Opendoor Technologies Inc. (OPEN) 2026 Q3 法說會逐字稿

(video playing)

(影片播放中)

Hey, everyone. Welcome to Opendoor's 3Q 2026 Financial Open House earnings livestream. I'm Michael Judd, Opendoor's Health Investor Relations. A few housekeeping items before we get started. Details of our results and additional management commentary are available in our earnings release, which can be found at investor.opendoor.com. The following discussion contains forward-looking statements within the meaning of the federal securities laws.

大家好。歡迎收看 Opendoor 2026 年第三季（3Q 2026）「Financial Open House」財報直播。我是 Michael Judd，負責 Opendoor 的健康投資人關係。在開始之前先說幾項例行事項。我們的業績細節與管理層的補充評論已發布於財報新聞稿，可在 investor.opendoor.com 查閱。以下討論包含符合聯邦證券法定義的前瞻性陳述。

All statements other than statements of historical fact are statements that could be deemed forward-looking, including but not limited to statements regarding Opendoor's financial condition, anticipated financial performance, business strategy and plans, market opportunity and expansion, and management objectives for future operations. These statements are neither promises nor guarantees and undue reliance should not be placed on them.

除歷史事實陳述外，所有陳述皆可能被視為前瞻性陳述，包括但不限於關於 Opendoor 的財務狀況、預期財務表現、業務策略與計畫、市場機會與擴張，以及管理層對未來營運的目標等。這些陳述既非承諾亦非保證，不應對其過度依賴。

Forward-looking statements involve risks and uncertainties that may cause actual results to differ materially from those discussed here. Initial information that could cause actual results to differ from forward-looking statements can be found in the Risk Factors section of Opendoor's most recent annual report on Form 10-K for the year ended December 31, 2025, as updated by our periodic reports and other filings with the SEC. Any forward-looking statements made on this webcast, including responses to your questions, are based on management's reasonable current expectations and assumptions as of today. And Opendoor assumes no obligations to update or revise them, whether as a result of new information, future events, or otherwise, except as required by law.

前瞻性陳述涉及風險與不確定性，可能導致實際結果與此處討論內容存在重大差異。可能導致實際結果與前瞻性陳述不同的初步資訊，可見於 Opendoor 最近一期截至 2025 年 12 月 31 日止年度的 Form 10-K 年報之「風險因素」章節，並以我們定期報告及向 SEC 提交的其他文件所更新者為準。本次網路直播中所作的任何前瞻性陳述（包括對各位提問的回覆），均基於管理層截至今日合理的當前預期與假設。除法律要求外，Opendoor 不承擔因新資訊、未來事件或其他原因而更新或修訂該等陳述的任何義務。

The following discussion contains references to certain non-GAAP financial measures. The company believes these non-GAAP financial measures are useful to investors as supplemental operational measurements to evaluate the company's financial performance. For a reconciliation of each of these non-GAAP financial measures to the most directly comparable GAAP metric, please see our website at investor.Opendoor.com.

以下討論將提及若干非 GAAP 財務衡量指標。公司認為，這些非 GAAP 財務衡量指標可作為補充性的營運衡量，協助投資人評估公司的財務表現。如需各項非 GAAP 財務衡量指標與最直接可比之 GAAP 指標的調節表，請至 investor.Opendoor.com 查閱。

And with that, let's get into the open house with Kaz and Christy.

那麼，接下來就請 Kaz 和 Christy 帶我們進入 open house。

Hey, everyone. I've been doing this job for just over a year. And like all anniversaries, I thought it'd be a good idea to take a second and look at what we've done in the past year. A year ago, I told you we would build a different Opendoor. Faster, stronger, fully committed to our mission of tilting the world in favor of homeowners.

大家好。我做這份工作剛滿一年多一點。就像所有週年紀念一樣，我覺得花點時間回顧過去一年我們做了什麼，是個好主意。一年前，我告訴各位我們會打造一個不同的 Opendoor。更快、更強，並且全心投入我們的使命：讓世界更偏向屋主。

Let's see what's changed. Today, our market share of U.S. real estate is up more than four times. We're now over half a percent of the market. We signed over 7,000 contracts in Q3. A year ago, that was just a little over 1,500.

我們來看看有哪些改變。今天，我們在美國房地產市場的市占率已提升超過四倍。我們現在已超過市場的 0.5%。我們在第三季簽下超過 7,000 份合約；一年前只有略高於 1,500 份。

In Q3, we shipped more than 500 new features and products across our stack. We had shipped almost nothing that was useful in the weeks before I started. This time last year, we didn't have a single mortgage license. Today, we have 25 state licenses and our mortgage origination volume is up 5x quarter over quarter, and October alone was over half of Q3.

第三季，我們在整個技術堆疊上推出了超過 500 項新功能與產品。在我上任前的幾週，我們幾乎沒有推出任何有用的東西。去年這個時候，我們連一張房貸執照都沒有。今天，我們已取得 25 個州的執照，房貸承作量（origination volume）季對季成長 5 倍，而光是 10 月就已超過第三季的一半。

Over the last 12 months, we had over $1 billion in Cash Now, More Later volume. This is $1 billion more than we had done in the previous 12 months. And even with the listings at unprecedented highs, we're selling homes faster than the market every single week.

過去 12 個月，我們的「Cash Now, More Later」交易量超過 10 億美元。這比再往前 12 個月多了 10 億美元。即使市場待售物件（listings）處於前所未見的高位，我們每一週賣房的速度都比市場更快。

Look, we should talk a bit about the market because if I don't talk about the market now, every follow-up question will be about the macro. It's not going to be a shock to anyone watching this that the market has turned, right? 2025 was the worst year housing had had in a long time and 2026 was tracking right on top of it through June.

我們得稍微談一下市場，因為如果我現在不談市場，後續每個問題都會是關於總體環境。對任何正在收看的人來說，市場已經轉向這件事應該不會令人意外，對吧？2025 年是房市很長一段時間以來最糟的一年，而 2026 年截至 6 月的走勢也幾乎與之重疊。

And honestly, we would have been just fine with that. But then in Q3, clearance just started to collapse and it hasn't come back. In September, pending home sales index fell to the lowest level it has ever seen since we started recording it, right?

老實說，我們本來對那樣的情況也能應付得很好。但到了第三季，去化（clearance）開始崩跌，而且一直沒有回來。9 月份，成屋待售指數（pending home sales index）跌到自我們開始記錄以來的最低水準，對吧？

It's lower than it was in April 2020 when going out to two homes was basically illegal. And mortgage rates are now above 7%. That's more than a full point higher than it was a year ago. And everyone in housing got hit our job is to build a business that can perform better in difficult conditions. And we're doing this in three important ways.

它甚至比 2020 年 4 月還低，當時外出看兩間房基本上都算違法。而房貸利率現在已高於 7%。這比一年前高出超過整整 1 個百分點。房地產產業的每個人都受到衝擊；我們的工作是打造一個在艱困環境下也能表現更好的企業。我們透過三個重要方式來做到這點。

First, a capital light volume right this past week around half of the first offers are D2C seller accepts. We're listing but Cash Now, More Later that lets us serve more sellers with less of our capital exposed to changes in the market. Second, velocity. The less time we own a home, the less time the market has opportunity to move against us. So we price homes to sell and focus on reducing the time between acquisition and resale.

第一，輕資本的交易量。就在上週，大約一半的首次出價是 D2C（直面消費者）賣方接受。我們以「Cash Now, More Later」的方式上架，讓我們在更少資本暴露於市場變動的情況下，服務更多賣方。第二，速度（velocity）。我們持有房屋的時間越短，市場對我們不利變動的機會就越少。因此我們以可售出的價格定價，並專注於縮短從收購到轉售之間的時間。

Third, mortgage. When buyers are hard to find, better financing gives them a reason to choose to work with Opendoor. That's why we launched Buy with Opendoor in beta this week. We need to grow, use less capital per transaction and sell homes faster.

第三，房貸。當買方難以尋找時，更好的融資方案能給他們一個選擇與 Opendoor 合作的理由。這就是為什麼我們本週以 beta 形式推出「Buy with Opendoor」。我們需要成長、在每筆交易中使用更少資本，並更快賣出房屋。

Those are things we can control and that's what we're focused on. What matters to us is solving the core problems of our business and those are what we need to focus on. Look, for most of its life, Opendoor has had basically three problems, right?

這些是我們能掌控的事情，而這正是我們專注的方向。對我們而言，重要的是解決業務的核心問題，而那才是我們需要聚焦的重點。你看，Opendoor 在其大部分發展歷程中，基本上有三個問題，對吧？

To grow, we had to overpay sellers for their home. To grow, we had to own more homes and take on more risk on our entire inventory. And to grow, we had to hire more and more people. That is a hard business to scale.

為了成長，我們必須以更高價格向賣方收購房屋。為了成長，我們必須持有更多房屋，並在整體庫存上承擔更多風險。為了成長，我們必須雇用越來越多的人。這是一門很難擴張的生意。

A year ago, I said we would change this and I gave you three objectives, right? One for each of these problems, we've made real progress on these objectives. And we're in a really good place on the first and third and we're making progress on the second. Let's go through each of these objectives.

一年前，我說我們會改變這一點，並提出三個目標，對吧？每個問題各對應一個目標；我們在這些目標上已取得實質進展。第一與第三項我們處於非常好的位置，第二項也正在推進。我們逐一來看這些目標。

Okay, objective one was scale acquisitions, right? The first problem was that for Opendoor growth always meant paying more for homes. To understand our core business opportunity, you need first to understand two numbers, just two simple numbers. First. In our core markets, nearly one in four Americans who go to sell their home first try to sell it to Opendoor. That's the first number.

好，第一個目標是擴大收購規模（scale acquisitions），對吧？第一個問題是：對 Opendoor 而言，成長總是意味著要用更高的價格買房。要理解我們核心業務的機會，你首先需要理解兩個數字，就兩個簡單的數字。第一。在我們的核心市場中，幾乎每四位準備賣房的美國人，就有一位會先嘗試把房子賣給 Opendoor。這是第一個數字。

Second, one in three Americans cry when trying to sell their home the traditional way. Our problem has never been that we don't have enough demand or that people love crying. Our problem has always been that we have not. Been able to serve them in a good way, right?

第二，有三分之一的美國人在用傳統方式賣房時會哭。我們的問題從來不是需求不夠，或是人們喜歡哭。我們的問題一直是我們沒有。能夠以良好的方式服務他們，對吧？

So have we scaled acquisitions? The answer is for sure, yes, right? Now I want to be careful because the easy and lazy story here is Opendoor took on more risks and bought more homes and then the market turned. That's not what happened, right? And the evidence is obvious when you look at our true seller conversion.

