Gemini Space Station Inc (GEMI) 2025 Q4 法說會逐字稿

Thanks, operator, and thank you, everyone, for joining this morning for Gemini's fourth-quarter and full-year 2025 earnings call. My name is Ryan Todd, Head of Investor Relations at Gemini. Joining me on the call today are Gemini's Founders, Cameron and Tyler Winklevoss; and Interim CFO, Danijela Stojanovic. Yesterday, we released our fourth-quarter and full-year 2025 financial results.

謝謝，接線員，也謝謝各位今天早上參加 Gemini 2025 年第四季與全年財報電話會議。我是 Ryan Todd，Gemini 投資人關係主管。今天與我一同參與電話會議的有 Gemini 創辦人 Cameron 與 Tyler Winklevoss，以及代理財務長 Danijela Stojanovic。昨天，我們已發布 2025 年第四季與全年財務結果。

During today's call, we may make forward-looking statements which may vary materially from actual results and are based on management's current expectations, forecasts, and assumptions. Information concerning the risks, uncertainties, and other factors that could cause these results to differ is included in our SEC filings.

在今天的電話會議中，我們可能會發表前瞻性陳述；此等陳述可能與實際結果存在重大差異，並係基於管理層目前的預期、預測與假設。有關可能導致結果出現差異之風險、不確定性及其他因素的資訊，已載於我們向 SEC 提交的文件中。

Our discussion today will also include certain non-GAAP financial measures. Reconciliation to the most directly comparable GAAP financial measures are provided in the shareholder letter on our investor Relations website and on the SEC's website. Non-GAAP financial measures should be considered in addition to, not as a substitute for, GAAP measures.

我們今天的討論也將包含若干非 GAAP 財務衡量指標。與最直接可比之 GAAP 財務衡量指標的調節表，已提供於我們投資人關係網站上的股東信以及 SEC 網站上。非 GAAP 財務衡量指標應作為 GAAP 指標的補充，而非替代。

We'll start today's call with prepared remarks and then take questions. And with that, let me turn the call over to our Founders, Cameron and Tyler.

我們將先進行預先準備的發言，接著進入問答。接下來，讓我把電話會議交給我們的創辦人 Cameron 與 Tyler。

Thanks, Ryan. Cameron here. 2025 was a remarkable year for Gemini. We crossed the threshold into the public markets and became a public company on September 12 after being a private company for over a decade. On that day, the price of Bitcoin was $115,000.

謝謝你，Ryan。我是 Cameron。2025 年對 Gemini 而言是非凡的一年。在作為私人公司超過十年後，我們於 9 月 12 日跨入公開資本市場並成為上市公司。當天，比特幣價格為 115,000 美元。

Since then, Bitcoin has traveled down to $60,000 and then back up to around $70,000 where it hovers today, a reminder that one of the biggest challenges for crypto builders and investors is its cyclical nature and a reminder that in order to move beyond these cycles, you need to build beyond them.

此後，比特幣一路下探至 60,000 美元，之後又回升至約 70,000 美元並在今日附近徘徊；這提醒我們，加密領域的建設者與投資人面臨的最大挑戰之一是其週期性，也提醒我們若要超越這些循環，就必須打造能超越循環的事物。

We started as a Bitcoin company. We became a crypto company. We are now becoming a markets company. If Gemini's first decade was building a bridge to the future of money, today, we are building a bridge to the future of money and markets via a super app.

我們起初是一家比特幣公司。我們後來成為一家加密公司。我們現在正成為一家市場公司。如果說 Gemini 的第一個十年是在搭建通往貨幣未來的橋樑，那麼今天，我們正透過一個超級 App 搭建通往貨幣與市場未來的橋樑。

Our first foray into people's daily financial lives, beyond buy, sell, and store crypto, began with the Gemini Credit Card which delivered strong growth last year. In 2025, card sign-ups grew nearly 15 times and credit card revenue reached $33.1 million, up 185% year over year. Many of these Gemini Credit Card customers engage with Gemini multiple times a day to earn crypto rewards when they spend with the Gemini Credit Card.

我們首次深入人們的日常金融生活、超越買入、賣出與存放加密資產的嘗試，始於 Gemini 信用卡；該產品在去年帶來強勁成長。2025 年，信用卡申請數成長近 15 倍，信用卡營收達 3,310 萬美元，年增 185%。許多 Gemini 信用卡客戶每天多次與 Gemini 互動，使用 Gemini 信用卡消費以賺取加密回饋。

December marked a new era for Gemini with the launch of Gemini Predictions. We believe prediction markets will be as big or bigger than today's capital markets. They offer a profound and boundless opportunity to leverage the wisdom of the crowds and the power of markets to provide unique insights into the future.

12 月，隨著 Gemini Predictions 的推出，Gemini 進入新時代。我們相信，預測市場將與當今資本市場同等規模，甚至更大。它們提供深刻且無邊界的機會，能運用群眾智慧與市場力量，對未來提供獨特洞見。

Our investment in securing a designated contract market DCM license from the CFTC to launch our own prediction marketplace positions us as an early mover on this new and exciting frontier. We've been building and operating regulated marketplace infrastructure for over a decade. Sequencers, matching engines, order books, real-time settlement, post-trade reporting, custody infrastructure, and more.

我們投資取得 CFTC 核發的指定合約市場（DCM）牌照，以推出自有的預測市場平台，使我們得以在這個令人振奮的新前沿成為早期進入者。我們建置並營運受監管的市場基礎設施已逾十年。包括排序器、撮合引擎、訂單簿、即時結算、交易後申報、託管基礎設施等。

This is a big part of what we do best. Our prediction markets are new instruments running on infrastructure we already know how to build and operate. As a result, we chose not to partner with a third party or license someone else's technology and instead build it ourselves.

這是我們最擅長的核心之一。我們的預測市場是新的金融工具，運行在我們已熟悉如何建置與營運的基礎設施之上。因此，我們選擇不與第三方合作或授權他人的技術，而是自行打造。

In doing so, this also means we have chosen to invest in developing the unique operational capabilities for creating and resolving thousands of contracts on a daily basis. A new and fascinating challenge with growing complexity as we expect the cardinality of these markets to continue to explode over time.

如此一來，也意味著我們選擇投資於開發獨特的營運能力，以便每日建立並結算數以千計的合約。這是一項新且引人入勝的挑戰；隨著我們預期這些市場的規模與種類數量（cardinality）將隨時間持續爆炸式成長，其複雜度也將不斷提高。

In short, we built Gemini Predictions from the ground up because we want to own and operate our prediction markets end-to-end for the long term. We believe in the power of markets. Bitcoin is a store of value that is a product of market forces. The best economies are market-based. Markets are truth over the long term, and we believe that we are just figuring out how to apply them to the world around us.

簡言之，我們從零開始打造 Gemini Predictions，因為我們希望長期端到端擁有並營運我們的預測市場。我們相信市場的力量。比特幣是一種價值儲存工具，是市場力量的產物。最好的經濟體是以市場為基礎的。長期而言，市場趨近於真相；我們相信，我們才剛開始摸索如何將市場應用到我們周遭的世界。

From politics to economic indicators, business, tech, culture, and sports, prediction markets are forecasting the future more accurately and more quickly than traditional pollsters, experts, and the media. This is a profound change in the world's source of truth and an equally profound solution to the loss of trust in our institutions and resulting epistemological crisis.

從政治到經濟指標、商業、科技、文化與體育，預測市場在預測未來方面，比傳統民調機構、專家與媒體更準確也更迅速。這是世界「真相來源」的深刻轉變，也同樣是對我們機構信任流失及其所導致的認識論危機的一項深刻解方。

The printing press created the fourth estate or the public press. The Internet created the fifth estate or decentralized public press. Prediction markets are creating the sixth estate. Decentralized information combined with the integrity and accountability of markets or skin in the game. Like money, markets are an innovation in technology that continue to evolve thousands of years after they were first invented.

