Upstart Holdings Inc (UPST) 2022 Q3 法說會逐字稿

Good day, and welcome to the Upstart Third Quarter 2022 Earnings Call. Today's conference is being recorded. At this time, I would like to turn the conference over to Mr. Jason Schmidt, Vice President of Investor Relations. Please go ahead, sir.

美好的一天，歡迎參加 Upstart 2022 年第三季度財報電話會議。今天的會議正在錄製中。此時，我想將會議轉交給投資者關係副總裁 Jason Schmidt 先生。請繼續，先生。

Good afternoon and thank you for joining us on today's conference call to discuss Upstart's third quarter 2022 financial results. With us on today's call are Dave Girouard, Upstart's Chief Executive Officer; and Sanjay Datta, our Chief Financial Officer.

下午好，感謝您參加今天的電話會議，討論 Upstart 的 2022 年第三季度財務業績。 Upstart 的首席執行官 Dave Girouard 與我們一起參加今天的電話會議；和我們的首席財務官 Sanjay Datta。

Before we begin, I want to remind you that shortly after the market closed today, Upstart issued a press release announcing its third quarter 2022 financial results and published an Investor Relations presentation. Both are available on our Investor Relations website, ir.upstart.com.

在我們開始之前，我想提醒您，在今天市場收盤後不久，Upstart 發布了一份新聞稿，宣布了其 2022 年第三季度的財務業績，並發布了一份投資者關係報告。兩者都可以在我們的投資者關係網站 ir.upstart.com 上找到。

During the call, we will make forward-looking statements, such as guidance for the fourth quarter of 2022 related to our business and our plans to extend our platform in the future. These statements are based on our current expectations and information available as of today and are subject to a variety of risks, uncertainties, and assumptions. Actual results may differ materially as a result of various risk factors that have been described in our filings with the SEC.

在電話會議期間，我們將做出前瞻性陳述，例如與我們的業務相關的 2022 年第四季度的指導以及我們未來擴展平台的計劃。這些陳述基於我們目前的預期和截至今天可用的信息，並受到各種風險、不確定性和假設的影響。由於我們在提交給美國證券交易委員會的文件中描述的各種風險因素，實際結果可能存在重大差異。

As a result, we caution against placing undue reliance on these forward-looking statements. We assume no obligation to update any forward-looking statements as a result of new information or future events, except as required by law. In addition, during today's call, unless otherwise stated, references to our results are provided as non-GAAP financial measures and are reconciled to our GAAP results, which can be found in the earnings release and supplemental tables. (Operator Instructions)

因此，我們告誡不要過分依賴這些前瞻性陳述。我們不承擔因新信息或未來事件而更新任何前瞻性陳述的義務，除非法律要求。此外，在今天的電話會議中，除非另有說明，否則對我們結果的引用作為非公認會計原則財務指標提供，並與我們的公認會計原則結果相一致，可在收益發布和補充表格中找到。 （操作員說明）

Later this quarter, Upstart will be participating in Citi's 2022 FinTech Conference on November 15 and Wedbush's Disruptive Finance Conference, December 2.

本季度晚些時候，Upstart 將參加 11 月 15 日的花旗 2022 年金融科技會議和 12 月 2 日的 Wedbush 的顛覆性金融會議。

Now I'd like to turn it over to Dave Girouard, CEO of Upstart.

現在我想把它交給 Upstart 的首席執行官 Dave Girouard。

Good afternoon, everyone. Thank you for joining us on our earnings call covering our third quarter 2022 results. I'm Dave Girouard, Co-Founder and CEO of Upstart. Our results in Q3 were certainly not what we wanted them to be, but I also believe they reflect the Upstart team making the right decisions in a very challenging economic environment for the long-term success of the company.

大家下午好。感謝您參加我們關於 2022 年第三季度業績的財報電話會議。我是 Upstart 的聯合創始人兼首席執行官 Dave Girouard。我們在第三季度的結果當然不是我們想要的，但我也相信它們反映了 Upstart 團隊在極具挑戰性的經濟環境中為公司的長期成功做出了正確的決定。

Our revenue was down primarily because loan volume in our platform is down and secondarily, because credit markets are extremely dislocated. Higher interest rates and significantly elevated risk in the economy means we're approving about 40% fewer applicants than we would have a year ago. And those approved today are seeing offers about 800 basis points higher than they would have a year ago.