那麼，我們是否擴大了收購規模？答案當然是：有，對吧？不過我想謹慎一點，因為這裡最簡單、也最偷懶的說法是：Opendoor 承擔了更多風險、買了更多房子，然後市場就轉向了。事情並不是那樣發生的，對吧？當你看我們真正的賣方轉換率（true seller conversion）時，證據就很明顯。

I want you to look at this chart. It takes homeowners who got an offer from us and within 90 days either sold to us or listed on the open market and measures how many chose Opendoor at the same period over the same spread. The blue line is this year, the gray lines are 2024 and 2025.

我希望你看看這張圖表。它以曾收到我們報價的屋主為對象，觀察他們在 90 天內不是賣給我們、就是在公開市場掛牌，並衡量在相同期間、相同價差區間下，有多少人選擇了 Opendoor。藍線代表今年，灰線代表 2024 年與 2025 年。

What this shows as we're winning more of these sellers at nearly every spread level. And we're literally doing three times better in some of these levels. We're winning the market without having to make riskier offers. So how are we doing this? And the answer is we've just built better products.

這顯示我們在幾乎每一個價差區間都贏得了更多這類賣方。而且在其中一些區間，我們的表現確實是三倍之多。我們不需要提出更高風險的報價，也能贏得市場。那我們是怎麼做到的？答案是：我們只是打造了更好的產品。

We have built more ways to meet sellers' needs and have made it easier for them to understand their options and. Make the choice they want for their own lives, right? Sellers can now adjust the offer sliders and make the choice that is good for their families.

我們建立了更多方式來滿足賣方需求，也讓他們更容易理解自己的選項，並且做出符合他們自身生活的選擇，對吧？賣方現在可以調整報價滑桿，做出對家人最有利的選擇。

Look, not that long ago, around 80% of the assessments we did on homes required an Opendoor employee and this could take many days. Today, that 80% is 2%. 98% of the time. You can ask for an offer Monday morning, complete the assessment before noon, and sign a contract before breakfast Tuesday.

你看，不久之前，我們對房屋所做的評估中，大約有 80% 需要 Opendoor 員工到場，這可能要花上好幾天。今天，那個 80% 已經變成 2%。有 98% 的情況下，你可以在週一早上要求報價，中午前完成評估，並在週二早餐前簽下合約。

We help sellers understand our options, we help them move fast, and sometimes we tell them we're not the best option for you, and that builds trust. Look, these things don't mean that we price every home perfectly, but by building the product the right way. We have managed to do two things that usually don't go hand in hand. We've increased growth and lowered risk.

我們幫助賣方理解我們的選項，幫助他們快速推進；有時我們也會告訴他們：我們不是最適合你的選擇，而這會建立信任。你看，這些並不代表我們能把每一間房都定價得完美無缺，但透過以正確方式打造產品，我們成功做到兩件通常無法兼得的事。我們提升了成長，同時降低了風險。

Every other time this company has scaled, it has done so by paying more for houses. This time it didn't. We proved that we can grow without buying growth. That's the single most important thing that has happened over the last year.

過去每一次這家公司擴張，都是靠用更高的價格買房來做到的。但這一次不是。我們證明了：我們可以在不「買來成長」的情況下成長。這是過去一年發生的最重要的一件事。

Our objective one is solid green. Okay, the second problem was that growth meant owning more homes and carrying all the risk, right? We're changing that with our capital-light products. But when we do own a home, we still have to renovate it, price it, and sell it, and we have to do it fast.

我們的目標一是穩穩的綠燈。好，第二個問題是：成長意味著要持有更多房屋、承擔所有風險，對吧？我們正透過資本輕量化產品來改變這點。但當我們確實持有房屋時，我們仍然必須翻新、定價並出售，而且必須做得快。

On objective two, which was like a unit economics and resale velocity, I think we're yellow. We have some work left to do, but things are much better than they were a year ago. We're not as good as we want to be, but we're in a much better place. Look, when the market moves against you and you own homes, there's two things you can do, right?

在目標二（有點像單位經濟與轉售速度）上，我認為我們是黃燈。我們還有一些工作要做，但情況比一年前好很多。我們還沒達到想要的水準，但我們的位置已經好很多。你看，當市場走勢對你不利、而你又持有房屋時，你有兩件事可以做，對吧？

You can sell fast or you can hold and wait and pray for the market to recover. We choose to clear every single time. The decline in late August and September knocked us off track on execution a little bit because we turned our attention to clearance rather than margin improvement. It costs us something real in a quarter like this, but that's the right way to run this company for the long run.

你可以快速賣出，或是持有、等待，並祈禱市場回升。我們每一次都選擇清庫存。8 月下旬與 9 月的下跌，讓我們在執行上稍微偏離軌道，因為我們把注意力轉向清庫存，而不是改善毛利。在像這樣的一個季度，這會讓我們付出真實代價，但從長期來看，這才是經營這家公司的正確方式。

You don't focus on a quarter, you focus on the mission. The economics that come out of this are not great. They're not what we want, but they're better than they would be otherwise because we're currently scaling Opendoor.

你不該只盯著一個季度，你要專注在使命上。由此產生的經濟結果並不理想。它們不是我們想要的，但比起其他做法會更好，因為我們目前正在擴張 Opendoor。

And we're learning about all the parts of the company that are not built for scale yet, right?

而且我們也在學習公司裡所有那些尚未為規模化而打造的部分，對吧？

We're fixing these things properly as we come across them rather than putting duct tape over cracks in the foundation because what we want to do. Is build a generational company and not one that just looks good in the short term.

我們在遇到這些問題時，會把它們正確地修好，而不是在地基裂縫上貼膠帶，因為我們想做的事是打造一家能傳承世代的公司，而不是一家短期看起來漂亮的公司。

Let me give you a couple of examples because these are things that if we had not changed, no one would have noticed and the numbers may have looked better for the quarter but would have been way worse for the long run, right? We fixed things that cost us margin this quarter but are incredibly important for the long run. Let's first talk about renovation.

我給你幾個例子，因為如果我們沒有改變這些事，沒有人會注意到，這一季的數字可能看起來更好，但長期會糟得多，對吧？我們修正了一些在本季會犧牲毛利、但對長期極其重要的事情。我們先談翻新。

Every real estate company has a manual, usually a physical manual that they renovate or build homes to as well. It was called the Universal Repair Standard. Opendoor had one of these until recently.

每一家房地產公司都有一本手冊，通常是一本實體手冊，用來作為翻新或建造房屋的標準。它叫做「通用修繕標準」（Universal Repair Standard）。直到最近，Opendoor 也有這樣一本。

The reason why all homes basically end up looking the same, right? You look at home after you acquired it and then you apply some universal set of rules to that home. So a home in Raleigh gets the same renovation as a home in Orlando.

這就是為什麼所有房子最後基本上看起來都一樣，對吧？你在收購房屋後看一眼，然後把一套通用規則套用到那間房子上。所以羅利（Raleigh）的一間房，會做出和奧蘭多（Orlando）一間房一樣的翻新。

And this all requires lots of humans doing lots of humans things. And if you don't have lots of humans doing those things, like it all just breaks. This past quarter, we rebuilt this entire system at Opendoor.

而這一切都需要大量的人去做大量「人才能做的事」。如果沒有足夠的人去做這些事，整個流程就會崩壞。在上一季，我們在 Opendoor 重建了整套系統。

We built an AI system that determines a specific repair plan for every single home before we buy it. Machines then draft a renovation scope and hand it to people doing renovation.

我們建立了一套 AI 系統，在我們買下每一間房之前，就為每一間房決定一個專屬的修繕計畫。接著由機器擬定翻新範圍（scope），再交給負責翻新的團隊執行。

This was a lot of work and because models don't behave perfectly as you build them. The first few homes that hit the market under this model were not built the way we would like them, right? For the last few weeks, this system has been outperforming.

這需要大量工作，而且模型在建置過程中不會完美運作。在這個模型下，最早幾間上市的房子並沒有按照我們希望的方式完成，對吧？但在過去幾週，這套系統的表現已經超越了。

The old process, just amazingly. Another example is resale, right? Every single real estate company ends up with humans managing resale of homes. If they're exceptional, one person can handle somewhere between 100 and 150 homes at a time.

舊流程，簡直是驚人地超越。另一個例子是轉售，對吧？每一家房地產公司最後都會由人來管理房屋的轉售。如果他們非常優秀，一個人同時大概能處理 100 到 150 間房。

That means that as we rescaled our acquisitions by 4x, we would have had to increase the number of people managing these acquisitions by 4x. We have done the exact opposite look until August our resale process was similar to literally everyone else in real estate we thought we were better at it but pricing every home we owned basically lived in like a glorified spreadsheet with.

這表示當我們把收購規模重新擴大到 4 倍時，我們就必須把管理這些收購的員工數也增加到 4 倍。我們做的正好相反。你看，直到 8 月，我們的轉售流程其實跟房地產業幾乎所有人都差不多；我們以為自己更擅長，但我們對每一間持有房屋的定價，基本上都放在一個像是加強版試算表的東西裡，裡面有。

70 something columns, right? So every Tuesday a portfolio manager would scroll through like 3,000 rows, pulling information from different tools and deciding what price to drop and what price to keep and how much. And then somebody would type this into a spreadsheet and into a separate system. And then like, it was all like just very weird, but this was actually state-of-the-art for the industry up until August, but that doesn't scale.

大概 70 多欄，對吧？所以每個星期二，一位投資組合經理會像這樣滑過大約 3,000 列，從不同工具拉資料，決定哪些要降價、哪些要維持、以及降多少。然後再由某個人把這些輸入到試算表，以及另一個獨立系統。然後就像……整件事都很怪，但直到 8 月為止，這其實是業界最先進的做法；但它無法規模化。

Like scaling this means adding more people and adding more errors and then adding more people to catch those errors, right? We replaced the entire thing, the single system, where a model recommends an action right in a tool that looks at comps and offer history and then the people keep the judgment, but software does the work.

因為要規模化就意味著要增加更多人、增加更多錯誤，然後再增加更多人去抓那些錯誤，對吧？我們把整套流程換掉了，改成單一系統：模型會在一個工具裡直接建議行動，該工具會看可比成交（comps）與報價歷史；人保留判斷，但工作由軟體完成。

This works amazingly well today. 12 weeks ago, it didn't exist. Mortgage, that's another example, right? Licensing here has taken us longer than we wanted. We can't rush governments, but we're making really good progress.