印刷術創造了第四權力，也就是公共媒體。網際網路創造了第五權力，也就是去中心化的公共媒體。預測市場正在創造第六權力。去中心化資訊結合市場的完整性與問責性，也就是「有真金白銀的利害關係（skin in the game）」。如同貨幣，市場也是一項技術創新；自其首次被發明以來，歷經數千年仍持續演進。

From the birth of the bond markets in the Italian city-states in the twelfth century to the launch of the first stock market in Amsterdam in the seventeenth century to electronic trading replacing the open outcry of humans in trading pits on Wall Street in the twenty-first century, markets continue to grow and develop.

從十二世紀義大利城邦債券市場的誕生，到十七世紀阿姆斯特丹第一個股票市場的啟動，再到二十一世紀電子交易取代華爾街交易場內的人聲喊價，市場持續成長並發展。

Just when you thought the money experiment had reached its terminal steady state, Bitcoin emerged. Just when you thought markets were done maturing, prediction markets caught fire. Gemini was founded to help build and shape a new era of money.

就在你以為貨幣實驗已達到最終穩態之時，比特幣出現了。就在你以為市場已不再成熟之時，預測市場迅速走紅。Gemini 的成立，是為了協助建構並塑造貨幣的新時代。

Today, we have a similar opportunity to help build and shape a new era of markets. Unfolding in parallel is the meteoric ascent of AI. Once these strains of technology, money, markets, and AI converge, we believe they will supercharge each other in dramatic and novel ways that generate new economic activity that we are uniquely positioned to be at the center of and help build and shape too. This cauldron of Promethean fire could make progress in these fields up to this point appear rather quaint.

今天，我們也有類似的機會，協助建構並塑造市場的新時代。與此同時，AI 正以流星般的速度崛起。一旦這些技術、貨幣、市場與 AI 的脈絡匯聚，我們相信它們將以戲劇性且新穎的方式彼此加乘，創造新的經濟活動；而我們具備獨特優勢，能站在其核心並參與建構與塑造。這鍋普羅米修斯之火的熔爐，可能會讓迄今為止在這些領域的進展顯得相當樸素。

We have long felt that it is only a matter of time before we have more machines as customers than humans. Machines can't open a bank account, but they can easily plug into protocols and use crypto to become rational economic actors. Humans may have built crypto, but crypto is not so much money for humans as it is money for machines. We're just starting to see this take shape.

我們長期以來一直認為，客戶是機器而非人類，這只是時間問題。機器無法開立銀行帳戶，但它們可以輕易接入各種協議，並使用加密資產成為理性的經濟行為者。人類或許打造了加密，但加密與其說是給人類的貨幣，不如說是給機器的貨幣。我們才剛開始看到這一切逐漸成形。

Here's one example. For the first decade, we had three API protocols, REST, WebSocket, and FIX. We're now adding a fourth, Model Context Protocol, or MCP, an open source API interface designed specifically for AI agents like large language models or LLMs.

這裡有一個例子。在第一個十年，我們有三種 API 協議：REST、WebSocket 與 FIX。我們現在正在新增第四種：模型上下文協議（Model Context Protocol，MCP），這是一個專為 AI 代理（例如大型語言模型或 LLM）設計的開源 API 介面。

While we believe AI is going to change the composition of our customer base, it's already changing the composition of our workforce and how we work. Up until recently, the impact of software engineers could differ by an order of magnitude or 10x. Great engineers would have 10x more impact than good engineers.

雖然我們相信 AI 將改變我們客戶群的組成，但它已經在改變我們員工隊伍的組成以及我們的工作方式。直到最近，軟體工程師的影響力可能相差一個數量級，也就是 10 倍。頂尖工程師的影響力會比優秀工程師高出 10 倍。

AI has completely changed the game, expanding this paradigm by another order of magnitude at a minimum, making a 10x-er now a 100x-er. Critically, we are seeing that this step change holds true for every engineer who adopts AI in their workflows, and it also holds true for non-engineering work as well.

AI 已徹底改變了遊戲規則，至少再把這個範式擴大一個數量級，讓原本的 10 倍高手如今成為 100 倍高手。關鍵在於，我們看到這種躍升對於每一位在工作流程中採用 AI 的工程師都成立，對非工程類工作同樣成立。

Doing more with less has never been more true or possible, and we believe this trend line is only just beginning. Notably, the force multiplier effect of AI for Gemini and our workforce is quite new. It wasn't until the end of last year that AI agents for coding and software development had a splitting of the atom moment.

以更少資源做更多事情，從未像現在這樣真實或可行，而我們相信這條趨勢線才剛開始。值得注意的是，AI 對 Gemini 與我們員工隊伍的「倍增器」效應相當新。直到去年年底，用於寫程式與軟體開發的 AI 代理才迎來了如同「分裂原子」般的關鍵時刻。

While different pockets of our technology organization had been experimenting with AI and their workflows for a while, AI was not core to them. For example, late last summer when we were in the middle of our IPO roadshow, AI was used in only 8% of the code being written and shipped to production. In December, however, the future arrived. Models hit an inflection point, and in combination with the internal tools we built for change set management, AI is now too powerful not to use at Gemini.

雖然我們技術組織中的不同團隊已經在其工作流程中試驗 AI 一段時間，但 AI 並非其核心。例如，去年夏末我們正處於 IPO 路演期間，寫入並部署到正式環境的程式碼中，只有 8% 使用了 AI。然而在 12 月，未來到來了。模型達到拐點，再加上我們為變更集管理打造的內部工具，AI 如今強大到在 Gemini 不用都不行。

Today, AI is used in more than 40% of our production code changes, and we expect that number to climb close to 100% in the not-too-distant future. Not using AI at Gemini will soon be the equivalent of showing up to work with a typewriter instead of a laptop. As a result, we have reduced the size of our workforce by roughly 30% since the start of 2026. We believe that a smaller organization leveraging the right tools isn't just more efficient, it's actually faster.

今天，AI 已用於我們超過 40% 的正式環境程式碼變更，我們預期在不久的將來這個比例將攀升至接近 100%。在 Gemini 不使用 AI，很快就等同於帶著打字機而不是筆電來上班。因此，自 2026 年初以來，我們已將員工規模縮減約 30%。我們相信，規模更小、並善用正確工具的組織不僅更有效率，實際上也更快。

Gemini started in America in 2015. Since then, we expanded our areas of operation to more than 60 countries. These foreign markets proved hard to win in for various reasons, and we found ourselves stretched thin with a level of organizational and operational complexity that drove our cost structure up and slowed us down.

Gemini 於 2015 年在美國起步。此後，我們將營運範圍擴展至 60 多個國家。由於各種原因，這些海外市場很難攻下，我們發現自己被拉得很分散，組織與營運的複雜度推升了成本結構並拖慢了我們的速度。

We didn't have the demand in these regions to justify them. The reality is that America has the world's greatest capital markets, and America has always been where it's at for Gemini. Furthermore, we are encouraged by the stated goals of the current SEC and CFTC and their efforts thus far to make the super app possible in America and usher in a new golden age of markets.

我們在這些地區的需求不足以支撐其合理性。現實是，美國擁有全球最強的資本市場，而美國一直都是 Gemini 的主場。此外，我們也受到現任 SEC 與 CFTC 所宣示目標的鼓舞，以及他們迄今為止為了讓「超級 App」在美國成為可能、並開啟市場新黃金時代所做的努力。

We decided it was time for us to focus and double down on America. This will allow us to build more meaningful and powerful relationships with new and existing customers. To that end, in addition to reducing the size of our workforce, we have reduced the areas in which we operate by exiting the UK, EU, and Australian markets.

