的。所以 Craig，謝謝你的問題。Andrew，這也會回應你問題的第二部分。Craig，我會把大部分答案先給你，你也能把這些線索串起來。

So the good news is that we are capturing growth from the AI infrastructure build out, just like many, many other companies that are part of this ecosystem. And at the moment, it's probably the strongest growth driver in the Ratings business, albeit one of a number of these elements of this financing super cycle that we're in.

好消息是，我們正在從 AI 基礎設施建置中捕捉到成長，就像這個生態系中許多、許多其他公司一樣。而目前來看，這大概是評級（Ratings）業務中最強的成長驅動因素，儘管它只是我們所處的這一輪融資超級循環（financing super cycle）中多個要素之一。

So from an issuance growth perspective, year-to-date, AI-related issuance has accounted for a little over half of total issuance growth. And that's AI infrastructure kind of broadly defined. I won't get into everything that's included in it, but it's not just hyperscalers. This is -- goes into data centers and things related to that.

因此，從發行成長的角度來看，截至目前（year-to-date），與 AI 相關的發行量占了總發行成長的一半以上。而這裡的 AI 基礎設施是以相當廣義的方式定義。我不會細講其中包含的所有項目，但不僅僅是 hyperscalers。這還包括資料中心以及與之相關的項目。

Now from a revenue standpoint, that number is closer to 40%, and that's very consistent with what we've been telling you for much of the year that a lot of this big hyperscaler issuance is obviously now frequent -- kind of a frequent issuer commercial construct to it.

而從營收角度來看，這個比例更接近 40%，這也非常符合我們今年大部分時間一直告訴各位的：這些大型 hyperscaler 的發行顯然現在已經帶有某種「常態化發行人」的商業特性（frequent issuer commercial construct）。

So before anyone tries to turn a positive into a negative, I just -- I want to reinforce that I think we're in kind of a form of an intelligence revolution. And I have been out talking to this ecosystem, CFOs, CEOs, tech companies very extensively over the last several months. And everything I hear from them is that there is not a demand issue. There's really -- many companies would issue more debt if they could.

所以在任何人試圖把正面因素解讀成負面之前，我只是——我想再次強調，我認為我們正處於某種形式的智慧革命（intelligence revolution）。過去幾個月我非常密集地與這個生態系交流，包括 CFO、CEO、科技公司。我從他們那裡聽到的一切都顯示，並不存在需求問題。實際上——很多公司如果可以的話，會發行更多債務。

So there is plenty of demand. You can see that in the remaining performance obligations with the hyperscalers, which continue to grow very substantially. And when you look at our rating analysts have estimated there's something like $800 billion of CapEx for 2026. They expect that to grow to somewhere $1 trillion or more. So what that tells you, Craig, is we expect that there will be continued growth in the hyperscaler issuance and AI infrastructure build-out.

因此需求非常充足。你可以從 hyperscalers 的剩餘履約義務（remaining performance obligations）中看到這一點，該數字仍在非常顯著地成長。而當你看我們的評級分析師估算，2026 年的資本支出（CapEx）大約有 8,000 億美元。他們預期這會成長到 1 兆美元或以上。所以這告訴你，Craig，我們預期 hyperscaler 的發行量與 AI 基礎設施建置將會持續成長。

Rob, sorry, is that $1 trillion you're talking about you guys are thinking for next year or a few years out? Or -- ?

Rob，不好意思確認一下，你們提到的 1 兆美元，是指你們對明年，還是未來幾年後的預期？還是——？

That's for next year. So that I think of $800 billion-ish this year of CapEx, call that $1 trillion -- north of $1 trillion next year. So we expect that there will be growth in the infrastructure build-out and hyperscaler issuance investment in CapEx in 2027.

是明年。所以我認為今年的資本支出（CapEx）大概在 8,000 億美元左右，把它算作 1 兆美元——明年會超過 1 兆美元。因此我們預期，基礎設施建置以及超大規模雲端業者（hyperscaler）在 2027 年的資本支出發行與投資將會成長。

Alex Kramm, UBS.

Alex Kramm，瑞銀（UBS）。

Just, can we just touch on private credit for a little bit here. There's a lot of cross currents. Obviously, I think issuance is struggling a little bit relative to public markets, but you've been investing in that space a lot. At the same time, I think there continue to be headlines around some of the smaller rating agencies, maybe doing a lot of activity that people don't really see as correct. If that's the right word to use here.