我們的收入下降主要是因為我們平台的貸款量下降，其次是因為信貸市場極度混亂。更高的利率和顯著增加的經濟風險意味著我們批准的申請人比一年前減少了約 40%。今天獲得批准的報價比一年前高出約 800 個基點。

This accounts for the vast majority of the reduction in volume. Many of our lending partners have reduced their originations, raised their rates or both. This is generally out of an abundance of caution with respect to the economy that their Upstart-powered loan portfolios have met or exceeded expectations since the program began in 2018.

這佔成交量減少的絕大部分。我們的許多貸款合作夥伴都減少了貸款額度，提高了利率或兩者兼而有之。自該計劃於 2018 年開始以來，他們的 Upstart 貸款組合已達到或超過預期，這通常是出於對經濟的高度謹慎。

But I want to be clear, contraction in lending volume in a time of rising rates and elevated consumer risk is a feature of our platform, not a bug. In fact, it's required in order to generate the returns lenders and investors expect. Whether due to an increase in expected loss rates, caution on the part of lenders or higher yield demanded by credit investors, higher interest rates and reduced volumes means that as unhappy as we are with the numbers, the system is working as intended. We're eyes wide open to the challenges of the current macro economy and determined to make the decisions that will optimize for the long-term success of Upstart. At the simplest level, we're improving our operational efficiency in the near term so that we can continue to maximize investment in our AI platform for the long term.

但我想明確一點，在利率上升和消費者風險升高的情況下，貸款量的收縮是我們平台的一個特點，而不是一個錯誤。事實上，它是產生貸方和投資者期望的回報所必需的。無論是由於預期損失率的增加、貸方的謹慎或信貸投資者要求更高的收益率、更高的利率和減少的交易量，都意味著儘管我們對這些數字不滿意，但該系統正在按預期運行。我們對當前宏觀經濟的挑戰睜大眼睛，並決心做出優化Upstart長期成功的決策。在最簡單的層面上，我們將在短期內提高我們的運營效率，以便我們可以繼續最大限度地長期投資於我們的人工智能平台。

First and foremost, we're continually calibrating our risk models to the market. Performance of credit is and always will be our highest priority. While we don't make predictions about the future, we've chosen to take a conservative position with respect to the direction of the economy in the coming quarters. In other words, we assume the worst is in front of us. We'll be pleasantly surprised if this turns out not to be the case.

首先，我們不斷根據市場調整我們的風險模型。信用表現是並將永遠是我們的重中之重。雖然我們不對未來做出預測，但我們選擇對未來幾個季度的經濟走向採取保守立場。換句話說，我們假設最壞的情況就在我們面前。如果事實並非如此，我們會感到驚喜。

Second, we're strengthening our unit economics both by increasing our revenue per loan as well as reducing marketing spend in our most expensive acquisition channels. And third, we're carefully managing our operational and fiscal plans to make sure that we're in a strong corporate footing for as long as this cycle lasts.

其次，我們正在通過增加每筆貸款的收入以及減少我們最昂貴的收購渠道的營銷支出來加強我們的單位經濟。第三，我們正在仔細管理我們的運營和財政計劃，以確保只要這個週期持續下去，我們就處於強大的企業基礎。

In recognition of the reduction in loan volume in our platform, we unfortunately eliminated approximately 140 hourly positions within our loan operations team, representing about 7% of our workforce. This was disappointing for sure, but necessary to keep our operational capacity in line with the current environment. No other teams at Upstart were affected. We're also limiting hiring in other functions to a small number of positions that are strategic to our business.

認識到我們平台的貸款量減少，我們不幸地在我們的貸款運營團隊中取消了大約 140 個小時工職位，約占我們員工總數的 7%。這肯定令人失望，但有必要使我們的運營能力與當前環境保持一致。 Upstart 的其他團隊沒有受到影響。我們還將其他職能部門的招聘限制在少數對我們的業務具有戰略意義的職位上。

With a healthy balance sheet, robust unit economics and strong pricing power, we believe we're well-positioned to navigate an extended period of economic uncertainty while continuing to invest strategically in future growth. Despite these challenges, I'm very optimistic about Upstart's future. There's broad recognition among technology leaders and industry pundits that AI is perhaps the most transformational technology of our time, and risk-based industries such as lending are at the forefront of this incredible opportunity.

憑藉健康的資產負債表、穩健的單位經濟和強大的定價能力，我們相信我們已做好準備，在長期的經濟不確定性中度過難關，同時繼續對未來增長進行戰略性投資。儘管存在這些挑戰，但我對 Upstart 的未來非常樂觀。技術領導者和行業專家普遍認為，人工智能可能是我們這個時代最具變革性的技術，而貸款等基於風險的行業正處於這一難以置信的機遇的最前沿。

As the leader in AI-enabled lending, we are well-positioned to capitalize on these growing trends and believe that market volatility will only strengthen our position and differentiation over time. While we dislike a weakened economy as much as you do, the increase in default rates that accompany this weakness serve to train our AI models faster. While other platforms continue to retreat to serving super-prime consumers, Upstart is rapidly learning how to price and serve mainstream Americans in all market conditions.