我們決定是時候聚焦並加碼美國市場。這將使我們能與新舊客戶建立更有意義、更強大的關係。為此，除了縮減員工規模之外，我們也透過退出英國、歐盟與澳洲市場，縮小了我們的營運區域。

We expect this will help reduce our total expenses in line with our headcount reduction and meaningfully accelerate our path to profitability, even in the backdrop of the current crypto market. Simplify, consolidate, then accelerate. We love being a public company.

我們預期這將有助於在員工人數縮減的同時降低總費用，並在當前加密市場背景下，顯著加速我們邁向獲利的進程。簡化、整併，然後加速。我們喜歡成為一家上市公司。

Perhaps a somewhat surprising statement when looking at the performance of our share price over the past six months since we've been public. Rather than be dispirited, we are motivated. While it's never fun to see your stock drop, we love the feedback loop. It forces us to confront what is working and what is not working, and it makes us sharper. It's challenging, but absolutely the right challenge.

考量到我們上市後過去六個月股價表現，這或許是個有些令人意外的說法。我們不但沒有氣餒，反而更有動力。雖然看到股價下跌從來都不好受，但我們喜歡這個回饋迴路。它迫使我們正視哪些有效、哪些無效，並讓我們更敏銳。這很具挑戰，但絕對是正確的挑戰。

We view this feedback loop as one of the greatest benefits of being a public company. As rowers, losing a race provided invaluable feedback on the changes you needed to make in order to win. The path to the Olympics is paved in lost races and the invaluable learning that comes from them.

我們將這個回饋迴路視為成為上市公司最大的好處之一。作為划船選手，輸掉一場比賽會提供無比珍貴的回饋，讓你知道為了贏需要做哪些改變。通往奧運之路，是由一場場失利與從中獲得的寶貴學習所鋪成。

We welcome the feedback and love the challenge. 2025 marked the end of Gemini 1.0, and 2026 marks the beginning of Gemini 2.0. This starts with our shift into becoming a markets company with Gemini Predictions and using the same infrastructure to power our perpetual futures contracts once these contracts are allowed in the US. It continues with our plan to launch US equities as the next phase of our platform, giving our customers access to the largest, most liquid markets in the world.

我們歡迎回饋，也熱愛挑戰。2025 年標誌著 Gemini 1.0 的結束，而 2026 年標誌著 Gemini 2.0 的開始。這首先體現在我們轉型成為一家市場公司：推出 Gemini Predictions，並在美國允許這些合約後，使用同一套基礎設施來支援我們的永續期貨合約。接著，我們計畫推出美國股票作為平台的下一階段，讓客戶得以進入全球最大、流動性最強的市場。

Altogether, we have developed the foundation and building blocks for a super app where users will be able to fulfill their existing and future financial needs all in one place. Amazing awaits.

總體而言，我們已打造出「超級 App」所需的基礎與積木，讓使用者能在同一個地方滿足其現有與未來的金融需求。精彩可期。

Thank you, Cameron and Tyler, and great to speak with everyone. Before I turn to the numbers, I'll briefly note that I stepped into the interim CFO role earlier this year after serving as Gemini's Chief Accounting Officer since May of 2025. I've been closely involved in the company's financial reporting, the IPO process, and the prior two quarters as a public company. The broader finance organization remains fully in place, and there has been no disruption to our financial reporting or operational execution.

謝謝 Cameron 和 Tyler，也很高興與各位交流。在我開始談數字之前，我先簡要說明：今年稍早，在自 2025 年 5 月起擔任 Gemini 首席會計官之後，我接任了代理 CFO 的職務。我一直密切參與公司的財務報導、IPO 流程，以及作為上市公司之前兩個季度的相關工作。更廣泛的財務組織仍完整運作，我們的財務報導或營運執行沒有受到任何干擾。

I will begin with a few key takeaways from the quarter before walking through their results in more detail. First, revenue grew sequentially despite a materially weaker crypto trading environment in Q4. Second, the business continued to diversify meaningfully. Services revenue more than doubled year-over-year and now represent over a third of our revenue.

在更詳細地說明本季結果之前，我先分享幾個關鍵重點。第一，儘管第四季加密交易環境明顯轉弱，營收仍較前一季成長。第二，業務持續顯著多元化。服務營收年增超過一倍，且目前已占我們營收的三分之一以上。

Third, the restructuring actions we announced earlier this year repositioned the company with a significantly lower cost base going into 2026. Now turning to the results. Net revenue for the fourth quarter was $56.4 million, up 13% from $49.8 million in Q3. This growth occurred despite a more challenging market backdrop.

第三，我們今年稍早宣布的重整措施，使公司在邁入 2026 年時擁有顯著更低的成本基礎。接著看本季結果。第四季淨營收為 5,640 萬美元，較第三季的 4,980 萬美元成長 13%。即使在更具挑戰的市場背景下，仍實現了這項成長。

The biggest driver of that change was volatility in the crypto market. Bitcoin fell nearly 47% from its October high, and that environment put real pressure on trading volumes and transaction fees. The credit card business kept growing through it, which helped offset some of that, but Q4 was a harder macro quarter than Q3. I'll walk through the key components.

造成這項變化的最大驅動因素是加密市場的波動。比特幣自 10 月高點下跌近 47%，這樣的環境對交易量與交易手續費帶來實質壓力。信用卡業務在此期間仍持續成長，幫助抵銷了部分影響，但第四季的總體環境比第三季更艱難。我將逐一說明主要組成。

Transaction revenue was $26.7 million, up slightly from $26.3 million in Q3 on spot volumes of $11.5 billion compared to $16.4 billion in Q3. Retail volumes came in at $1.6 billion and institutional at $9.9 billion. As a reminder, we earn fees from both retail and institutional customers with rates varying by order type. Instant orders at the top of the range and active trader orders lower.

交易營收為 2,670 萬美元，較第三季的 2,630 萬美元小幅上升；現貨交易量為 115 億美元，相較第三季的 164 億美元。其中，零售交易量為 16 億美元，機構交易量為 99 億美元。提醒一下，我們向零售與機構客戶皆收取費用，費率會依訂單類型而有所不同。即時訂單的費率位於區間上緣，而活躍交易者訂單的費率較低。

While volumes declined, transaction revenue proved relatively resilient. This reflects improvement in fee economics across both retail and institutional trading, as well as a mix shift in retail trading towards higher fee order types. Services revenue for the quarter was $26.5 million, up 33% sequentially from $19.9 million in Q3. This category continues to grow quickly and represents one of the most important structural shifts in our business.

儘管交易量下滑，交易收入仍展現出相對韌性。這反映出零售與機構交易兩端的費率經濟性均有所改善，以及零售交易的組合轉向手續費較高的訂單類型。本季服務收入為 2,650 萬美元，較第三季的 1,990 萬美元季增 33%。此類別持續快速成長，並代表我們業務中最重要的結構性轉變之一。

A few things worth calling out here. Credit card revenue was $16 million, up 87% from Q3's $8.5 million. We added nearly 30,000 new card sign-ups in the quarter, compared to 64,000 in Q3, and receivable balances grew to $219.8 million.

這裡有幾點值得特別說明。信用卡收入為 1,600 萬美元，較第三季的 850 萬美元成長 87%。本季新增近 30,000 名新卡申辦，相較第三季為 64,000 名，且應收帳款餘額增至 2.198 億美元。

Staking revenue was $5.1 million, down 13% from Q3's $5.9 million, largely reflecting lower crypto asset prices during the quarter. However, we continue to see adoption of staking across the platform, including through auto-staking features integrated with the credit card rewards program. Q4 was our first full quarter with card auto-staking rewards live, which came alongside the Solana card launch in October.