這套系統今天運作得非常好。12 週前，它還不存在。房貸也是另一個例子，對吧？我們在這裡的牌照申請花的時間比我們希望的更久。我們無法催促政府，但我們正在取得非常好的進展。

Today, when a buyer finances a home through Opendoor and closes it quickly, we can improve our contribution margin on that home significantly.

今天，當買方透過 Opendoor 為房屋辦理融資並快速完成交割時，我們可以顯著提升該房屋的貢獻毛利。

And through Buy with Opendoor, which we just launched in beta this week, we can earn mortgage revenue. Even on homes that we never own. Let's put this all together. Improving resale velocity isn't that complicated.

而透過我們本週剛以 Beta 版推出的 Buy with Opendoor，我們還能賺取房貸收入。即使是在我們從未持有過的房屋上也是如此。把這些整合起來看。提升轉售速度其實沒那麼複雜。

You make the homes better, you get better pricing them, you renovate them quickly, and then you buy something people actually want. And then mortgage makes it something they can afford. Better pricing gets the right buyer to act faster. The proof will be in how fast these homes sell.

你把房子變得更好、把定價做得更準、快速完成翻新，然後去買人們真正想要的房子。接著，房貸讓它變成他們負擔得起的選項。更好的定價能讓合適的買方更快採取行動。證明就在於這些房子賣得有多快。

And what we make on them when they do so. Look, I don't think we're there yet. We're much better off than we were a year ago, but I think this objective is yellow, and we have work to do. Okay, the third problem. The third problem was that growth meant hiring more people.

以及它們賣出時我們能賺多少。坦白說，我不認為我們已經做到位了。我們比一年前好得多，但我認為這個目標目前是黃色警示，我們還有工作要做。好，第三個問題。第三個問題是，成長意味著要雇用更多人。

And here, like we're clearly green. We're doing exceptionally well and I think we're doing better than any public company I know of. Here's what this looks like. On a trailing-12-month basis, our operations expense per acquired home is down 36% and fixed operations expense per acquired home is down 34%. Opendoor was founded under this idea that you could take a real estate transaction apart, put it together again in a much better way, right? That thesis still holds, but AI fundamentally changes the math.

而在這一點上，我們顯然是綠燈。我們表現得非常出色，我認為比我所知道的任何上市公司都做得更好。具體是這樣。以過去 12 個月（TTM）計算，我們每取得一戶房屋的營運費用下降了 36%，每取得一戶房屋的固定營運費用下降了 34%。Opendoor 的創立理念是：你可以把一筆房地產交易拆解開來，再用更好的方式重新組裝，對吧？這個論點仍然成立，但 AI 從根本上改變了這筆帳的計算方式。

I want to talk about three things here. First, automated intelligence. Look at this chart. Over the past year, our pricing teams throughput per person is up about 2.5 times. Rather than having people spending their days looking at pictures and doing mundane tasks, our people now each manage an army of agents. In July, we had 5,400 homes. That's 5,400 sellers sitting in a queue waiting for days for a human to confirm the price of their home.

我想在這裡談三件事。第一，智慧自動化。看看這張圖。過去一年，我們定價團隊的人均產出約提升了 2.5 倍。與其讓人們整天看照片、做瑣碎工作，我們的員工現在每個人都在管理一支代理人軍團。在 7 月，我們有 5,400 間房屋。那代表有 5,400 位賣方在佇列中等待，得等上好幾天，才會有真人確認他們房屋的價格。

Two weeks ago, that queue was zero. Zero and it will remain zero. As far as I can tell, that's the first time in Opendoor's history where we can say that the queue is zero and we expect it to remain zero. Look, this is not unique to pricing. Last year, the most efficient Opendoor employee could turn on a few dozen utilities in a week on homes we had bought.

兩週前，那個佇列是零。零，而且會一直維持為零。就我所知，這是 Opendoor 歷史上第一次我們可以說佇列為零，並且預期它會持續為零。而且，這不只發生在定價。去年，Opendoor 最有效率的員工，一週也只能在我們買下的房屋上開通幾十項公用事業服務。

Last week, that number was 400. Last year, the most efficient closer at Opendoor could close around 135 escrow files. Today, that number is 431. Second. I want to talk about how workflows connect to each other, right? Opendoor used to have all of the entire industry's fragmentation inside the company.

上週，這個數字是 400。去年，Opendoor 最有效率的結案人員大約能完成 135 份託管（escrow）檔案的結案。今天，這個數字是 431。第二。我想談談工作流程如何彼此串接，對吧？Opendoor 過去把整個產業的碎片化都搬進了公司內部。

Our title and escrow team had to open a legacy software system just to look up which Opendoor entity was buying a home. Now that entire information is in one system they already use. That's not just in title and escrow across the entire stack. We're doing the same thing across the entire transaction where our people can benefit from having one tool to manage their day.

我們的產權與託管團隊必須打開一套舊式軟體系統，才能查到是哪一個 Opendoor 法人實體在買這間房。現在，這整套資訊都在他們原本就使用的同一個系統裡。這不只是產權與託管，而是貫穿整個技術堆疊。我們也在整個交易流程做同樣的事，讓員工能用一個工具來管理他們的一天並從中受益。

The last part is our shared platform, right? In a fragmented industry, when you improve something, that makes that something better but doesn't benefit anything else, right? So if you improve acquisition gets better, but resale doesn't improve.

最後一部分是我們的共享平台，對吧？在一個碎片化的產業裡，當你改善某個環節，只會讓那個環節變好，卻不會讓其他環節受益，對吧？所以即使你把收購做得更好，轉售也不一定會改善。

At Opendoor, when we improve something, it compounds across the transaction, right? The data we use to buy a home, renovate a home, list a home, give a mortgage on a home, close a transaction, it all works together rather than hopping across multiple parts of the business. Look, there is a test being run right now for all companies.

但在 Opendoor，當我們改善某件事，它會在整個交易中產生複利效應，對吧？我們用來買房、翻新、上架、提供房貸、完成交易的資料，全部是協同運作的，而不是在業務的不同部分之間跳來跳去。坦白說，現在所有公司都正在接受一項測試。

When a new model comes out, does it make you better or does it make you more vulnerable? For a lot of companies, the honest answer is more vulnerable, right? Because the thing they sell is a thin layer on top of someone else's model. And the next release just eats that layer away. For us, we get better. Everyone has access to the same model, but what they do not have is what we feed them.

當一個新模型推出時，它會讓你變得更好，還是讓你更脆弱？對很多公司而言，誠實的答案是更脆弱，對吧？因為他們賣的東西，只是覆蓋在別人模型之上的一層很薄的介面。而下一次發布就會把那一層侵蝕掉。對我們而言，我們會變得更好。每個人都能使用同樣的模型，但他們沒有的是我們餵給模型的資料。

Every home we have bought every repair we have scoped hundreds of millions of listing photos and trillions of other photos over a decade of what our homes have actually sold for and not the misleading data that's publicly available no one else has our data advantage and because we're the largest buyer of homes in America. Our data advantage is growing. We see inside of homes and we see foundational data on homes for hundreds and hundreds and hundreds of homes and no one else gets that data.

我們買過的每一間房、我們界定過的每一項修繕範圍、十年來數以億計的上架照片與數兆張其他照片，以及我們的房屋實際成交價格（而不是公開可得、但具誤導性的資料）——沒有人擁有我們這樣的資料優勢，而且因為我們是美國最大的房屋買家。我們的資料優勢還在擴大。我們能看到房屋內部，也能取得數以百計、數以百計、數以百計房屋的基礎性資料，而其他人拿不到這些資料。

A better model makes Opendoor's core economics better. It doesn't just increase our operational leverage, right? This week I sent. a note to our entire team, basically wholesale copied Jeff Bezos. There's a famous shareholder letter that he wrote in 2000 where he referenced the Benjamin Graham line that in the short run the market is a voting machine, but in the long run it's a weighing machine.

更好的模型會讓 Opendoor 的核心經濟性變得更好。它不只是提高我們的營運槓桿，對吧？本週我寄出。一封給全體團隊的訊息，基本上是直接大量照抄 Jeff Bezos。他在 2000 年寫過一封著名的股東信，引用了 Benjamin Graham 的一句話：短期來看，市場是一台投票機；長期來看，市場是一台稱重機。

Bezos said that he'd rather be weighed, right? And I also want to put Opendoor on a scale. If there were a machine to weigh Opendoor, I think it would give you five data points about our business. One, very few people inside or outside Opendoor believed that this would be true a year ago, right? But we've now proven that we can grow without overpaying for houses.

Bezos 說他寧願被稱重，對吧？我也想把 Opendoor 放到秤上。如果有一台機器可以衡量 Opendoor，我認為它會給你五個關於我們業務的數據點。第一，一年前無論在 Opendoor 內部或外部，相信這會成真的人都很少，對吧？但我們現在已經證明，我們可以在不為房屋付出過高價格的情況下成長。

Spreads went up this past year and conversion went up even more. We improved the product and the funnel and more sellers chose us. Two, we're generating volume that needs less of our balance sheet.

過去一年，價差（spreads）上升，而轉換率（conversion）上升得更多。我們改善了產品與漏斗，更多賣方選擇了我們。第二，我們正在創造需要更少資產負債表支撐的交易量。

Around half of what sellers are accepting through our D2C funnel right now is listing and Cash Now, More Later. We can serve these sellers while carrying much less risk and paying much less interest.

目前透過我們 D2C 漏斗被賣方接受的方案中，大約有一半是「代售（listing）」以及「現在拿現金、之後再拿更多（Cash Now, More Later）」。我們可以在承擔更低風險、支付更少利息的情況下服務這些賣方。

Three, we're supporting more acquisitions at a lower cost per home, down more than a third on both the measures we talked about earlier. Four, we get paid more than once per home, mortgage, title, and escrow. This is an entirely different business model. And we've added it while we're still fixing the old one, right? Now, every home we touch gives us multiple revenue opportunities. This wasn't true before.

第三，我們以更低的單戶成本支援更多收購——在我們先前談到的兩項衡量指標上都下降了超過三分之一。第四，我們在每一間房屋上不只賺一次：房貸、產權、以及託管。這是一種完全不同的商業模式。而且我們是在仍在修復舊模式的同時把它加進來的，對吧？現在，我們接觸到的每一間房屋都帶來多個營收機會。以前不是這樣。

Five, every new AI model, every single new model that comes up makes us better because we are sitting on data no one else has. So put these things together, demand that does not require us to pay higher prices. Volume that doesn't require a heavy balance sheet growth that does not cause OpEx to go up linearly a new tech stack that improves itself every single week a data advantage that is just growing look we are taking market share and making a stronger company while everyone else is backing down we have spent the past year. Building a machine, and that machine is getting better every single day. We're not done yet, it's not perfect, but it's getting better every single day. So the question is this, is Opendoor in a better economic condition than it was a year ago? The answer is obviously yes. We told you we would build a faster, stronger, better Opendoor, and what I just shared with you shows that we have.