作為人工智能貸款領域的領導者，我們有能力利用這些不斷增長的趨勢，並相信隨著時間的推移，市場波動只會加強我們的地位和差異化。雖然我們和你一樣不喜歡疲軟的經濟，但伴隨這種疲軟而導致的違約率增加有助於更快地訓練我們的 AI 模型。在其他平台繼續退居至為超級消費者提供服務的同時，Upstart 正在迅速學習如何在所有市場條件下為主流美國人定價和服務。

Beyond my conviction in AI and the impact it can have in lending, my optimism also stems from seeing the rapid progress made by each of our product and machine learning teams. Some of the areas where we're making fast progress include, first, model accuracy. Our AI models have never been more accurate relative to a traditional FICO-based model, and our pace of model development has increased significantly. To be more specific, the increase in Upstart's model accuracy in the last 4 months is as much as we saw in the prior 2 years.

除了我對人工智能的信念及其對貸款的影響之外，我的樂觀還源於看到我們每個產品和機器學習團隊取得的快速進展。我們取得快速進展的一些領域首先包括模型準確性。相對於傳統的基於 FICO 的模型，我們的 AI 模型從未像現在這樣精確，並且我們的模型開發速度顯著加快。更具體地說，Upstart 的模型準確性在過去 4 個月中的增長與我們在前 2 年看到的一樣多。

Second, macro reporting and responsiveness. An important goal for Upstart is to help lenders understand the direction our economy is trending in order to make more informed decisions about their lending programs. To support this goal, we have developed and are beginning to productize the Upstart Macro Index, or UMI. This index is a monthly indication of the state of the economy, specifically with regard to consumer financial health and credit performance.

二是宏觀報告和反應能力。 Upstart 的一個重要目標是幫助貸方了解我們的經濟發展方向，以便對其貸款計劃做出更明智的決定。為了支持這一目標，我們已經開發並開始將 Upstart Macro Index 或 UMI 產品化。該指數是經濟狀況的月度指標，特別是在消費者財務健康和信貸表現方面。

At the simplest level, UMI is designed to estimate the level of default to expect in a time period, holding underwriting models and borrowers constant. What's even more interesting is that we have determined that a handful of common economic variables present in the Dodd-Frank stress test can estimate UMI with a high degree of accuracy. We continue to iterate our methodology with the intention of translating widely available forecast of macro indicators into an expectation for future levels of default.

在最簡單的層面上，UMI 旨在估計一段時間內預期的違約水平，保持承保模型和借款人不變。更有趣的是，我們已經確定 Dodd-Frank 壓力測試中存在的少數常見經濟變量可以高度準確地估計 UMI。我們繼續迭代我們的方法，旨在將廣泛可用的宏觀指標預測轉化為對未來違約水平的預期。

We believe this is the first time that commonly understood and broadly forecasted economic indicators can predict credit performance and are looking forward to sharing more with you as we refine this tool. Our goal is to be the fastest platform to respond to macro changes and to provide the most relevant and up-to-date information to our lenders, and UMI is a big step in that direction.

我們相信這是第一次普遍理解和廣泛預測的經濟指標可以預測信貸表現，並期待在我們完善這一工具時與您分享更多信息。我們的目標是成為響應宏觀變化的最快平台，並向我們的貸方提供最相關和最新的信息，而 UMI 是朝著這個方向邁出的一大步。

Third, automation. In the third quarter, we saw a record 75% of loans fully automated. This came from a variety of efforts, including an experiment to help applicants enter information more accurately that led to an absolute 1.8% lift in instant approvals.

第三，自動化。在第三季度，我們看到了創紀錄的 75% 的貸款完全自動化。這來自各種努力，包括一項幫助申請人更準確地輸入信息的實驗，這導致即時批准的絕對提升 1.8%。

Fourth, auto refinance. This quarter saw 3 significant improvements to our auto refi product. We launched a new model to more accurately identify loan payoff amounts. We fine-tuned our income verification models. And finally, we improved the process of reviewing registration cards. These upgrades collectively led to a 20% improvement to our auto refi conversion funnel.