質押收入為 510 萬美元，較第三季的 590 萬美元下降 13%，主要反映本季加密資產價格較低。然而，我們仍持續看到平台上的質押採用度提升，包括透過與信用卡回饋計畫整合的自動質押功能。第四季是我們信用卡自動質押回饋上線後的第一個完整季度，並與 10 月推出的 Solana 卡同步。

That feature is a great example of natural multi-product engagement in providing customers a way to stake organically. They pick a stakeable reward, it gets staked automatically on every card transaction, and they're a staking customer without any extra steps. Staking balances at quarter end were approximately $509 million. Staking fee rate adjustment we made in Q3 also ran through a full quarter for the first time.

該功能是自然的多產品互動的絕佳例子，為客戶提供一種可自然進行質押的方式。客戶選擇可質押的回饋項目後，每筆刷卡交易都會自動進行質押，無需任何額外步驟就成為質押客戶。季度末的質押餘額約為 5.09 億美元。我們在第三季進行的質押費率調整也首次完整反映於一整個季度。

Let me turn to expenses. Total operating expenses for Q4 were $171.7 million, essentially flat compared to Q3. Compensation and headcount expenses declined to $72.3 million from $82.5 million in Q3, reflecting lower stock-based compensation expense. Stock-based comp in Q4 was $36 million. Headcount at quarter end was 650 compared to 677 in Q3.

接著我談費用。第四季總營運費用為 1.717 億美元，與第三季基本持平。薪酬與人力成本由第三季的 8,250 萬美元降至 7,230 萬美元，反映股權激勵費用下降。第四季股權激勵費用為 3,600 萬美元。季度末員工人數為 650 人，第三季為 677 人。

Importantly, the roughly 30% workforce reduction that occurred in early 2026 is not yet reflected in those numbers. That impact starts flowing through in Q1 of 2026, with the full run rate savings expected to be reflected by Q3 and beyond. As of March 1, total headcount was approximately 445.

重要的是，2026 年初約 30% 的人力縮減尚未反映在上述數字中。其影響將自 2026 年第一季開始逐步反映，預期至第三季及之後將呈現完整的年度化節省效果。截至 3 月 1 日，總員工人數約為 445 人。

Sales and marketing was $39 million, up from Q3's $32.9 million, reflecting the continued growth and momentum of the credit card portfolio and increased cardholder spending, which drove higher crypto rewards during the fourth quarter. As we've said consistently, we treat marketing as a variable line and calibrate it to what we're seeing in acquisition performance and growth opportunities.

銷售與行銷費用為 3,900 萬美元，高於第三季的 3,290 萬美元，反映信用卡產品組合持續成長與動能，以及持卡人支出增加，帶動第四季加密回饋提高。如我們一貫所述，我們將行銷視為可變動的費用項目，並依據獲客成效與成長機會進行校準。

For the full year, sales and marketing was $97.1 million or $52.5 million, excluding credit card rewards and promotions, which remained in line with the $45 million to $60 million range we previously guided to. Transaction processing expenses were $7.3 million, down from Q3's $8.6 million, reflecting lower trading volumes during the quarter. Transaction losses were at $6 million, down from Q3's $7.7 million.

全年銷售與行銷費用為 9,710 萬美元；若排除信用卡回饋與促銷，則為 5,250 萬美元，仍符合我們先前指引的 4,500 萬至 6,000 萬美元區間。交易處理費用為 730 萬美元，低於第三季的 860 萬美元，反映本季交易量較低。交易損失為 600 萬美元，低於第三季的 770 萬美元。

This includes our provision for credit losses on the card of $2.8 million, which remain broadly consistent with the prior quarter. Overall, credit quality across the card portfolio continues to remain stable as the book scales. Technology and infrastructure was $22.3 million, up from Q3's $20.3 million, mainly reflecting higher cloud infrastructure and software licensing costs as the platform scaled.

其中包含信用卡信用損失準備 280 萬美元，與前一季大致一致。整體而言，隨著信用卡業務規模擴大，卡片產品組合的信用品質仍維持穩定。科技與基礎設施費用為 2,230 萬美元，高於第三季的 2,030 萬美元，主要反映平台擴張帶來較高的雲端基礎設施與軟體授權成本。

General and administrative was $24.9 million, up from Q3's $19.3 million, driven mainly by higher professional services and ongoing public company operating costs. Full-year tech and G&A came in at $154.6 million, in line with our guidance range.

一般及行政費用為 2,490 萬美元，高於第三季的 1,930 萬美元，主要由較高的專業服務費用及持續的上市公司營運成本所帶動。全年科技與一般及行政費用合計為 1.546 億美元，符合我們的指引區間。

Now turning briefly on to full-year metrics. We served approximately 601,000 MTUs as of December 31st, up 17% year-over-year, reflecting continued growth and engagement as users adopt additional products across the platform. Full-year net revenue was $174 million compared to $141 million in 2024, up 24% year-over-year.

接著簡要說明全年指標。截至 12 月 31 日，我們服務約 601,000 名 MTU（每月交易用戶），年增 17%，反映用戶在平台上採用更多產品所帶來的持續成長與互動提升。全年淨收入為 1.74 億美元，較 2024 年的 1.41 億美元年增 24%。

Transaction revenue for the year was $98 million, while services and interest revenue reached $76 million, representing a significant and growing portion of our overall revenue base. This shift towards services is a key structural change, reducing dependence on trading activity.

全年交易收入為 9,800 萬美元，而服務與利息收入達 7,600 萬美元，已成為我們整體收入基礎中顯著且持續成長的一部分。此一向服務收入的轉移是關鍵的結構性變化，可降低對交易活動的依賴。

Services and interest revenue came in ahead of the $60 million to $70 million range we provided at our third quarter earnings call. This was driven primarily by stronger than expected card growth, with more than 116,000 new card sign-ups during the year in response to card edition launches such as the XRP card.

服務與利息收入高於我們在第三季財報電話會議所提供的 6,000 萬至 7,000 萬美元區間。主要原因是信用卡成長優於預期，全年新增超過 116,000 名新卡申辦，係受 XRP 卡等卡片版本推出所帶動。

We saw growth across several other services categories. Custodial fee revenue increased 25% year-over-year, driven by higher average crypto assets under custody. We also recognized approximately $4.8 million of advisory revenue related to services provided to a strategic customer, as well as $1.2 million from new on-chain offerings, including integrations and token listing services.

我們也看到其他多個服務類別的成長。託管費收入年增 25%，主要由託管加密資產的平均規模提升所帶動。此外，我們確認約 480 萬美元的顧問收入，係與向一名策略性客戶提供服務相關；另有 120 萬美元來自新的鏈上產品，包括整合服務與代幣上架服務。

As we continue expanding the platform, we see increasing opportunities to drive monetization across multiple services as users engage with additional products beyond trading. Total operating expenses for the full year were $525 million versus $308 million in 2024. The year-over-year increase was driven largely by three main things.

隨著我們持續擴展平台，我們看到越來越多機會，當用戶在交易以外使用更多產品時，可透過多項服務推動變現。全年總營運費用為 5.25 億美元，2024 年為 3.08 億美元。年增主要由三個因素所驅動。

First, stock-based compensation tied to the IPO, including the Q3 bonus accrual that settled in equity. Second, the significant marketing investments we made after going public to drive card growth. And third, continued spend in technology, compliance, and public company infrastructure costs.

第一，與 IPO 相關的股權激勵，包括第三季以股權結算的獎金應計。第二，我們在上市後為推動信用卡成長所進行的大幅行銷投資。第三，持續投入於科技、合規，以及上市公司基礎設施成本。

These investments were deliberate, and the restructuring actions we announced are designed to reset the company's cost structure going forward. Full-year adjusted EBITDA was a loss of $258 million, which is inclusive of $33.4 million of net realized and unrealized losses.