第五，每一個新的 AI 模型、每一個出現的新模型都讓我們變得更好，因為我們掌握著別人沒有的數據。把這些因素放在一起：不需要我們支付更高價格的需求；不需要沉重資產負債表擴張、也不會讓營運費用（OpEx）線性上升的交易量；每週都能自我改進的新技術堆疊；持續擴大的數據優勢。你看，我們在其他人退縮之際，正在搶占市占並打造更強的公司；過去一年我們一直在。打造一台機器，而這台機器每天都在變得更好。我們還沒完成，它不完美，但每天都在變得更好。所以問題是：Opendoor 的經濟狀況是否比一年前更好？答案顯然是肯定的。我們告訴過你們，我們會打造一個更快、更強、更好的 Opendoor，而我剛才分享的內容顯示我們做到了。

And with that, I will turn to Christy to tell you about the numbers.

接下來，我把時間交給 Christy，請她為各位說明數字表現。

Thank you, Kaz. Financial highlights. Third quarter revenue was over $1 billion, up 12% year over year and in line with the revised expectations we provided in September, despite a weaker housing market. Contribution profit was $34 million, up 70%.

謝謝你，Kaz。財務重點。第三季營收超過 10 億美元，年增 12%，並符合我們在 9 月提供的修正後預期，儘管房市更疲弱。貢獻毛利（contribution profit）為 3,400 萬美元，年增 70%。

We sold 2,728 homes in the quarter, up from 2,568 a year ago, and purchased 4,865 homes, more than four times the 1,169 we purchased in Q3 last year. We ended the quarter with 1,780 homes under contract to purchase compared with 526 a year ago. Contribution margin was 3.3%, down from 5.8% in Q2, but up from 2.2% a year ago and in line with the revised expectations we provided in September.

本季我們售出 2,728 間房屋，高於一年前的 2,568 間；並購入 4,865 間房屋，超過去年第三季購入 1,169 間的四倍。季末我們有 1,780 間房屋處於簽約待購（under contract to purchase）狀態，相較一年前為 526 間。貢獻毛利率（contribution margin）為 3.3%，低於第二季的 5.8%，但高於一年前的 2.2%，並符合我們在 9 月提供的修正後預期。

Our adjusted EBITDA loss was $31 million, an improvement from a $33 million loss a year ago, but wider than the $4 million loss in Q2 2026. Most of the sequential decline came from the homes we bought in the spring selling into a much weaker market, along with deeper price cuts as we chose to clear rather than hold. We expect margin to decline again in Q4 and will continue selling through the remaining spring acquisitions into Q1.

我們的調整後 EBITDA 虧損為 3,100 萬美元，較一年前的 3,300 萬美元虧損有所改善，但較 2026 年第二季的 400 萬美元虧損擴大。環比下滑主要來自我們在春季買入的房屋，於更疲弱的市場中出售；同時，因我們選擇加速出清而非持有，採取了更深的降價。我們預期第四季毛利率將再次下滑，並將持續出售剩餘的春季收購房源至第一季。

Slower market clearance also showed up in our inventory aging. At quarter end, 18% of listed homes had been on the market for more than 120 days, up from 9% last quarter as clearance slowed, but still well below the market at 30% and the 51% we carried a year ago.

市場去化變慢也反映在我們的庫存屋齡（inventory aging）上。截至季末，已上架房屋中有 18% 在市場上超過 120 天，高於上季的 9%（因去化放緩），但仍明顯低於整體市場的 30%，也低於我們一年前的 51%。

We're continuing to move homes through a difficult market, releasing capital and limiting how long we remain exposed to further changes in prices. While homes we purchased before the market slowed continued to weigh margins, the business we're building is becoming more efficient. On a trailing-12-month basis through September, operations expense per acquired home was $4,000, down 36% from $6,300 a year ago.

我們持續在艱難的市場中推動房屋成交，釋放資本，並限制我們暴露於進一步價格變動的時間。雖然在市場轉弱前購入的房屋仍持續拖累毛利率，但我們正在打造的業務正變得更有效率。以截至 9 月的過去 12 個月（TTM）計算，每收購一間房屋的營運費用為 4,000 美元，較一年前的 6,300 美元下降 36%。

Fixed operating expense was $37 million in Q3, slightly up from $35 million in Q2. We've also become more efficient at acquiring customers. On a trailing-12-month basis through September, marketing expense per acquired home was $2,800, down 57% from $6,500 a year ago.

第三季固定營運費用為 3,700 萬美元，較第二季的 3,500 萬美元略增。我們在獲客方面也變得更有效率。以截至 9 月的過去 12 個月（TTM）計算，每收購一間房屋的行銷費用為 2,800 美元，較一年前的 6,500 美元下降 57%。

That efficiency, together with our operational improvements, gives us confidence to invest more. Earlier this year, we held marketing spend back while strengthening our products and our ability to deliver them at scale. We're now increasing Q4 marketing to around $20 million, up from $5 million in Q3 and above what we planned last quarter.

這些效率提升，加上我們的營運改善，讓我們有信心加大投資。今年稍早，我們在強化產品以及提升規模化交付能力的同時，控制了行銷支出。我們現在將第四季行銷費用提高至約 2,000 萬美元，高於第三季的 500 萬美元，也高於我們上季的規劃。

We've improved seller conversion, expanded our capital-light offerings. And addressed key bottlenecks using AI to scale underwriting, tailor renovations to buyer demand and better inform resale pricing we still need to prove those improvements show up in realized margins but we're no longer relying on the same manual processes to support higher volume and we're not loosening offers to buy volume while others are retreating, we're advancing.

我們提升了賣方轉換率，並擴大了輕資本（capital-light）產品。同時，我們運用 AI 解決關鍵瓶頸，以擴大承保（underwriting）規模、依買方需求調整翻修內容，並更好地為轉售定價提供資訊。我們仍需證明這些改善能反映在已實現毛利率上，但我們不再依賴相同的人工流程來支撐更高的交易量；而且在其他人退縮之際，我們並未為了追求量而放鬆收購出價——我們正在前進。

We intend to win more sellers, take share and build a stronger business. We won't let near-term pressure distract us from our long-term opportunity.

我們打算贏得更多賣方、提升市占，並打造更強健的業務。我們不會讓短期壓力分散我們對長期機會的專注。

Turning to the balance sheet, we strengthened our capital position while keeping existing shareholders at the center of those decisions. We ended the third quarter with $910 million in cash and cash equivalents. The August financing was an important part of that, and I want to spend a moment on what we accomplished.

接著談資產負債表：我們在強化資本實力的同時，也把既有股東的利益放在決策核心。第三季末，我們持有 9.10 億美元的現金及約當現金。8 月的融資是其中的重要一環，我想花點時間說明我們完成了什麼。

In August, we raised $650 million through 0% coupon convertible notes due in 2030. After funding a concurrent share repurchase and capped calls, the transaction added approximately $440 million of capital before transaction expenses. We used $135 million to repay our expiring notes at maturity.

8 月，我們透過 2030 年到期、0% 票息的可轉換公司債募集 6.50 億美元。在同時進行的股份回購與上限買權（capped calls）完成資金安排後，扣除交易費用前，本交易約增加 4.40 億美元資本。我們使用 1.35 億美元償還到期的票據。

The transaction reduced shares outstanding by approximately 5% through our first share repurchase as a public company. The capped calls were designed to limit potential future dilution. Together, these actions extended our debt maturity profile, added capital to support growth and protected more of the upside for our existing shareholders.

透過我們作為上市公司以來的首次股份回購，本次交易使流通在外股數約減少 5%。上限買權（capped calls）的設計旨在限制未來可能的稀釋。綜合而言，這些行動延長了我們的債務到期結構，增加了支持成長的資本，並為既有股東保留更多上行空間。

Our responsibility is to now put that capital to work productively. For example, Cash Now, More Later requires less capital upfront per acquisition, while list with Opendoor lets us serve sellers without buying their homes.

我們的責任是把這些資本有效率地投入運用。例如，Cash Now, More Later 每筆收購所需的前期資本較少；而「在 Opendoor 上架」（list with Opendoor）則讓我們能在不買下房屋的情況下服務賣方。

As those products grow, we can serve more customers without a proportional increase in inventory. Faster resale complements that shift by releasing capital sooner and reducing the time we're exposed to changes in the market. Together, these give us more capacity to grow with the capital we have. Before turning to guidance, I want to put the opportunity ahead of us into perspective.

隨著這些產品成長，我們可以在不按比例增加庫存的情況下服務更多客戶。更快的轉售也與這一轉變相輔相成，能更早釋放資本並縮短我們暴露於市場變動的時間。綜合起來，這些讓我們能用現有資本取得更大的成長容量。在進入財測指引之前，我想先把我們面前的機會放到更宏觀的角度來看。

In the third quarter, we signed more than 7,000 acquisition contracts, representing just over 0.5% of home-selling activity in our addressable market. That's more than four times our share a year ago, but we're still serving fewer than 1 in 100 sellers. Even in this slow market, we're seeing roughly 5 million annualized true sellers. Each incremental 10 basis points of market share is roughly 5,000 homes a year, or roughly $1.9 billion of annual transaction value at today's average price.

第三季，我們簽署了超過 7,000 份收購合約，約占我們可服務市場（addressable market）房屋出售活動的 0.5% 多一點。這比一年前的市占高出四倍以上，但我們仍然服務不到每 100 位賣方中的 1 位。即使在這個低迷市場，我們仍看到約 500 萬名年化的真實賣方（true sellers）。市占每增加 10 個基點（0.10%），約等於每年增加 5,000 間房屋，或以目前平均價格計算約 19 億美元的年化交易價值。

We've added 40 basis points of share in the past year alone. Our own funnel shows how much room we have to serve more of those sellers. Every week, around 9,000 true sellers come to Opendoor, and we go into contract with around 500. Most of the rest go on to list their homes. Each week, around 9,000 true sellers come to Opendoor.