第四，汽車再融資。本季度我們的汽車改裝產品有 3 項重大改進。我們推出了一種新模型，以更準確地識別貸款償還金額。我們微調了收入驗證模型。最後，我們改進了註冊卡審核流程。這些升級共同使我們的自動 refi 轉換漏斗改進了 20%。

Fifth, auto retail. In Q3, we shipped our largest software release of the year, including a new build and price feature, which allows consumers to build, configure and price autos that the dealer doesn't yet have on the lot. Our software is in more than 700 dealers now. We've also turned on retail lending with 3 more dealer groups and are now in 4 states, representing 25% of the U.S. auto market by population. And more than 1 in 3 auto loan applications were automatically verified, about double the prior quarter.

五是汽車零售。在第三季度，我們發布了今年最大的軟件版本，包括一個新的構建和定價功能，它允許消費者構建、配置和定價經銷商尚未擁有的汽車。我們的軟件現已在 700 多家經銷商中銷售。我們還與另外 3 個經銷商集團開啟了零售貸款業務，目前分佈在 4 個州，占美國汽車市場人口的 25%。超過三分之一的汽車貸款申請得到了自動驗證，大約是上一季度的兩倍。

Sixth, small-dollar loans. This team shipped too many improvements to name, but in Q3, we saw more than 9,000 small-dollar loans on our platform, almost 4x the prior quarter. And all these loans were to borrowers only otherwise would have declined. Smaller and shorter-term loans are critical to reach more consumers and to help our AI models learn as quickly as possible. So we're very excited about this progress.

第六，小額貸款。這個團隊對名稱進行了太多改進，但在第三季度，我們在我們的平台上看到了超過 9,000 筆小額美元貸款，幾乎是上一季度的 4 倍。所有這些貸款都是給借款人的，否則就會下降。較小和短期的貸款對於吸引更多消費者並幫助我們的 AI 模型盡快學習至關重要。因此，我們對這一進展感到非常興奮。

And seventh, small business loans. I told you last quarter that we had reached our first $1 million in SMB loans. Well now, we're close to $10 million in loans originated, and the team is rapidly shipping improvements as we look to refine that product. With the financial impact of these upgrades to our products is muted in the current environment, we're confident that they'll set us up for a giant leap forward once the economy and credit markets normalize.

第七，小企業貸款。上個季度我告訴過你，我們已經達到了第一筆 100 萬美元的 SMB 貸款。現在，我們的貸款總額接近 1000 萬美元，團隊正在快速交付改進，因為我們希望改進該產品。在當前環境下，這些升級對我們產品的財務影響不大，我們相信一旦經濟和信貸市場正常化，它們將為我們的巨大飛躍做好準備。

Finally, while there's no shortage of caution among banks and credit unions, I'm also happy to report that we deployed a record 17 new lenders onto our platform in Q3, including Alliant Credit Union, which is a top 10 credit union by asset size. This compares to 17 lenders launched in all of 2021. While these lenders are starting up cautiously, it's encouraging that we're planting seeds for funding capacity in our future. As of today, we have 83 lenders under contract on the Upstart platform.

最後，雖然銀行和信用合作社並不缺乏謹慎，但我也很高興地報告，我們在第三季度在我們的平台上部署了創紀錄的 17 家新貸方，其中包括資產規模排名前 10 的信用合作社 Alliant Credit Union .相比之下，2021 年全年推出了 17 家貸方。雖然這些貸方起步謹慎，但令人鼓舞的是，我們正在為未來的融資能力播下種子。截至今天，我們在 Upstart 平台上擁有 83 家簽約貸方。

Before I wrap up, I want to say again, we're not pleased with the results we shared with you today. But when interest rates are rising and the economy is in flux, lenders and credit investors naturally become cautious. Despite this caution, our lenders will tell you that the performance of the Upstart-powered credit has met or exceeded expectations over time.

在結束之前，我想再說一遍，我們對今天與大家分享的結果並不滿意。但當利率上升且經濟動盪時，貸方和信貸投資者自然會變得謹慎。儘管有這種謹慎，我們的貸方會告訴您，Upstart 驅動的信貸的表現隨著時間的推移已經達到或超過了預期。

We don't like volatility any more than you do, but we won't allow it to set us off course from our long-term goal to reinvent how credit works. Our goal is to become the destination with the best rates and the best process for all forms of credit for everyone. This can't and won't be done by a single bank, but it can be done by a vast network of banks, credit unions and credit investors powered by a modern cloud-based AI platform. Great companies separate themselves from merely good ones during the hardest of times. They are clear-eyed about how the environment has changed.