我想稍微談一下私募信貸（private credit）。目前有很多相互牽動的因素。很明顯，相較於公開市場，我認為發行（issuance）有點吃力，但你們在這個領域投入很多。同時，我想也持續看到一些關於較小型評等機構的新聞標題，可能在做很多活動，但大家不太認為那是正確的。如果這樣用詞合適的話。

So yes, maybe just talk about your appetite to do more there, what you're seeing, and how your role has been evolving here over the last few quarters.

所以，是的，想請你談談你們在那邊進一步擴大的意願、你們看到的情況，以及過去幾季你們的角色是如何演變的。

Yes, Alex, thanks. And we've talked about this a lot over the last number of quarters and we had some work to do to be able to get the methodologies and the analytical resources and that, frankly, the engagement model to be able to support the growth in private credit over the last several years. And I think we've done a good job of that. And you've seen that come into the rating results. But you've also heard me say, in some ways, I wanted diagnostic to whether it's private credit funding or public credit funding.

是的，Alex，謝謝。我們在過去幾季談了很多，為了能支援過去幾年私募信貸的成長，我們確實需要做一些工作，包含方法論、分析資源，以及坦白說，能支援這種成長的合作／互動模式（engagement model）。我認為我們做得不錯。你也在評等結果中看到這些反映。但你也聽我說過，某種程度上，我希望我們對於是私募信貸融資或公開信貸融資能保持中立（agnostic）。

It's actually interesting, Alex, if we look at this past quarter, there was a significant rotation of private credit back into the broadly syndicated loan market. We actually -- there was a pitch book stat about $12 billion of direct lending deals refi from private credit into the BSL market in the third quarter. That was the highest quarterly volume since the tracking began back in 2022.

其實很有意思，Alex，如果我們看上一季，私募信貸明顯輪動回到廣泛聯貸（broadly syndicated loan, BSL）市場。我們其實——有一個 PitchBook 的統計指出，第三季大約有 120 億美元的直接放款（direct lending）交易從私募信貸再融資（refi）到 BSL 市場。那是自 2022 年開始追蹤以來的最高單季量。

So you know I guess what, we're going to rate that stuff, too. And so I'd say a few things just in terms of credit quality. We've got the same analyst rating the public credit and the private credit, right? And that makes me feel very comfortable. It's not some special different team. We have a lot of engagement with regulators who want to understand our approaches and how we're thinking about all of this. So that's great.

所以你知道，我想說的是，我們也會去評等那些東西。就信用品質而言，我會說幾點：我們是由同一批分析師同時評等公開信貸與私募信貸，對吧？這讓我非常放心。不是另外一個特殊、不同的團隊。我們也和監管機構有很多互動，他們想了解我們的方法與我們如何思考這一切。所以這很好。

And then you've seen us do the things like the MSCI and now the Allvue partnership. And what that's really about, Alex, is about all of these headlines are just leading investors to ask more questions about the credit risk that they are taking.

然後你也看到我們做了像是 MSCI，以及現在與 Allvue 的合作夥伴關係。而這背後真正的重點，Alex，是這些新聞標題正促使投資人對自己承擔的信用風險提出更多問題。

And I talked, for a few years, I'd hear kind of the talking points. I'd say, yes, but don't you want to have an independent assessment of credit, how much credit for the return that you say you're getting, do you understand how much risk you're taking. And so what we did with Allvue is very similar to what we did with MSCI. We had another data set that allows us to further calibrate our credit models and be able to make those available to Allvue and other customers so that they can get a probability of an independent rigorous probability default, it's mapped to a rating level to help investors.

而我在過去幾年常聽到一些既定說法。我會說，是的，但你們難道不希望有一個獨立的信用評估嗎？你們所說拿到的報酬，對應承擔了多少信用風險？你們了解自己承擔了多少風險嗎？所以我們和 Allvue 做的事情，和我們與 MSCI 做的非常類似。我們取得另一套資料集，讓我們能進一步校準（calibrate）信用模型，並把這些能力提供給 Allvue 與其他客戶，讓他們能得到一個獨立且嚴謹的違約機率（probability of default），並映射到評等等級，來協助投資人。

What we really want to do is start to drive the usage and utility of our tools and our credit language with investors in the private credit markets, just like we've done in the public credit markets. And ultimately, be agnostic to whether it's public or private credit.