僅在過去一年，我們就增加了 40 個基點的市占。我們自身的漏斗也顯示，我們仍有很大空間去服務更多賣方。每週約有 9,000 名真實賣方來到 Opendoor，而我們約與其中 500 名進入簽約。其餘大多數則會選擇把房屋掛牌出售。每週約有 9,000 名真實賣方來到 Opendoor。

By offering more ways to sell, we can help more than 28% in our tests. We can build a much larger business by meeting the needs of more sellers who are already coming to us. We can also earn more from each customer relationship. Historically, most of our economics have come from buying and reselling homes, complemented by title and escrow.

透過提供更多元的出售方式，在測試中我們可以多協助超過 28% 的賣方。只要滿足更多已經來到我們這裡的賣方需求，我們就能打造規模大得多的業務。我們也能從每一段客戶關係中獲得更多收益。過去，我們的大部分經濟效益來自買入並轉售房屋，並由產權（title）與代管（escrow）服務加以補充。

Our emerging mortgage offering allows us to earn more from each relationship while meeting a critical customer need and making the transaction easier. Those services also open up a broader market. Our emerging Buy with Opendoor offering can help buyers whether or not they are buying a home from Opendoor, and we see an opportunity to extend title and escrow to customers who aren't buying or selling through Opendoor.

我們新推出的房貸產品，讓我們在滿足客戶關鍵需求並讓交易更簡便的同時，也能從每一段客戶關係中創造更高收益。這些服務也為我們打開了更廣闊的市場。我們新推出的「透過 Opendoor 購買（Buy with Opendoor）」方案，無論買方是否購買 Opendoor 的房屋都能提供協助；同時，我們也看到機會將產權（Title）與代管（Escrow）服務延伸至那些未透過 Opendoor 買賣的客戶。

Similar to listings, the offerings extend our reach beyond our home acquisition business and create opportunities to earn without taking ownership of a home. Addressing more of the sellers who come into our funnel, financing more buyers and expanding title and escrow to a broader market gives us multiple ways to grow earnings and build toward profitability. That is why we're investing through this market.

與房源刊登（listings）類似，這些產品讓我們的觸及範圍超越房屋收購業務，並創造在不取得房屋所有權的情況下也能獲利的機會。在我們的漏斗中承接更多賣方、為更多買方提供融資，並將產權與代管服務擴展到更廣泛的市場，讓我們有多種方式提升收益並邁向獲利。這就是我們在此市場環境下仍持續投資的原因。

The opportunity ahead of us is significant and we're building the products and capabilities to capture it. Now let me turn to the outlook. We're rebuilding Opendoor while the housing market is shifting every week, and that means our near-term results can move significantly.

我們面前的機會相當可觀，我們正在打造產品與能力以把握它。接下來我談談展望。我們在房市每週都在變動的情況下重建 Opendoor，這意味著我們的短期業績可能會出現顯著波動。

These guideposts reflect what we see today. We'll keep making the right decisions we believe are right for the business even when they weigh on near-term results. Our focus remains on our long-term mission, tilting the world in favor of homeowners. With that in mind, here's what we expect.

這些指引反映的是我們今天所看到的情況。即使會對短期結果造成壓力，我們仍會持續做出我們認為對公司正確的決策。我們的重點仍是長期使命：讓世界更偏向房主。基於此，以下是我們的預期。

We are behind schedule on our goal of reaching positive adjusted net income by the end of 2026, measured on a 12-month go-forward basis. The same also applies to our prior expectation for adjusted EBITDA profitability on a 12-month go-forward basis.

就以未來 12 個月滾動基礎衡量而言，我們在「於 2026 年底前達成正向調整後淨利」的目標上落後於進度。同樣地，我們先前對於以未來 12 個月滾動基礎達成調整後 EBITDA 獲利的預期也不例外。

We continue to drive toward profitability, and we remain focused on closing the gap. But our fourth quarter guideposts reflect the challenging market conditions we see today.

我們仍持續推進獲利，並專注於縮小差距。但我們對第四季的指引反映了我們目前所見的嚴峻市場環境。

On revenue, we expect growth of at least 40% year over year. You can continue to follow acquisition contracts and resale closings on accountable.Opendoor.com. On contribution margin, we expect a sequential decline as we continue selling into a difficult market.

在營收方面，我們預期年增至少 40%。您可持續在 accountable.Opendoor.com 追蹤收購合約與轉售成交（resale closings）。在貢獻毛利率（contribution margin）方面，隨著我們持續在艱難市場中銷售，我們預期將出現季比下滑。

Our focus remains on moving homes and managing the capital and market exposure associated with holding them. We expect an increase in adjusted operating expenses from Q3, which will be almost entirely driven by investment in growth, including the higher marketing spend I described earlier and increased holding costs as we rebuild inventory.

我們的重點仍是加速房屋流轉，並管理持有房屋所帶來的資金占用與市場曝險。我們預期調整後營業費用將較第三季增加，幾乎完全由成長投資所驅動，包括我先前提到的更高行銷支出，以及在我們重建庫存時增加的持有成本。

The higher marketing spend also reflects seasonality. Marketing in the fall and winter helps us acquire homes ahead of the spring and summer selling seasons when buyer demand is typically stronger. We incur that marketing expense now while the revenue and margin from those homes come later. This is why we look at marketing efficiency on a trailing-12-month basis and why you should expect our quarterly marketing to move with seasonality.

較高的行銷支出也反映季節性因素。在秋冬進行行銷，有助於我們在春夏銷售旺季（買方需求通常較強）之前先行取得房源。我們現在先承擔行銷費用，而這些房屋帶來的營收與毛利會在之後才反映。這也是為什麼我們以過去 12 個月（trailing-12-month）為基礎衡量行銷效率，以及為什麼您應預期我們的季度行銷支出會隨季節性波動。

Together, sequentially lower contribution margin and higher adjusted operating expenses lead to an expected fourth quarter adjusted EBITDA loss of $70 million to $85 million. Comparing year over year, roughly a third of the increase in our expected loss is the additional marketing and the remaining two-thirds is primarily higher holding costs as we've been rebuilding our inventory.

綜合而言，貢獻毛利率的季比下降與調整後營業費用的上升，將導致我們預期第四季調整後 EBITDA 虧損為 7,000 萬至 8,500 萬美元。若以年比來看，我們預期虧損增加中約三分之一來自額外行銷支出，其餘約三分之二主要來自我們重建庫存期間較高的持有成本。

We have more work to do. We'll continue to report against the objectives we set and be clear about where we're delivering and where we're behind. With that, Michael, I'll turn it over to you for questions.

我們還有更多工作要做。我們將持續就既定目標進行揭露，並清楚說明我們在哪些方面達成、哪些方面落後。接下來，Michael，我把時間交給你進行提問。

Our first question comes to us from Will down in Phoenix.

我們的第一個問題來自鳳凰城的 Will。

Hey, Kaz, Will Bieger here. Super excited to be in the Opendoor office and installed this painting here today.

嗨，Kaz，我是 Will Bieger。非常興奮能在 Opendoor 辦公室，今天也在這裡把這幅畫裝上了。

Got a question for you. Thanks for answering my DM, by the way, and getting me involved. Really appreciate that. But I know you studied Opendoor a lot even before becoming the CEO.

我有個問題想請教你。順帶一提，謝謝你回覆我的私訊（DM）並讓我參與其中。真的很感謝。不過我知道你在成為 CEO 之前就對 Opendoor 研究很多。

Wanted to ask the question, what's something or an opportunity that you understand now that Opendoor 2.0 is here that you didn't kind of think of before joining the company? Really appreciate you answering my question and looking forward to watching this earnings call.

我想問的是，現在 Opendoor 2.0 上線了，有沒有什麼你現在才理解到、但在加入公司之前沒有想到的事情或機會？很感謝你回答我的問題，也期待收看這場財報電話會議。

Thanks, Will. Before I answer the question, let me take a second to say this.

謝謝你，Will。在回答問題之前，讓我先花一點時間說明一下。

One of the reasons I love working at Opendoor is because our shareholders believe in the company and our product that we are on a mission together to tilt the world in favor of homeowners because we believe owning a home is like owning a share in your country. It's like owning a share in your community and that homeownership just makes the world a better place.

我喜歡在 Opendoor 工作的原因之一，是因為我們的股東相信公司與我們的產品；我們共同肩負使命，讓世界更偏向房主，因為我們相信擁有房屋就像擁有你國家的股份一樣。也像是擁有你社區的一份股份，而房屋所有權確實能讓世界變得更美好。

To answer your question, look, I think there are lots of ways to answer this, but the one that has honestly surprised me the most is Title and Escrow. Title and Escrow is like the infrastructure that lets a real estate transaction happen, right? It's a lot like payments in e-commerce. And I've spent basically my entire life being a payments nerd. I took apart my first credit card terminal when I was a teenager.

回到你的問題，我想有很多種回答方式，但老實說，最讓我驚訝的是產權（Title）與代管（Escrow）。產權與代管就像是讓不動產交易得以發生的基礎設施，對吧？它很像電商中的支付系統。而我幾乎一輩子都是個支付迷。我十幾歲時就拆解過我的第一台信用卡刷卡機。

And when I joined, I thought most of my payment experience would go to building a mortgage product. And to be clear, I'm really bullish on our mortgage product and I'm much more bullish on it today than when I joined. But Title and Escrow feels like such an amazing opportunity.

我加入公司時，以為我大部分的支付經驗會用在打造房貸產品上。先說清楚，我非常看好我們的房貸產品，而且我現在比剛加入時更看好。但產權與代管看起來是一個非常驚人的機會。

Because with mortgage, what we're doing is we're pulling a system that was built in the 2000s into 2026. With title and escrow, we're building a system that operates not pre-internet, but pre-computers into modern AI world. And we're doing two things. First, we're making our title and escrow operators just way more effective, right? In some cases, they're handling three times more volume.

因為在房貸方面，我們做的是把一套建立於 2000 年代的系統拉到 2026 年。而在產權與代管方面，我們是在把一套不是「網際網路之前」、而是「電腦之前」的系統帶進現代 AI 世界。而我們正在做兩件事。第一，我們讓產權與代管的作業人員變得更有效率，對吧？在某些情況下，他們處理的量是三倍。

Than they were a year ago, right? And this is in a business where the biggest cost is people's time, right? And that goes straight to the cost of the transaction, right? So every operator that can handle more transactions gives us capacity to grow. And I'm really excited about it because when we have this capacity, we can build products. That we cannot build today and no one can build today.

比一年前多，對吧？而在這個業務中，最大的成本是人力時間，對吧？而這會直接反映在交易成本上，對吧？因此，每位作業人員能處理更多交易，就能帶給我們成長的產能。我對此非常興奮，因為當我們擁有這樣的產能，就能打造產品。那些是我們今天做不到、而且沒有人今天做得到的產品。

And this is why we're doing the second thing, we're connecting the workflows in title and escrow and mortgage to the actual buying and selling activities, right? When these systems work together, there's a lot less chasing information around. It's almost like you're going from faxing your order with a credit card number into it, into like Sears with a catalog.