這些投資是經過審慎規劃的，而我們所宣布的重整措施旨在重設公司未來的成本結構。全年調整後 EBITDA 為虧損 2.58 億美元，其中包含 3,340 萬美元的已實現與未實現淨損失。

On a GAAP basis, full-year net loss was $582.8 million. It is important to note that a substantial portion of the net loss relates to non-cash items. These include $178.5 million of fair value losses on our prior related party instruments and mark-to-market adjustments on crypto assets, as well as $85 million of stock-based compensation expense associated with the equity awards issued in connection with our IPO. We believe that adjusted EBITDA is a useful way to look at the underlying performance of the business.

按 GAAP 基準，全年淨損失為 5.828 億美元。需要強調的是，淨損失中相當大一部分與非現金項目有關。其中包括 1.785 億美元的公允價值損失（來自我們先前的關係人金融工具）以及加密資產的按市價評價調整，另包括與 IPO 相關所發行股權獎勵的 8,500 萬美元股權激勵費用。我們認為，調整後 EBITDA 是觀察業務基本面表現的一種有用方式。

That said, our adjusted EBITDA result is not where we want it to be, and we've made decisions since year-end that are designed to change that. Now briefly on the balance sheet. We ended the year with approximately $252 million in cash and cash equivalents.

話雖如此，我們的調整後 EBITDA 結果仍未達到我們期望的水準，而我們已在年末之後做出一些決策，旨在改變這一點。接下來簡要談一下資產負債表。我們在年底時持有約 2.52 億美元的現金及約當現金。

The largest cash outflow in the quarter was the $117 million repayment of the Galaxy loan, which was completed in Q4 and removed that obligation from our balance sheet. As a result, we enter 2026 with a simpler balance sheet and lower debt levels. Following the restructuring actions announced earlier this year, we expect our normalized operating cash losses to decline meaningfully.

本季最大的現金流出是償還 Galaxy 貸款的 1.17 億美元，該償還已於第四季完成，並將該項義務自我們的資產負債表中移除。因此，我們以更簡化的資產負債表與更低的負債水準進入 2026 年。在今年稍早宣布的重組行動之後，我們預期正常化的營運現金損失將顯著下降。

Going forward, our focus is on continuing to narrow the gap to profitability through disciplined cost management and growth in higher margin services revenue. The card warehouse facility had $154.4 million outstanding at year-end against a $188 million in pledged receivables, supporting capacity of $250 million. As the receivables book grows, we'll execute additional funding capacity to support expected growth.

展望未來，我們的重點是透過嚴謹的成本管理，以及高毛利服務收入的成長，持續縮小與獲利之間的差距。截至年底，卡片倉儲融資工具未償餘額為 1.544 億美元，對應質押應收款為 1.88 億美元，支援的額度容量為 2.5 億美元。隨著應收款帳冊成長，我們將執行額外的融資容量，以支援預期的成長。

On restructuring costs, the $11 million in pretax charges associated with the Gemini 2.0 plan will land almost entirely in Q1 of 2026 and are expected to be cash charges. They cover the UK, EU, and Australia wind down and the headcount reductions. Timing on some of the international pieces will depend on local consultation requirements, but we expect the full plan to be substantially complete by mid-year. We expect these actions to simplify the organization and reduce our operating cost base going forward.

關於重組成本，與 Gemini 2.0 計畫相關的 1,100 萬美元稅前費用幾乎將全部落在 2026 年第一季，且預期為現金費用。這些費用涵蓋英國、歐盟與澳洲的收尾退出，以及人力縮減。部分國際項目的時程將取決於當地諮詢要求，但我們預期整體計畫將在年中前大致完成。我們預期這些行動將簡化組織，並在未來降低我們的營運成本基礎。

Before I turn to the full-year outlook, let me share what we are seeing so far in Q1 2026. Through February, trading volume was approximately $5.3 billion, down from Q4 levels as broader trading activity has continued to soften.

在我轉到全年展望之前，先分享我們目前在 2026 年第一季所看到的情況。截至 2 月，交易量約為 53 億美元，較第四季水準下滑，因為整體交易活動持續轉弱。

On the card, payment volume has exceeded $330 million with over 150,000 open card accounts. On predictions, approximately 15,000 users have traded since launch across more than 12,000 listed contracts. Total monthly transacting users across the platform were approximately 606,000. As always, we urge caution in extrapolating partial quarter activity.

在卡片方面，支付量已超過 3.3 億美元，開立中的卡片帳戶超過 150,000 個。在預測（Predictions）方面，自上線以來約有 15,000 名用戶進行交易，涵蓋超過 12,000 份已上架合約。平台每月總交易用戶約為 606,000 人。一如既往，我們提醒各位在外推部分季度的活動時需保持審慎。

With that context, let me turn to how we're thinking about fiscal year 2026. At this time, we are not providing total operating expense guidance for the year. With the restructured cost base still taking shape and a macro environment that is difficult to forecast, we think the more useful approach is to frame the key expense categories individually.

在此背景下，接下來談談我們對 2026 財年的看法。目前我們不提供全年總營運費用的指引。由於重組後的成本基礎仍在成形，加上宏觀環境難以預測，我們認為更有用的方式是分別就主要費用類別進行說明。

The restructuring actions we implemented earlier this year began flowing through the cost structure in Q2. Since year-end, we have reduced headcount by approximately 30% from peak levels. Because 2025 compensation reflected the full year at pre-restructuring staffing levels, the year-over-year decline is more moderate than the underlying headcount reductions.

我們今年稍早實施的重組行動自第二季開始反映在成本結構中。自年底以來，我們已將人力自高峰水準削減約 30%。由於 2025 年的薪酬反映了重組前人力配置水準的全年，因此年對年的下降幅度較底層的人力縮減幅度更為溫和。

We expect compensation excluding stock-based comp and restructuring charges to decline 15% to 20% relative to 2025. Stock-based compensation is expected to total $100 million to $115 million in 2026. 2025 included only two quarters of stock-based compensation at post-IPO levels following our September listing. The full-year figure is higher in absolute terms, but the quarterly run rate is stabilizing as the IPO-related grant cycle normalizes.

我們預期，扣除股票基礎薪酬與重組費用後的薪酬成本，將較 2025 年下降 15% 至 20%。2026 年的股票基礎薪酬預期合計為 1 億至 1.15 億美元。2025 年在我們 9 月上市後，僅包含兩個季度、按 IPO 後水準計提的股票基礎薪酬。全年數字在絕對金額上較高，但隨著 IPO 相關授予週期正常化，季度運行率正趨於穩定。

Technology and G&A is expected to range from $155 million to $190 million. The lower end reflects the post-restructuring normalized base. The width of the range reflects variable costs that scale with card and trading activity, and we plan to narrow this range as we gain visibility through the year. Marketing expenses, excluding rewards and promotions, are expected at 10% to 15% of revenue, depending on market conditions and the opportunities we see in our highest returning acquisition channels.

技術與一般及行政（G&A）費用預期介於 1.55 億至 1.9 億美元。區間下緣反映重組後的正常化基礎。區間寬度反映會隨卡片與交易活動而變動的可變成本；我們計畫在年度進行中、能見度提升後縮小此區間。行銷費用（不含回饋與促銷）預期為營收的 10% 至 15%，取決於市場狀況以及我們在最高投報率獲客渠道所看到的機會。

On the revenue side, our credit card product remains the principal engine for acquisition and growth. Predictions are still early, but with more than 15,000 users since December, we see early traction as encouraging, and it is central to where we are taking the company. While 2025 was the most expensive year in the company's history, given our IPO, the card investments, and international expansion, the actions we've taken since then are designed to ensure that 2026 looks very different financially.