我們真正想做的是，開始推動投資人在私募信貸市場使用並提升我們工具與信用語言的效用，就像我們在公開信貸市場做過的一樣。而最終，對於是公開或私募信貸保持中立。

David Motemaden, Evercore.

David Motemaden，Evercore。

I had a question on the insurance business, and it's interesting to hear that casualty is one of the fastest growth areas in the quarter. We're just seeing and hearing a lot of interest from third-party capital to access casualty business. So I'm wondering if you are seeing any impact of that yet in the results and how you're viewing that opportunity going forward?

我有一個關於保險業務的問題。聽到本季意外險（casualty）是成長最快的領域之一很有意思。我們看到、也聽到很多第三方資本對於進入意外險業務的興趣。所以我想知道，你們是否已經在結果中看到任何影響？以及你們如何看待未來這個機會？

Yes, David, thanks for the question. And like a lot of this stuff where we're innovating, the casualty is still a pretty small part of the business, but it's -- but we're seeing very good traction and interest. And why is that? In the US, casualty, I think, is something like 60% of gross written premiums. It's a huge part of the insurance business, but it is very immature in terms of its use of third-party data and analytics.

是的，David，謝謝你的問題。就像我們在創新的一些領域一樣，意外險目前仍是業務中相對小的一部分，但——我們看到非常好的進展與興趣。為什麼呢？在美國，意外險我記得大概占總簽單保費（gross written premiums）的 60%。它是保險業務非常大的一塊，但在使用第三方資料與分析方面仍非常不成熟。

When you think about property, property has been using these sophisticated CAT models for three decades. But what about casualty. We go into casualty, and people are talking about Google searches and all sorts of stuff. So that's why we made the investment a couple of years ago in that casualty analytics business. There's a strong kind of AI component to it.

你想想財產險（property），財產險已經使用這些精密的巨災（CAT）模型三十年了。但意外險呢？我們進到意外險領域，人們在談 Google 搜尋以及各式各樣的東西。所以這就是為什麼我們在幾年前投資了意外險分析業務。其中有很強的 AI 成分。

One of the things it does -- and I think I've talked about this in the past, it's scouring law journals and medical journals and finding emerging mass torts. And guess what? We alerted our customers to the mass tort risk of addictive software back in, I think, 2018, right?

它做的一件事——我想我以前提過——是掃描法律期刊與醫學期刊，找出新興的大規模侵權（mass tort）。你猜怎麼著？我們早在我記得是 2018 年，就提醒客戶注意成癮性軟體（addictive software）的大規模侵權風險，對吧？

And of course, we saw the Facebook settlement this year. So that's what that business is about. And what you heard us do with Lloyd's is Lloyd's understands there's potentially a lot of signal value in all of the company data that we have and connecting that data to be able to better understand casualty risk.

當然，我們今年也看到了 Facebook 的和解案。所以那就是那項業務在做的事。而你們聽到我們與勞合社（Lloyd's）合作的內容是：勞合社理解，我們掌握的所有公司資料可能具有很高的訊號價值，把這些資料串接起來，能更好地理解意外險風險。

So it was interesting, David, I can't remember if you were at Exceedance with us. But the casualty rooms and sessions, they were packed. And it was the same thing at Rendez-vous. There was so much discussion about casualty because it's a big part of the market, but people need better tools to be able to get comfortable with underwriting it.

所以很有意思，David，我不記得你是否有和我們一起參加 Exceedance。但意外險的會議室與場次都擠滿了人。在 Rendez-vous 也是同樣情況。大家對意外險有非常多討論，因為它是市場很大的一部分，但人們需要更好的工具，才能更有把握地承保（underwrite）它。

So this is one where I feel really good about what we're bringing to this market, and the insurance industry partnering with us to be able to bring these solutions to market. So you're going to keep hearing about this as a growth driver, I think, on future calls.

所以這一塊我對我們帶給市場的東西感覺非常好，而保險業也正與我們合作，把這些解決方案推向市場。因此我想，在未來的電話會議上，你們會持續聽到這是成長驅動因素之一。

Owen Lau, Clear Street.

Owen Lau，Clear Street。

Could you please talk about MA's longer-term model of selling data. So given the pros and cons, do you think MA can get more revenue by entitling more users to get data through AI models and MCP or sell the data directly through your own channel. I know you're doing both, but I'm wondering how Moody’s think about protecting data getting incremental revenue, but not cannibalizing your own platform?