這也是為什麼我們在做第二件事：把產權與代管以及房貸的工作流程，與實際的買賣活動串接起來，對吧？當這些系統協同運作時，就不需要到處追著資訊跑。這幾乎就像你從把訂單連同信用卡號碼一起傳真出去、寄到像 Sears 那樣的型錄商店。

You're going from that to Shop Pay today, right? This is what the big jump will be. And I generally underestimated how big a business title and escrow could become just in its own right.

一路進化到今天的 Shop Pay，對吧？這就是那個巨大的躍升。而我過去確實低估了產權與代管本身能成為多大的業務。

So the capabilities we're building for Opendoor transactions today are going to be used for non-Opendoor transactions and they can help us deliver title and closing services to people who never buy a home from us or sell a home to us. And that's why I'm just so excited about this because it's making the core business better, but it also allows us to extend the reach of the software we have built outside Opendoor. So we don't have to buy your home to make homeownership better for everyone.

因此，我們今天為 Opendoor 交易所打造的能力，未來也會用在非 Opendoor 的交易上，並能協助我們向那些從未向我們買房或賣房的人提供產權與交割（closing）服務。這也是我對此如此興奮的原因：它讓核心業務變得更好，同時也讓我們能把已打造的軟體觸角延伸到 Opendoor 之外。所以，我們不必買下你的房子，也能讓每個人的房屋所有權體驗變得更好。

Our next question comes to us from [Nia N]. who asks when is Opendoor going to be profitable?

我們的下一個問題來自 [Nia N]。她問 Opendoor 什麼時候會獲利？

Yeah, this is simple. Our goal is to be ANI positive on a 12-month go-forward basis at the end of this year, right? We're a bit behind on this today, but we're going to go through some walls and catch up. Like this isn't going to be easy, nothing is for certain, but that's what we're going to do.

是的，這很簡單。我們的目標是在今年年底，以未來 12 個月的滾動基礎達到 ANI 為正，對吧？我們目前在這方面稍微落後，但我們會突破一些障礙並追趕上來。這不會容易，沒有任何事是確定的，但這就是我們要做的。

Our next question comes to us. Angelo E. asks. You said we don't need rates to fall. If Dalio is right that the bond bear market has a long way to go and mortgages hit 8% through 2027, do we stay solvent without raising money? And how are we hedging our inventory risk so open as a battle on our technology, not home prices?

我們的下一個問題是。Angelo E. 問：你說我們不需要利率下跌。如果 Dalio 的判斷正確，債券熊市還有很長的路要走，且房貸利率一路到 2027 年都達到 8%，我們在不募資的情況下能否維持償付能力？另外，我們如何對沖庫存風險，讓競爭焦點放在我們的技術，而不是房價？

Yeah. Hey, thanks, Angel, for the question. Let me start with like the easy answer, which is we ended the third quarter with over $900 million in cash. And I'm very happy about our balance sheet position.

是的。嘿，Angel，謝謝你的問題。我先從比較簡單的答案開始：我們在第三季末有超過 9 億美元的現金。我對我們的資產負債表狀況非常滿意。

So, we have all the business, right? Our job is to provide, the bigger question is how we're changing the right solution for the user who comes to us, right? And different customers want different things. They want different amounts of certainty. They will take different amounts of risk.

所以，我們擁有所有業務，對吧？我們的工作是提供——更大的問題是，我們如何為來找我們的使用者提供合適的解決方案，對吧？不同客戶想要不同的東西。他們想要不同程度的確定性。他們也願意承擔不同程度的風險。

And basically all of our products are going to end up being like the same underlying infrastructure and models. That we have built basically from scratch for Cash Now, More Later but used with different like differing risk levels right so we are rebasing the entire company back on top of Cash Now, More Later with basically kind of like a risk toggle like let's just think about the products for a second so with cash.

基本上，我們所有產品最後都會建立在相同的底層基礎設施與模型之上。這些是我們幾乎從零開始打造、用於 Cash Now, More Later 的，但會以不同的風險水準來使用；因此我們正把整家公司重新建立在 Cash Now, More Later 之上，並加入一個類似「風險切換」的機制——先稍微想一下產品：以現金（cash）來說。

You get this maximum certainty, right? We take on the resale risk and like economically that is a lot like insurance, right? It has value, it has a cost, and some people want insurance on some things and some people don't, but this is becoming a smaller and smaller part of our portfolio.

你會得到最高程度的確定性，對吧？我們承擔轉售風險，而從經濟本質上看，這很像保險，對吧？它有價值，也有成本；有些人希望在某些事情上有保險，有些人則不需要，但這正逐漸成為我們投資組合中越來越小的一部分。

Look, I think cash is going to be a massive business for us, but it'll be a smaller and smaller part of our portfolio overall. Cash Now, More Later is where you receive the cash you need upfront to do the thing you need to do, but then you participate in the outcome of the home and it sells, right?

你看，我認為現金（cash）對我們會是一個非常龐大的業務，但它在我們整體投資組合中的占比會越來越小。Cash Now, More Later 是你先拿到你需要的現金去完成你需要做的事，但之後你會參與房屋最終出售結果，對吧？

That changes how risk is shared. And listing is like, we're early here, but listing is basically cash down or later, except with zero risk upfront, right? We use exactly the same tools we use for cash down or later. And we help you sell your home without having to buy it from you, right?

這改變了風險如何被分攤。而「刊登（listing）」也是——我們在這方面還很早期——但 listing 基本上就是 Cash Now, More Later，只是前期零風險，對吧？我們使用的工具完全與 Cash Now, More Later 相同。我們幫你賣房，但不需要向你買下房子，對吧？

The seller will retain the risk and the exposure to the eventual sale price, but they will use all of our tools, workflows, title and escrow technology and mortgage technology to make the transaction work, right? We don't need to buy every home to help every customer, to be clear, like even today. When we're like super early on this, there are folks who don't buy or sell a home to Opendoor who use our mortgage product, who use our title and escrow products, right? This gives us more flexibility as the market changes and it gives us more ways to grow without taking risk. And for homes that we do own.

賣方會保留風險以及對最終成交價的曝險，但他們會使用我們所有工具、工作流程、產權與託管（title and escrow）技術，以及房貸技術，來讓交易順利完成，對吧？需要說清楚的是，我們不必買下每一棟房子才能服務每一位客戶——即使是今天也是如此。在我們仍處於非常早期階段時，就已經有人沒有向 Opendoor 買賣房屋，但會使用我們的房貸產品、使用我們的產權與託管產品，對吧？這讓我們在市場變化時更有彈性，也讓我們能在不承擔風險的情況下有更多成長方式。至於我們確實持有的房屋。

Underwriting and selling efficiently still matter like a lot, right? But our direction is very clear. Serve more customers, build one infrastructure and one set of models and use it for everything and be very deliberate about the risk we take and how we price it.

高效率地承保與銷售仍然非常重要，對吧？但我們的方向非常清楚。服務更多客戶，打造一套基礎設施與一組模型，並把它用在所有事情上；同時對我們承擔的風險以及如何定價保持非常審慎。

Our next question comes to us from Calvin R. who asks, as a long-term shareholder, I've watched age inventory drop from 36% to 9%. And home loans and Opendoor owned go live. What's the one metric you want shareholders like me to judge you on over the next 12 months and what's the target?

我們的下一個問題來自 Calvin R.，他問：作為長期股東，我看到逾齡庫存（age inventory）從 36% 降到 9%。而房屋貸款與 Opendoor Owned 也上線了。未來 12 個月，你希望像我這樣的股東用哪一個指標來評價你們？目標是什麼？

Yeah, thanks Kevin. We don't really have just one hero metric and that's like very deliberate because Goodhart's law is like real, right? When a measure becomes a target, people start optimizing for the measure rather than the outcome it is like designed to produce.

是的，謝謝你，Kevin。我們其實沒有單一的「英雄指標」，而這是刻意為之，因為古德哈特定律（Goodhart's law）確實存在，對吧？當一個衡量指標變成目標時，人們就會開始優化這個指標，而不是它原本要產生的結果。

Like this is why lots of tech companies just go right over their skis while improving their KPIs. And that's why we talk about the three objectives because they're kind of like three legs of the stool and they all work together, right? Scale acquisitions, improving economics and resale velocity and build operating leverage. I'm not going to repeat everything we said on the call, but I think we're very green on one and three and for two, what you're seeing is that we're scaling our asset-light products. We're growing our mortgage product, and we're improving the core pricing and home management systems.

這也是為什麼很多科技公司在改善 KPI 的同時，卻會過度擴張、失控。所以我們談三個目標，因為它們有點像三腳凳的三條腿，彼此配合，對吧？

We have work to do, but we're making great progress here. And that's kind of the beauty of these three objectives. They kind of keep each other honest. You can't take an easy shortcut. We have to do the hard work required to make all of them green.

我們還有工作要做，但我們在這裡取得了很大的進展。而這三個目標的妙處就在於：它們會彼此制衡。你不能走簡單的捷徑。我們必須做艱難的工作，才能讓三項都變成綠燈。

Great. David C. asks, what milestones would tell investors that Opendoor is successfully shifting from a capital-heavy 1P model toward a more capital light 2P or 3P structure.

很好。David C. 問：哪些里程碑能讓投資人看出 Opendoor 正成功地從資本密集的 1P 模式，轉向更輕資本的 2P 或 3P 架構？

That's a good question. Thanks, David. I think it's kind of useful. Let me just tell you how I think you should think about it, right?

這是個好問題。謝謝你，David。我覺得這樣思考會比較有幫助。我先告訴你我認為你應該怎麼看待它，好嗎？

So Christy said that we see around 9,000 true sellers every week. And to be clear, this is not like all the sellers we see like a lot more, but these are 9,000 people that we know are going to sell or list a home, right? We can't serve all of them if the only product we offer requires to buy every home in America, right? We just can't do that, right?

Christy 說我們每週大約看到 9,000 位真正的賣方。需要說清楚的是，這不是我們看到的所有賣方——實際上更多——但這 9,000 人是我們知道他們一定會賣房或刊登房源的人，對吧？如果我們提供的唯一產品都要求買下美國的每一棟房子，那我們就無法服務所有人，對吧？我們就是做不到，對吧？

And when we were like a single product company, we were effectively saying. To every customer, take this certainty at this price and take it or leave it, right? And for some customers, that was exactly what they wanted, but for others, we were solving a problem they did not have. They did not want to pay us to take a risk they didn't want to give away, right?