在營收方面，我們的信用卡產品仍是獲客與成長的主要引擎。預測（Predictions）仍處於早期階段，但自 12 月以來已有超過 15,000 名用戶，我們認為初期動能令人鼓舞，且這是我們推動公司發展方向的核心。雖然 2025 年因 IPO、信用卡投資與國際擴張而成為公司史上成本最高的一年，但我們此後採取的行動旨在確保 2026 年在財務表現上將截然不同。

Overall, we believe that the organization we enter 2026 with is leaner, more focused, and positioned to drive improved operating leverage as we continue to scale our business. Together, we expect these dynamics to result in an improvement in adjusted EBITDA in 2026 as we operate with a more disciplined cost structure and a more diversified revenue base.

整體而言，我們相信進入 2026 年時的組織更精實、更聚焦，並在我們持續擴大業務規模的同時，具備推動營運槓桿改善的定位。綜合而言，我們預期在更具紀律的成本結構與更為多元的營收基礎下，這些動態將使 2026 年的調整後 EBITDA 改善。

To summarize, 2025 was a year of significant transformation for Gemini. We went public, scaled our credit card program, expanded and diversified revenue through services, launched prediction markets, and took decisive steps to reset our cost structure. We enter 2026 with a simpler organization, a lower expense base, and a more durable business model. We see the core story of Gemini today as straightforward. The business is becoming less dependent on crypto trading volumes and increasingly driven by recurring and diversified platform revenue.

總結來說，2025 年是 Gemini 顯著轉型的一年。我們完成上市、擴大信用卡計畫規模、透過服務擴張並多元化營收、推出預測市場，並果斷採取行動以重設成本結構。我們以更簡化的組織、更低的費用基礎，以及更具韌性的商業模式進入 2026 年。我們認為當前 Gemini 的核心故事很直接。公司對加密貨幣交易量的依賴正在降低，並日益由經常性且多元化的平台營收所驅動。

With that, we will now turn to questions. Thanks, everyone.

接下來，我們將進入問答環節。謝謝大家。

We will now take questions from our research analysts. Questions were submitted to us in writing, and we will take one question per analyst. Our first question comes from James Yaro at Goldman Sachs, who asks, Could you update us on the drivers of the recent executive departures and how this fits into your new strategy?

我們現在將接受研究分析師的提問。問題以書面方式提交，我們將每位分析師回答一題。第一個問題來自高盛（Goldman Sachs）的 James Yaro，他問：能否更新近期高階主管離職的驅動因素，以及這如何融入你們的新策略？

Thanks for this question. This summer was a different world, and when we IPO'd in September, the price of Bitcoin was about $115,000 per coin. Of course, the markets dropped significantly from that point in time. But in addition, our ability to build a super app in America with predictions, there's now a path forward for that. With the inflection point of AI, we have determined that we can move faster as a smaller, flatter, AI-enabled organization that is still, of course, very much founder-led. We think that we have the right team and the right organizational structure for today and tomorrow.

謝謝這個問題。今年夏天的情勢已大不相同，而我們在 9 月 IPO 時，比特幣價格約為每枚 115,000 美元。當然，市場自那時起大幅下跌。但此外，我們在美國打造包含預測功能的超級 App 的能力，如今也出現了可行的前進路徑。隨著 AI 的拐點到來，我們判斷以更小、更扁平、由 AI 賦能、且當然仍然高度由創辦人主導的組織，我們可以更快推進。我們認為我們擁有適合當下與未來的團隊，以及正確的組織架構。

Our next question comes from Matt Coad at Truist, who asks, You continue to see traction growing your user base despite the rough crypto market backdrop. What do you believe is driving this user growth, and how do you plan to cross-sell prediction markets into this large and growing user base?

下一個問題來自 Truist 的 Matt Coad，他問：儘管加密市場環境艱困，你們仍持續看到用戶基礎成長的動能。你們認為是什麼在驅動用戶成長？以及你們計畫如何將預測市場交叉銷售給這個龐大且持續成長的用戶群？

Thanks for the question. I think I can start here and then maybe kick it off to Cameron or Tyler to speak a little bit more on predictions. We're very pleased by the continued growth we see in our user base, particularly given the broader market backdrop. I think one of the key drivers here is we're continuing to see meaningful user acquisition through our credit card program, and just alongside broader engagement driven by new products that we're introducing and diversifying our revenue base, such as predictions. We'll hand it over to see if Cameron or Tyler want to touch on predictions a little bit more.

感謝提問。我想我可以先從這裡開始，然後也許把問題交給 Cameron 或 Tyler，請他們再多談一點關於預測（predictions）的看法。在更廣泛的市場背景之下，我們對用戶基礎持續成長感到非常滿意。我認為其中一個關鍵驅動因素是，我們持續透過信用卡計畫看到顯著的用戶獲取，同時也因為我們推出的新產品並使收入基礎更多元（例如預測），帶動更廣泛的參與度。我把時間交給 Cameron 或 Tyler，看看他們是否想再多談一點預測。

When we started in 2015, we were a Bitcoin company, and people came to us, and they could buy, sell, and store Bitcoin. Over time, we became a crypto company, and we added additional money verbs like staking, where users could stake their assets with us. Then we added the Gemini Credit Card, and that's become an active part of people's financial lives who want to earn crypto back every time they swipe. We're gonna continue to add things to our product where users have reasons to do more with us over time. Eventually, like, a number of these activities will continue to be independent of crypto cycles.

我們在 2015 年起步時是一家比特幣公司，人們來到我們這裡，可以買入、賣出並儲存比特幣。隨著時間推移，我們成為一家加密貨幣公司，並新增了更多「金錢動詞」（money verbs），例如質押（staking），用戶可以把資產在我們這裡進行質押。接著我們推出了 Gemini 信用卡，對於想要每次刷卡都能賺取加密貨幣回饋的人而言，這已成為其金融生活中活躍的一部分。我們將持續在產品上新增功能，讓用戶隨著時間推移有更多理由與我們進行更多互動。最終，像其中不少活動將會越來越不受加密市場週期影響而保持獨立運作。

I think that we're excited to see the engagement with prediction markets, our credit card, and other things that we're gonna bring to the Gemini app so that people don't have a reason to go elsewhere.

我想我們很期待看到大家對預測市場、我們的信用卡，以及我們即將帶到 Gemini App 的其他功能的參與度，讓人們沒有理由去別的地方。

The next question comes from Adam Frisch at Evercore, who asks, Can you help us frame the path to sustained positive standalone card economics, specifically the relative contributions from rewards optimization, lower acquisition costs, provision and credit normalization, and cheaper, broader funding capacity?

下一個問題來自 Evercore 的 Adam Frisch，他問：你們能否協助我們梳理實現信用卡業務獨立、可持續正向經濟效益的路徑，特別是獎勵最佳化、降低獲客成本、備抵與信用正常化，以及更便宜、更廣泛的資金來源能力，各自的相對貢獻？

I can take this one. Thanks for the question, Adam. We're really encouraged by the progress that we made in Q4, reaching near breakeven on the card. The card business has scaled really quickly, and we believe it has a clear path to profitability as the portfolio matures. There's a few primary levers that we think of. First, on the revenue side, we're seeing strong growth driven by interchange as spend increases.