可以請你談談 MA 較長期的資料銷售模式嗎？也就是說，考量利弊，你認為 MA 透過 AI 模型與 MCP 讓更多使用者取得資料授權（entitling more users to get data），與透過你們自有通路直接銷售資料相比，哪一種能帶來更多營收？我知道你們兩者都在做，但我想了解穆迪（Moody’s）如何思考在保護資料、取得增量營收的同時，又不會侵蝕（cannibalize）你們自家平台？

Yes. Great question, Owen. So let me think about how I want to answer this. There are different speeds of customers, right? First of all, we have to understand that. So we have the big sophisticated financial institutions and most of them want to consume our content via MCPs into either their own AI platforms or some of the third-party UIs they're using.

是的。問得很好，Owen。讓我想想我想怎麼回答這個問題。客戶採用的速度各不相同，對吧？首先，我們必須先理解這一點。我們有大型且成熟的金融機構，而其中多數希望透過 MCP 將我們的內容導入他們自有的 AI 平台，或他們正在使用的某些第三方 UI。

And so we're doing that. And we're doing that either via MCPs or in some cases, through specialized agents that will be in their platform. The good thing about that, Owen, is that if you think about over the years, Moody's has done a great job of monetizing the expert user community, right? These are people that are analyzing fixed income portfolios or CAT modeling professionals, but there's a lot of utility for our content kind of down that sophistication curve or user curve.

所以我們正在這麼做。我們要嘛透過 MCP 來做，要嘛在某些情況下，透過會部署在他們平台內的專用代理（agents）。Owen，這樣做的好處是，如果你回顧過去多年，穆迪一直很擅長將專家使用者社群變現，對吧？這些人是在分析固定收益投資組合或做 CAT（巨災）模型的專業人士，但沿著那條成熟度曲線或使用者曲線往下，其實我們的內容也有很大的效用。

And so when you think about deploying, I was just at one of the largest banks in the country, I guess it was Monday with Cristina. And what we were talking about is they've got 8,000 users on one of these AI surfaces, and they said, "Look, we want to bring your content into that surface". Great. Right now, we have a fraction of that, that's using some of that content.

所以當你想到部署時，我週一和 Cristina 一起去了國內最大的銀行之一。我們談到的是，他們在其中一個 AI 介面上有 8,000 名使用者，他們說：「我們希望把你們的內容帶進這個介面。」很好。而目前使用其中部分內容的，只是其中一小部分。

And so this gives us an opportunity to expose our content to thousands and thousands more users and get an uplift in doing that and then thinking about elements of consumption-based pricing that will include in that model.

因此，這讓我們有機會把內容曝光給成千上萬更多的使用者，並在此過程中帶來提升，同時也思考在該模式中納入以使用量為基礎的定價元素。

That said, I think you'll hear more from Cristina about this early in 2027. I don't think we just want to have data pipes and the customers, we also want to think about how we move up the value stack. And ultimately, have kind of embedded solutions with the customers that are driving very significant outcomes in terms of efficiency and effectiveness.

話雖如此，我想你會在 2027 年初從 Cristina 那裡聽到更多相關內容。我不認為我們只想做資料管道、把資料送到客戶那裡；我們也想思考如何往價值鏈上移。最終，與客戶一起打造某種嵌入式解決方案，在效率與效能方面帶來非常顯著的成果。

And so we have a lot to offer there. You've heard some of these examples that I've been talking about. But that means, I think, I talked about that kind of FD model of co-building with our customers is ultimately a number of customers are going to be migrating over time from software solutions who more agentic workflows, and we want to do that with our customers.

所以我們在那方面有很多可以提供的。你也聽過我提到的一些例子。但我認為，這也意味著我談到的那種與客戶共同打造的 FD 模式，最終會有不少客戶隨時間從軟體解決方案遷移到更具代理式（agentic）的工作流程，而我們希望與客戶一起完成這件事。

And we want to be with them with AI-enabled software. We want to be with them piping our data into their platforms, and we want to be co-building the next generation of agentic solutions with them, all of which are ways to consume our connected intelligence.