而當我們還像是一家單一產品公司時，我們等於是在對每位客戶說：以這個價格換取這份確定性，要就要、不然就算了，對吧？對某些客戶來說，那正是他們想要的；但對另一些客戶，我們是在解決一個他們根本沒有的問題。他們不想付錢給我們，讓我們承擔一個他們不想轉讓出去的風險，對吧？

That's why we're building a listing and Cash Now, More Later alongside cash. So if you want certainty at an agreed price.

這就是為什麼我們在現金（cash）之外，同時打造 listing 與 Cash Now, More Later。所以如果你想要在約定價格下的確定性。

And you want to be done with the sale, we have cash. If you want something now to move, to put a down payment in your next place, but want to participate with us in the upside, we have Cash Now, More Later. And if you just want to use all the tools we have built to get great listing product, we have listed Opendoor, right? But you shouldn't have to pay us to take a risk that you are comfortable keeping.

而且你想把出售這件事一次做完，我們有現金（cash）。如果你想先拿到一些錢來搬家、在下一個住處支付頭期款，但又想和我們一起分享上行空間，我們有 Cash Now, More Later。如果你只是想使用我們打造的所有工具，來獲得很棒的刊登（listing）產品，我們有 Opendoor 的 listing，對吧？但你不應該必須付錢給我們，讓我們承擔一個你自己也願意保留的風險。

And we shouldn't have to use our balance sheet to provide you certainty that you don't want. That's better for customer and it's more efficient for Opendoor. And I think it's like relatively well understood that Opendoor's capital costs are not linear, right? The first 80% of the home that we financed cost a lot less in interest than the last 20%, right?

而且我們不應該必須動用資產負債表，去為你提供你其實不想要的確定性。這對客戶更好，對 Opendoor 也更有效率。而且我想大家相對都理解，Opendoor 的資金成本並不是線性的，對吧？我們為房屋融資的前 80%，其利息成本遠低於最後 20%，對吧？

So when a product requires us to take less risk over time, our cost of capital should reflect that. So I think if I were outside the company, the best way I would track this is how much interest is Opendoor paying as a percentage of our volume. And if we're successful in this shift from 1P to 2P to 3P, that number should go down over time. by a lot.

所以當某個產品要求我們隨時間承擔更少風險時，我們的資金成本就應該反映這一點。所以我想如果我是公司外部的人，我追蹤這件事的最佳方式會是：Opendoor 支付的利息占我們交易量的百分比是多少。如果我們成功完成從 1P 到 2P 再到 3P 的轉型，這個數字應該會隨時間下降。而且會下降很多。

Our next question comes from Elmir C., who's curious. With a little over a million homes listed on Zillow and Redfin, can Opendoor realistically capture 10% of that market within the next five years? What would it take to achieve this scale and is its current business model capable of supporting it?

下一個問題來自 Elmir C.，他很好奇。在 Zillow 和 Redfin 上掛牌的房屋略超過一百萬套的情況下，Opendoor 是否能在未來五年內實際拿下其中 10% 的市場？要達到這樣的規模需要什麼條件？目前的商業模式是否有能力支撐？

Thanks, Elmir. This is a hard question for me to answer because I mostly think about the customer who's in front of us right now. Rather than like metrics many years out, but let me just answer it. Look, we grew our market share 4x over the past year, right? To about half a percent of the U.S. real estate market.

謝謝你，Elmir。這題我很難回答，因為我大多時候只在想眼前正在服務的客戶。而不是像多年後的指標，但我還是回答一下。你看，我們在過去一年把市占率提高了 4 倍，對吧？到大約占美國房地產市場的 0.5%。

Another 4x gets us to 2% and then another 4x gets us to 8%, right? That's the math. And today, like one in four people in our core markets who go on to list their home. Already come to Opendoor.com. That isn't a forecast, but I don't think anyone should tell us job well done if we end up being 10% on market in five years, right?

再乘 4 倍就是 2%，再乘 4 倍就是 8%，對吧？數學就是這樣。而且今天，在我們核心市場裡，最後真的去掛牌賣房的人，大概每四個就有一個。已經會先來 Opendoor.com。這不是預測，但我不覺得如果五年後我們只有 10% 市占，任何人應該跟我們說「做得好」，對吧？

Because we have lots of room to grow. The question is, how do we build a business that can support that scale, right? Like first of all, you clearly can't do it all with cash alone, we've already talked about this.

因為我們還有很大的成長空間。問題是，我們要怎麼打造一個能支撐那種規模的企業，對吧？首先，你顯然不可能只靠現金就做到，我們已經談過了。

How listing and Cash Now, More Later help us serve more customers right but the second thing that's important is that you cannot do it if you're a traditional real estate company right it's not possible like real estate companies have had this bad habit.

掛牌（listing）以及「現在拿現金、之後拿更多」（Cash Now, More Later）如何幫助我們服務更多客戶，這點沒錯；但第二個重要的是：如果你是一家傳統房地產公司，你做不到，對吧？這不可能，因為房地產公司一直有個壞習慣。

When the market gets hot and revenue grows, these companies hire a crap ton of people to keep up with this. All the money that they thought they were going to get when volume comes gets lost either because of bad operations or expensive operations.

當市場火熱、營收成長時，這些公司就會為了跟上而大量招人。他們以為交易量上來就能賺到的錢，最後不是因為營運不佳，就是因為營運成本太高而被吃掉。

And then when the market slows down. These companies are left with a cost structure and a system built for different business, right? That's the disease that Opendoor 2.0 is designed to cure, right? That's why I spent so much of my time on these calls talking about operator efficiency, right? When we say we're investing in growth a lot, what I mean is we are building the capacity to handle the growth that will arrive.

然後當市場放緩時。這些公司就會背著一套成本結構與系統，但那是為了另一種生意而打造的，對吧？這就是 Opendoor 2.0 要治的病，對吧？這也是為什麼我在這些電話會議上花這麼多時間談營運效率，對吧？當我們說我們正在大量投資成長時，我的意思是：我們正在建立能承接即將到來的成長的產能。

Like we could have postponed building these systems, right? We could have kept the old processes, put duct tape over everything and hired more people as volume grew, right? But that would have meant that we would have carried a broken Opendoor into growth when growth arrives. So instead what we're doing is we're changing the work. Software moves information between systems and AI.

像是我們本來可以延後建置這些系統，對吧？我們可以維持舊流程，把所有問題都用膠帶先貼起來，然後隨著量增就多招人，對吧？但那代表當成長來臨時，我們會帶著一個有缺陷的 Opendoor 一起進入成長。所以相反地，我們正在做的是改變工作的方式。由軟體在系統之間移動資訊，並由 AI。

Does the risk scoring and handling the straightforward cases, right? And people focus only on decisions that require human judgment. Like you can see that in our progress. In Q3, we purchased more than four times as many homes as a year ago, right? And our operation expense was basically the same.

負責風險評分並處理那些比較直接的案例，對吧？而人只專注在需要人類判斷的決策上。你可以從我們的進展看出來。在第三季，我們購買的房屋數量比一年前多了四倍以上，對吧？而我們的營運費用基本上差不多。

Like the point of investing now is to make the next stage of growth profitable. Every time we grow fourfold, we are not going to increase the size of our OpEx. We are going to put that volume through a better system and we're going to be able to do the thing real estate companies have not traditionally been able to do in good times. Look, I'm not setting a five year market share target today, but I genuinely think that if we end up having 10% of the market.

現在投資的重點，是讓下一階段的成長能夠獲利。每次我們成長四倍，我們不會把 OpEx（營運費用）的規模也跟著增加。我們會把這些量導入更好的系統，並且做到房地產公司在景氣好時傳統上做不到的事。你看，我今天不會設定五年後的市占目標，但我真心認為，如果最後我們只拿到 10% 的市場。

Five years from now, we should all be kind of disappointed, right? The products we have built give customers a reason to choose us and the software and the AI that we are building allows us to serve them and serve them really well.

五年後，我們大家都應該會有點失望，對吧？我們打造的產品給了客戶選擇我們的理由，而我們正在建置的軟體與 AI 讓我們能服務他們，而且服務得非常好。

Moving over, our next question comes to us from Dae Lee from JPMorgan, curious, Kaz. With the macro shifting against Opendoor, has that reinforced your approach of underwriting new homes for clearing speed? Does anything need to change or are you more confident?

接著，下一個問題來自摩根大通（JPMorgan）的 Dae Lee，他想問，Kaz：在總體環境轉而不利於 Opendoor 的情況下，這是否更強化了你們以「快速出清速度」為核心來承作新房源的核保方式？有什麼需要改變嗎？還是你更有信心了？

Thanks, Dae. I think you asked one question, but let me answer two things. Let me talk about dispersion and underwriting. First, dispersion. I basically went around for my first three weeks at Opendoor and said dispersion.

謝謝你，Dae。我想你問的是一個問題，但我想回答兩件事。我先談分散度（dispersion）和核保（underwriting）。第一，分散度。我在加入 Opendoor 的前三週基本上到處都在講「分散度」。

Because we are focused on diversifying the homes we are buying across multiple different vectors, right? Geography, home type, price point, etc.. There's a reason why people call diversification the only free lunch in finance, right? But to diversify well, you have to understand how the assets are actually different, right?

因為我們專注於沿著多個不同維度分散我們買入的房屋，對吧？地理位置、房屋類型、價格帶等等。人們之所以說分散投資是金融裡唯一的免費午餐，是有原因的，對吧？但要做好分散，你必須理解這些資產實際上到底哪裡不同，對吧？

And those differences go much deeper than like geography or price, right? People talk about the housing market as though every home is going through the same thing.

而且那些差異遠比地理或價格更深，對吧？人們談住房市場時，好像每一間房子都在經歷同樣的事情。

That's not true. It's not even true in the same neighborhood. Two homes can look identical on MLS but be very different properties and have very different margin profiles.

那不是真的。即使在同一個社區也不成立。兩間房子在 MLS 上看起來可能一模一樣，但可能是非常不同的物件，且有非常不同的毛利結構。

The condition of the home relative to what the buyer expects, the slope of the driveway, the noise of the train, and the feedback of the people who are actually touring the home. Like those tell very different stories, right?

房屋狀況相對於買方期待的落差、車道的坡度、火車噪音，以及實際去看房的人給的回饋。這些都在講非常不同的故事，對吧？

In a market like this, those differences matter like a lot, and averages can hide exactly the things that will make you money, right? That's also the advantage that we have with scale. The more homes we work on, the more homes we see, the more we learn and we see what attracts buyers and what objections come up and how we can prepare the homes.