這題我來回答。謝謝你的問題，Adam。我們對第四季取得的進展感到非常振奮，信用卡業務已接近損益兩平。信用卡業務擴張非常快，我們相信隨著資產組合成熟，將有清晰的獲利路徑。我們認為主要有幾個槓桿。首先在收入端，隨著刷卡消費增加，我們看到由交換手續費（interchange）帶動的強勁成長。

Then also important to note, interest income is still underearning relative to the size of the receivables base. As the portfolio seasons and matures, interest income becomes a meaningful tailwind. Then on the cost side, we have several levers really. Rewards are the largest expense today, but these were intentional and front-loaded to drive adoption and really establish the credit card. It worked.

另外也要注意，利息收入相對於應收帳款規模仍處於偏低、尚未充分貢獻的狀態。隨著資產組合經過時間沉澱並成熟，利息收入將成為重要的順風因素。在成本端，我們也有幾個槓桿。獎勵回饋目前是最大支出，但這是有意為之且前置投入，用來推動採用並真正建立信用卡產品。而且確實奏效。

We went from roughly 30,000 open accounts at the start of 2025 to now over 150,000 as of March 1. When you think about it, the Bitcoin card has only been out for about nine months, XRP for about five months. We're just getting started with this program. Rewards are really fully within our control, and we expect to optimize those over time.

我們從 2025 年初大約 30,000 個開立帳戶，成長到截至 3 月 1 日已超過 150,000 個。如果你想想看，比特幣卡推出至今大約只有九個月，XRP 卡大約五個月。我們這個計畫才剛起步。獎勵回饋完全在我們可控範圍內，我們預期會隨時間推移持續最佳化。

To add to that, we're also really been pleased with the organic sign-up traction on a smaller spend base. We're still averaging well north of 100 sign-ups a day, which is more than double where we were a year ago. We're also seeing improvements in bank fees as we scale which will reflect better underlying economics. From a credit perspective, the performance is trending in the right direction.

補充一下，我們也很滿意在較小的消費基礎上，仍能看到自然（organic）的註冊動能。我們目前平均每天仍遠高於 100 個註冊，較一年前的水準增加一倍以上。隨著規模擴大，我們也看到銀行費用有所改善，這將反映出更好的基礎經濟性。從信用角度來看，表現正朝正確方向發展。

We see loss rates as stabilizing and also continuing to improve as the book matures. Finally on funding. While funding costs are now coming into the model, we expect those to become more efficient as the portfolio grows and also as financing options expand. The expansion of the funding facility is really an important step, and longer term, we see opportunities to lower the cost of capital and diversify funding sources as the portfolio grows.

我們看到損失率正在趨於穩定，並且隨著資產組合成熟而持續改善。最後談資金面。雖然資金成本現在已納入模型，我們預期隨著資產組合成長以及融資選項擴大，資金效率會提升。擴大融資額度是一個非常重要的步驟；長期而言，隨著資產組合成長，我們也看到降低資本成本並分散資金來源的機會。

Putting it all together, we're already near breakeven on a pre-provision basis, and the path to sustained profitability is driven really by a combination of portfolio seasoning, cost optimization, and scale-driven efficiencies. We don't need one single lever to do all the work. It's really incremental improvements across each of these areas that we believe will drive the card business into consistent profitability.

綜合來看，我們在扣除備抵前（pre-provision basis）已接近損益兩平，而通往可持續獲利的路徑，主要由資產組合的時間沉澱（seasoning）、成本最佳化，以及規模帶來的效率提升共同驅動。我們不需要單一槓桿就能完成所有工作。我們相信，正是這些領域的逐步改善，將推動信用卡業務走向穩定且一致的獲利。

The next question comes from Michael Cyprys at Morgan Stanley. 15,000 users have used prediction markets through the end of February. How has that translated to revenue? Where do you see the growth potential from there? How do you compete versus peers that have a higher number of active users?

下一個問題來自 Morgan Stanley 的 Michael Cyprys。截至 2 月底，有 15,000 名用戶使用過預測市場。這如何轉化為收入？你們認為後續的成長潛力在哪裡？你們如何與活躍用戶數更高的同業競爭？

Thanks for this question, Michael. We will provide an update on revenue in the near future, but it's very early at this point. We are very encouraged with the fact that 15,000 customers have already engaged with this marketplace, which is brand new and we did not have even a quarter ago. We're very excited that our users are engaging with the product. We continue to grow that number on a daily basis and add many new contracts to the offering.

謝謝你的問題，Michael。我們會在不久的將來提供收入方面的更新，但目前仍非常早期。我們對於已有 15,000 名客戶參與這個全新市場感到非常振奮，這是我們在甚至一個季度前都還沒有的產品。我們非常興奮看到用戶正在使用這項產品。我們每天都在持續增加這個數字，並在產品中加入許多新的合約。

I think crypto is a great example. I think we started with monthly contracts. We are now down -- moved down to weekly, daily, hourly, 15-minute, and just offering all these different types of intervals and ways for people to hedge and trade around the price of crypto. We're just getting started.

我認為加密貨幣是一個很好的例子。我們一開始提供的是月度合約。現在我們已經往下延伸——改為提供週、日、每小時、15 分鐘等不同週期，並提供各種不同的區間與方式，讓人們可以針對加密貨幣價格進行避險與交易。我們才剛開始。

We're very encouraged. I think that looking at the market as a whole, it's also very early for this market, and we see the pie only growing from here. We think we're one of the few people who are building the full end-to-end marketplace for predictions, and we're excited that our customers, it's resonating with them.

我們非常受到鼓舞。我認為從整體市場來看，這個市場也仍處於非常早期階段，我們看到這塊「餅」只會從這裡開始持續變大。我們認為我們是少數正在打造端到端（end-to-end）預測市場的人之一，我們也很高興這項產品與我們的客戶產生共鳴。

Great. Just to add on to that, we've been building technology trading systems in marketplaces for well over a decade. These are the kind of things that we know how to do very well. We have a website, we have a mobile app, we have API interfaces, and we've been doing market surveillance. We know how to onboard customers, KYC them, and build great trading and marketplace experiences.

很好。再補充一下，我們在市場型態的交易系統技術方面已經建置超過十年。這些正是我們非常擅長的事情。我們有網站、有行動 App、有 API 介面，也一直在做市場監控（market surveillance）。我們知道如何導入客戶、完成 KYC，並打造優秀的交易與市場體驗。

This is very much an extension of the over a decade of experience and expertise that we've developed over the years.

這在很大程度上是我們多年來累積的十多年經驗與專業能力的延伸。

The next question comes from John Todaro at Needham. How are you thinking about capital raising and liquidity if we assume crypto volumes remain lower than 2025 levels through 2026 and 2027?

下一個問題來自 Needham 的 John Todaro。如果我們假設加密貨幣交易量在 2026 與 2027 年都維持低於 2025 年的水準，你們如何看待募資與流動性管理？

Thanks for the question, John. We really appreciate it, and we're planning the business with a conservative set of assumptions, which include a scenario where volumes remain below 25 levels through 2026 and 2027 as well. From a liquidity standpoint, we've taken really meaningful steps to reduce our cost base and improve cash efficiency. Really, we're focused on the scaling a more durable recurring revenue streams that are less dependent on trading volumes.

謝謝你的提問，John。我們非常感謝，我們也正以一套保守的假設來規劃業務，其中包括一種情境：交易量在2026年與2027年也都維持在低於25年水準。從流動性角度來看，我們已採取非常有意義的措施來降低成本基礎並提升現金效率。我們真正聚焦的是擴大更具韌性的經常性收入來源，讓其對交易量的依賴更低。

Our main focus is to execute on our operating plan with that discipline in mind. With that said, we're always evaluating opportunities to strengthen our balance sheet and support sustainable growth. If there are opportunities for this on attractive terms, we would consider them.