我們希望以 AI 賦能的軟體與他們同行。我們希望把資料輸送到他們的平台中與他們同行；我們也希望與他們共同打造下一代的代理式解決方案——這些都是消費我們「連結式智慧」的方式。

Peter Christiansen, Citigroup.

Peter Christiansen，花旗集團。

Rob, I guess, as you head into the largest renewal quarter, what are you seeing from a pricing perspective across Moody's Analytics? Are customers demonstrating a willingness to pay more for AI capabilities, workflow and integration, what have you? Just wondering if you could talk about what trends you're seeing in regards to that dynamic. Thank you.

Rob，我想問的是，當你們進入最大的一個續約季度時，從穆迪分析（Moody's Analytics）的定價角度來看，你們看到了什麼？客戶是否展現出願意為 AI 能力、工作流程與整合等付出更高費用的意願？想請你談談你看到的、與這種動態相關的趨勢。謝謝。

Yes. Thanks. And I know Noemie has bumped out that she did not get a chance to answer a question on this call. But, so look, I think it comes back to a couple of things. A lot of these institutions have very significant investment agendas around their AI platforms and very critically, the data that goes into the platforms.

是的，謝謝。我也知道 Noemie 已經先離線了，她沒有機會在這通電話上回答相關問題。不過你看，我認為這回到幾個重點。許多這類機構在其 AI 平台上有非常重大的投資規劃，而更關鍵的是，投入平台的資料。

So there is budget for all of that. But it comes back to you've got to prove the value of your stuff, right? And so that's why you hear us talking about the outcomes and the efficiency and the effectiveness. And by the way, we did a really interesting brief POC with -- it was with our credit MCP integrated into one of these AI surfaces. And the answer is with and without the Moody's kind of was actually shocking to me.

所以這些都有預算。但重點回到：你必須證明你提供的東西有價值，對吧？因此你會聽到我們一直在談成果、效率與效能。順帶一提，我們做了一個非常有意思、簡短的 POC——是把我們的信用 MCP 整合到其中一個 AI 介面裡。結果是：有沒有穆迪的差異，實際上讓我很震驚。

How many errors there were without the Moody's credit MCP when you were querying around credit topic. So banks understand, financial institutions understand the value of this stuff. We're trying to make it -- that's why we're trying to be very crisp about the outcomes that we think we can deliver. And if you can deliver those kinds of outcomes, those kinds of time savings, those kinds of increased effectiveness.

當你在信用主題上進行查詢時，如果沒有穆迪的信用 MCP，錯誤竟然那麼多。所以銀行懂、金融機構懂這些東西的價值。我們正努力讓它——這也是為什麼我們要非常清楚、精準地說明我們認為能交付的成果。如果你能交付那樣的成果、那樣的節省時間、那樣的效能提升。

I talked about 95% first-pass accuracy and that Risk Data Refinery. Institutions are going to pay for that because the ROI is very clear. But in many cases, the customers want to test drive this first to make sure that they can confirm the business case for it. And so, -- I would say the selling motions are a little different, but I feel very good about the engagement we’ve got with customers, the pricing opportunity, where the customers are seeing real value.

我提到過 Risk Data Refinery 的首次通過準確率達到 95%。機構會為此付費，因為投資報酬率（ROI）非常清楚。但在很多情況下，客戶會希望先試用，確認其商業案例成立。因此——我會說銷售推進方式有些不同，但我對我們與客戶的互動、定價機會，以及客戶所看到的真實價值，都感到非常有信心。

We have reached the end of the Q&A session. I will now turn the call back to Rob for closing remarks.

問答環節到此結束。我現在把電話交回給 Rob 作結語。

All right. Thanks, everybody for joining, and we will talk to you early in the new year. All right. Take care. Bye-bye.

好的。感謝各位參與，我們會在新年初再與各位交流。好的。保重。再見。

This concludes Moody's Corporation third-quarter 2026 earnings call.

以上為穆迪公司 2026 年第三季財報電話會議。

As a reminder, immediately following this call, the company will post the MIS revenue breakdown under the Investor Resources section of the Moody's IR homepage. Additionally, a replay will be made available after the call on the Moody's IR website. Thank you for attending. You may disconnect.

提醒各位，本次電話會議結束後，公司將在穆迪投資人關係（IR）首頁的「投資人資源」區發布 MIS 營收拆分明細。此外，會後也將在穆迪 IR 網站提供重播。感謝各位出席。您現在可以掛線。