在像現在這樣的市場裡，這些差異非常重要，而平均值可能會把那些真正能讓你賺錢的因素完全掩蓋掉，對吧？這也是我們規模化的優勢。我們處理的房子越多、看到的房子越多，我們學到的就越多；我們會看到什麼吸引買家、會出現哪些反對意見，以及我們該如何把房子準備好。

In order to get an offer, right? That was actually one of the bad things about Opendoor shrinking is that our dataset shrank when we stopped buying homes, right? And that's just a thing you don't want to do. You don't want to lose volume because it costs you your data advantage that makes everything else better.

才能拿到出價，對吧？其實 Opendoor 縮小規模的一個壞處是：當我們停止買房時，我們的資料集也跟著縮小了，對吧？而那正是你不想做的事。你不想失去交易量，因為那會讓你失去資料優勢，而資料優勢會讓其他一切都變得更好。

Okay, so that was on dispersion. I also want to talk about. Underwriting, right? That's an incredibly important thing. So for every home that you underwrite, you need to answer three questions, right?

好，所以以上是分散度。我也想談談。核保（underwriting），對吧？那是一件極其重要的事。所以對於你核保的每一間房子，你都需要回答三個問題，對吧？

How much will the home sell for? How long will it take to sell? And three, and most importantly, what are the odds that you're wrong about these two things, right? Opendoor's old valuation model was built in the pre-AI era, right? It was incredibly compute constrained.

房子會賣多少錢？需要多久才能賣掉？第三個，也是最重要的，你在這兩件事上判斷錯誤的機率有多大，對吧？Opendoor 舊的估值模型是在 AI 出現之前的時代建立的，對吧？它在運算資源上受到非常大的限制。

We also had like just generally underinvested in it, especially when it came to predicting days in possession. So we've invested a lot here, right? And these models are generally very good now.

而且我們整體上對它的投入也不足，尤其是在預測持有天數（days in possession）方面。所以我們在這裡投入了很多，對吧？而這些模型現在整體而言都非常好。

And now we're investing even more in the third question, like how wrong could we be? The confidence around the estimate matters. Just as much as the estimate itself, right? Two homes can have the same expected sale price or holding period but have very different uncertainty bands around them. And obviously we shouldn't underwrite those things the same way. Okay, that was the one, Sorry.

而現在我們在第三個問題上投入得更多，也就是：我們可能會錯到什麼程度？對估計值的信心程度很重要。它的重要性不亞於估計值本身，對吧？兩棟房子可能有相同的預期售價或持有期間，但其不確定性區間可能差異很大。而顯然，我們不應該用同樣的方式去承作／核保這些案子。好，剛剛那個是第一個，抱歉。

Our next question comes to us from Andrew from Citizens. I think this might be a better one for Christy.

下一個問題來自 Citizens 的 Andrew。我想這題可能更適合 Christy 來回答。

Explain to us the cadence of repricing homes in the quarter. Has this now been fully recognized or should we expect this to flow through contribution margins over the next one to two quarters as inventory is sold?

請向我們說明在一個季度內對房源重新定價的節奏。這部分現在是否已經完全反映，還是我們應該預期，隨著庫存售出，未來一到兩個季度會逐步反映到貢獻毛利率（contribution margins）上？

Thank you for the question, Andrew. And I'm going to take it in two parts. One, how we're driving demand for our homes, including pricing. And two, how the current inventory will flow through to contribution margin or how we expect it to do so. On demand, we're improving both price and product. So historically, our primary lever was reducing price. We've become more sophisticated about how we do that, moving from largely linear reductions to models that leverage multiple data points such as buyer activity, agent feedback, and comparable homes to inform the timing and size of adjustments for each home.

謝謝你的問題，Andrew。我會分兩個部分回答。第一，我們如何推動房源需求（包括定價）。第二，目前庫存將如何流入貢獻毛利率，或我們預期會如何反映。在需求面，我們同時改善價格與產品。過去我們主要的槓桿是降價。我們在做法上變得更成熟，從過去多為線性的降價，轉向使用能利用多個資料點的模型，例如買方活動、經紀人回饋與可比房源，來決定每一棟房子的調整時點與調整幅度。

And better initial listing strategies, learning from buyer and agent response, and adjusting faster as conditions change. But lowering the price is only one way to attract a buyer. We're focused on making our homes easier to shop, more appealing compared to other homes in the same neighborhood, and more affordable by connecting financing to the purchase. So the goal is to generate stronger buyer demand at the same asking price. We still obviously need a price competitively, but a more appealing home, more attractive financing, and an easier buyer experience gives us more ways to win buyers.

以及更好的初始上市策略，從買方與經紀人的反應中學習，並在條件變化時更快調整。但降價只是吸引買方的一種方式。我們專注於讓我們的房子更容易選購、相較同一社區的其他房子更具吸引力，並透過把融資與購屋連結起來，讓它更可負擔。因此目標是在相同的開價下，產生更強的買方需求。我們當然仍需要具競爭力的定價，但更吸引人的房子、更有吸引力的融資，以及更容易的買方體驗，讓我們有更多方式贏得買方。

On contribution margin, the impact of the current inventory is flowing through. Most of the sequential decline came from homes we bought in the spring, selling into a weaker market, along with deeper price cuts as we chose to clear rather than hold. We expect margins to decline again in Q4 and will continue selling through the remaining spring acquisitions into Q1.

在貢獻毛利率方面，目前庫存的影響正在反映中。大部分的環比下滑來自我們在春季買入的房子，在較疲弱的市場中出售，同時我們選擇加大降價幅度以加速出清而非持有。我們預期第四季毛利率會再下滑，並將持續在第一季賣出剩餘的春季收購庫存。

I think we have time for one more. This one is also from Andrew from Citizens, this will be for you, Kaz. With mortgage out of beta, what are the keys looking ahead to driving better attach? What is the opportunity for driving demand for mortgage with non-Opendoor home buyers?

我想我們還有時間再回答一題。這題同樣來自 Citizens 的 Andrew，這題給你，Kaz。房貸產品已經結束 beta 測試，展望未來，推動更好的綁定率（attach）的關鍵是什麼？對於非 Opendoor 房屋買家，推動房貸需求的機會有多大？

Thanks, Andrew. I want you to know that I noticed that you hacked our system and got two questions in but let me answer it. So I think it helps to start with who the customer is. Like most people who ask us about our mortgage attach rate. Underestimate what we're building, right?

謝謝你，Andrew。我想讓你知道，我注意到你「駭」了我們的系統塞進兩個問題，但我還是回答。我認為從客戶是誰開始談會有幫助。就像大多數問我們房貸綁定率的人一樣，他們低估了我們正在打造的東西，對吧？

Look, we expect to finance plenty of buyers who buy Opendoor homes, but we also have sellers who don't sell to us and buyers who don't buy one of our homes, right? Those people can still take a mortgage from us because they have a transaction to complete, right? So we have like around 9,000 true sellers come to us every week.

你看，我們預期會為許多購買 Opendoor 房屋的買家提供融資，但我們也有賣方沒有把房子賣給我們，以及買方沒有買我們的房子，對吧？這些人仍然可以向我們申請房貸，因為他們有一筆交易要完成，對吧？所以我們每週大約有 9,000 位真正的賣方來找我們。

Thousands of buyers, right? Someone might decide to list their home instead of taking cash offers, but like that doesn't mean they don't need a mortgage, right? So if your denominator is just the homes that Opendoor sells, you're just underestimating the product, I think we can generate way more mortgages than homes we sell. Like we will be over 100% on that metric, right? That's why listing and buy with Opendoor matter. Like the customer relationship does not end because we didn't buy their home.

還有數千名買方，對吧？有人可能決定把房子掛牌出售，而不是接受現金收購報價，但這不代表他們不需要房貸，對吧？所以如果你的分母只算 Opendoor 賣出的房子，你其實是在低估這個產品；我認為我們能產生的房貸量會遠超過我們賣出的房屋量。在那個指標上，我們會超過 100%，對吧？這就是為什麼「掛牌」以及「透過 Opendoor 買房」很重要。客戶關係不會因為我們沒有買下他們的房子就結束。

Look, I think I'll end this section by borrowing from Keith, who's the Chair of our Board. I think this is actually literally his pinned tweet. There's like this secret startup success formula, like find a large, highly fragmented industry with low NPS. Vertically integrate it, build a solution that adds value by building a great product, right?

你看，我想用我們董事會主席 Keith 的話來結束這一段。我想這其實就是他置頂推文的內容。有一個像是祕密的新創成功公式：找到一個規模大、高度分散、NPS 低的產業；把它做垂直整合，透過打造很棒的產品來提供價值的解決方案，對吧？

And then you end up having a very large business and buying a home fits that description perfectly. Look, we can connect brokerage and title and escrow and mortgage and buying and selling and simplify the entire process and use the combined economics to not just give the customer a better deal, but a better product.

然後你最後會擁有一個非常大的事業，而買房完全符合這個描述。你看，我們可以把房仲、產權（title）、代管（escrow）、房貸、買與賣串接起來，簡化整個流程，並運用整合後的經濟效益，不只是讓客戶拿到更好的交易條件，也得到更好的產品。

When it comes to mortgage, I think we basically have just two jobs. One is discovery, make sure, and the second is make the value obvious, right? The mortgage industry has spent a lot of time making the cost of mortgage very confusing.

談到房貸，我認為我們基本上只有兩個工作。第一是「被發現」（discovery），確保……；第二是讓價值一目了然，對吧？房貸產業花了很多時間把房貸成本弄得非常令人困惑。

So our job is to simplify away the BS that harms customers. And I think we are going to do a really good job here. I think what's going to happen in a little while is that the number of mortgages we issue will be higher than the number of homes that we sell. Okay, I think we have run out of time. With that, let me say thank you.

所以我們的工作是把那些傷害客戶的鬼扯（BS）簡化掉。而我認為我們會在這裡做得非常好。我認為過一段時間會發生的事是：我們核發的房貸數量會高於我們賣出的房屋數量。好，我想我們時間到了。那麼，讓我說聲謝謝。

Thank you for joining us. And as always, we'll continue to update you on accountable.Opendoor.com because we want to be held accountable as we build this company to tilt the world in favor of homeowners. Thanks, folks.

感謝各位加入。一如既往，我們會持續在 accountable.Opendoor.com 向各位更新，因為我們希望在打造這家公司、讓世界更偏向房主的過程中，接受各位的監督與問責。謝謝各位。