我們的主要重點是以這樣的紀律來落實營運計畫。話雖如此，我們也一直在評估強化資產負債表並支持可持續成長的機會。如果有條件具吸引力的機會，我們會加以考慮。

We're first and foremost focused on demonstrating the operating improvements and letting the results really create the conditions for any future transaction or capital raise. The key point is that we aim to build a model that can sustain itself across cycles and not one that depends on near-term recovery and volumes.

我們首先且最重要的是專注於展現營運改善，並讓成果真正為任何未來交易或募資創造條件。關鍵在於，我們的目標是建立一個能跨週期自我維持的模式，而不是依賴短期復甦與交易量的模式。

Yeah. Look, as founders, we've been building Gemini for over a decade. We don't just have our skin in the game, we have our entire bodies in the game. We're deeply committed to Gemini and the mission and very excited to continue building it and as we expand the mission into the super app. I think one of the things that we've talked about is that that really helps us break free of the crypto cycles and give customers things that they can do throughout their daily financial lives, whether it's using a credit card or trading predictions.

是的。你看，作為創辦人，我們打造Gemini已經超過十年。我們不只是把利益押在這裡，我們是把整個人都投入其中。我們對Gemini與其使命有深度承諾，也非常期待持續建設，並在我們把使命擴展到超級App時繼續前進。我想我們談過的一點是：這確實能幫助我們擺脫加密週期，並在客戶的日常金融生活中提供可使用的功能，不論是使用信用卡或交易預測。

We're hoping to launch US equities as well, investing in US capital markets and really building out a more durable story of revenue and engagement that moves beyond simply buy, sell, store, or say crypto, which is obviously very core to the business. We wanna build on that and give our customers more reasons to use Gemini.

我們也希望推出美國股票，投資美國資本市場，並真正打造一個更具韌性的收入與互動故事，超越單純的買、賣、存放，或說加密資產——這當然仍是業務的核心。我們想在此基礎上再往前，給客戶更多使用Gemini的理由。

We're seeing the beginnings of that, and I think we're really excited to keep doing that. Even if crypto prices do remain depressed for some prolonged period of time, we will be building other products that continue to drive engagement and growth of our business.

我們已經看到一些初步跡象，我想我們也非常興奮能持續推進。即使加密價格在一段較長時間內仍然低迷，我們也會打造其他產品，持續帶動互動與業務成長。

The next question comes from Peter Christiansen at Citi. What is Gemini's OpEx discipline going forward, and has management put in place guardrails that helps ensure eventual profitability at the EBITDA level?

下一個問題來自花旗（Citi）的Peter Christiansen。Gemini未來在營運費用（OpEx）上的紀律是什麼？管理層是否已設置護欄，以確保最終能在EBITDA層面實現獲利？

Thanks for the question, Pete. OpEx discipline is a core focus for us coming out of the restructuring. We've reset the business to a lower fixed cost base, and we've put clear guardrails in place around any incremental spend. That includes being very selective on headcount growth and tying it directly to revenue or strategic priorities, and also continuing to manage marketing as a variable lever really based on ROI and market conditions.

謝謝你的提問，Pete。在重組之後，營運費用（OpEx）紀律是我們的核心重點。我們已將業務重置到較低的固定成本基礎，並對任何新增支出設立了明確的護欄。這包括在人力擴張上非常審慎，並將其直接與收入或策略優先事項掛鉤；同時也持續把行銷作為可變的槓桿，真正依據投資報酬率（ROI）與市場狀況來管理。

That's a real lever that we can dial up or down, depending on market conditions, requiring clear payback thresholds for any new investments. Just as importantly, we've become much more focused as an organization. Prioritizing a smaller set of high-impact initiatives and exiting or scaling back areas that just didn't meet our return thresholds.

這是一個我們可以依市場狀況上調或下調的真正槓桿，並對任何新投資要求明確的回收門檻。同樣重要的是，我們作為一個組織已變得更聚焦。我們優先推動較少數的高影響力計畫，並退出或縮減那些未達到我們報酬門檻的領域。

That really allows us to concentrate our resources and our capital where we have the strongest product market fit and demand. We believe that the organization is now structured to drive really operating leverage as volumes and engagement recovers. What's important to add is we don't need to meaningfully re-expand the cost base to achieve our growth target.

這讓我們能把資源與資本集中在我們擁有最強產品市場契合度與需求的地方。我們相信，當交易量與互動回升時，組織目前的架構已能真正帶來營運槓桿。另外重要的一點是：為了達成成長目標，我們不需要大幅重新擴張成本基礎。

A lot of the growth from here really comes from just better monetization of our existing user base and also just leveraging the infrastructure that we've already built. I'd say the right way to think about this is we have a relatively stable OpEx base coming out of the restructuring with modest or highly targeted investments layered on top rather than us returning to a broad-based spending.

從現在開始的許多成長，實際上來自於更好地變現我們既有的用戶基礎，以及運用我們已經建好的基礎設施。我會說，正確的理解方式是：重組後我們有相對穩定的OpEx基礎，上面再疊加適度或高度精準的投資，而不是回到全面性支出。

If 2025 was the year of investment, I'd say 2026 is really the year of focus and discipline.

如果說2025年是投資之年，我會說2026年真正是聚焦與紀律之年。

The final question comes from Dan Dolev at Mizuho. Given the regulatory and competitive landscape in crypto and prediction markets, what are the biggest external risks you're managing against in 2026, and what would you point to as your most underappreciated competitive advantage?

最後一個問題來自瑞穗（Mizuho）的Dan Dolev。考量加密與預測市場的監管與競爭格局，你們在2026年正在管理的最大外部風險是什麼？以及你們認為最被低估的競爭優勢是什麼？

Thanks for the question, Dan. I think one of the things that we wanna talk about is the fact that obviously there's a lot of effort to pass a crypto market structure bill, and I think what is very encouraging to see is the SEC and the CFTC in parallel are doing great work to bring about the super app era independent of a bill. While we are hopeful that a good bill will ultimately get passed, there is a lot of great work going on at both agencies to create a path for super apps in the event that a bill does not pass for whatever reason. We believe, like, the future for crypto in America has never been brighter.

謝謝你的提問，Dan。我想我們想談的一點是：顯然目前有很多努力在推動通過加密市場結構法案；而令人非常鼓舞的是，SEC與CFTC也在同步做出很棒的工作，即使不依賴法案，也能推動超級App時代的到來。雖然我們希望最終能通過一部好的法案，但即便因任何原因法案未能通過，兩個機構也正在做大量出色的工作，為超級App建立一條可行的路徑。我們相信，美國加密的未來從未如此光明。

I think there is a lot of great work being done that we're excited about. I think the second point that I'd like to make is that we are one of the, I think, the few end-to-end prediction market marketplaces that also has a crypto marketplace within the same organization.

我認為有很多令人振奮的工作正在進行，我們也很期待。第二點我想說的是：我們是少數幾家端到端的預測市場交易平台之一，同時在同一個組織內也擁有加密交易市場。

We believe there's a lot of synergies for people who want to trade, for example, a Bitcoin event contract, but also be able to trade spot Bitcoin within the same place, and hopefully eventually perpetual futures down the road, and US equities. I think that being an end-to-end marketplace for both predictions and spot as opposed to plugging into another marketplace, we believe that's an advantage for us going forward.

我們相信，對於想交易的人而言存在許多協同效應，例如想交易比特幣事件合約的人，也能在同一個地方交易比特幣現貨，並希望未來也能交易永續期貨，以及美國股票。我認為，作為同時涵蓋預測與現貨的端到端市場，而不是接入另一個市場，我們相信這將是我們未來的優勢。

At this time, there are no more questions. Thank you all for listening, and we'll talk to you soon.

目前沒有更多問題。感謝各位收聽，我們很快再與各位交流。

That concludes today's conference call. You may now disconnect.

今天的電話會議到此結束。您現在可以掛線。