我們不喜歡波動性，但我們不會讓它偏離我們重塑信貸運作方式的長期目標。我們的目標是成為為所有人提供各種形式信貸的最優惠利率和最佳流程的目的地。這不能也不會由一家銀行完成，但它可以由一個由現代基於雲的人工智能平台提供支持的龐大的銀行、信用合作社和信貸投資者網絡來完成。在最艱難的時期，偉大的公司將自己與優秀的公司區分開來。他們對環境的變化非常清楚。

They make smart and fast decisions in order to ride out the turbulence, but they also retain an optimistic focus on the horizon as they continue to invest in the future. You have my full commitment to ensure Upstart is exactly that type of company.

他們做出明智而快速的決定以渡過難關，但在繼續投資未來時，他們也保持對地平線的樂觀關注。你有我的全部承諾，以確保 Upstart 正是那種類型的公司。

Thank you. And I would like to turn it over to Sanjay, our Chief Financial Officer, to walk through our Q3 financial results and guidance. Sanjay?

謝謝你。我想把它交給我們的首席財務官 Sanjay 來介紹我們的第三季度財務業績和指導。桑傑？

Thanks, Dave. And thanks to all for joining us today. As Dave has alluded to, the external environment continues to be a challenging one, particularly for those less affluent borrowers with limited access to credit that are at the core of Upstart's mission. Consumers have simultaneously whittled personal savings rates from pre-pandemic levels of roughly 9% down to 3.3% in Q3, a level not seen since the great financial crisis, and have swelled credit card balances to all-time record highs. Savings rates have dwindled and credit card balances have inflated to pay for what has been a continuing expansion in real consumption, so far with no corresponding increase in either real wages or labor force participation since the advent of COVID. As a consequence, defaults are on the rise.

謝謝，戴夫。感謝大家今天加入我們。正如戴夫所暗示的那樣，外部環境仍然是一個充滿挑戰的環境，特別是對於那些獲得信貸的不太富裕的借款人來說，這是 Upstart 使命的核心。消費者同時將個人儲蓄率從大流行前的大約 9% 降至第三季度的 3.3%，這是自大金融危機以來從未見過的水平，並將信用卡餘額膨脹至歷史最高水平。儲蓄率下降，信用卡餘額膨脹，以支付實際消費的持續擴張，但自 COVID 出現以來，實際工資或勞動力參與率沒有相應增加。因此，違約率呈上升趨勢。

Industry-wide data shows that less affluent borrowers are leading the way with impairment levels on unsecured personal loans that are about twice as high as they were prior to the onset of COVID. By way of comparison, highly affluent borrowers are now roughly back to being in line with pre-COVID impairment levels, although they continue to be on the rise. The Upstart Macro Index previously referenced by Dave is our internal way of articulating the impact of the external macro environment on loan defaults in our particular borrower portfolio by controlling for underwriting model changes and shifting borrower characteristics over time.

整個行業的數據顯示，較不富裕的借款人處於領先地位，其無擔保個人貸款的減值水平約為新冠疫情爆發前的兩倍。相比之下，高度富裕的借款人現在大致恢復到與 COVID 之前的減值水平一致，儘管他們繼續上升。 Dave 之前引用的 Upstart 宏觀指數是我們通過控制承保模式變化和隨時間變化的借款人特徵來闡明外部宏觀環境對我們特定借款人投資組合中貸款違約影響的內部方式。

The most recent index level of around 1.7 tells us that the Q3 environment produced 70% more defaults than we would expect from our borrower base in a long-run normal macro environment. This number is also approximately 20% higher than what we had observed when we last reported earnings in August. As a result of our model's adjustments to these changing macroeconomic conditions, our loans today are being priced at APRs that are significantly higher than those from the beginning of the year, which is one of the principal driving factors behind the overall volume contraction our business is currently experiencing. As David said, this is, in fact, working as intended.

最近大約 1.7 的指數水平告訴我們，在長期正常的宏觀環境中，第三季度環境產生的違約率比我們對借款人群體的預期高出 70%。這個數字也比我們上次在 8 月份報告收益時觀察到的高出約 20%。由於我們的模型對這些不斷變化的宏觀經濟條件進行了調整，我們今天的貸款定價明顯高於年初的年利率，這是我們業務整體銷量收縮背後的主要驅動因素之一目前正在經歷。正如大衛所說，這實際上是按預期工作的。

On the loan funding side, a brief period of late summer optimism in the ABS market has since receded, and loan funding in general remains challenging. Overall financing costs for our securitization investors are up about 500 basis points since last year. These higher financing costs and the general scarcity of available capital has contributed to the volume pressure on the business. With the preceding context, here now are some of the financial highlights from the past quarter. On the top line, revenue from fees of $179 million was largely in line with our expectations. However, negative fair value adjustments and losses on sale incurred by the loans on our balance sheet brought overall net revenue down to $157 million, short of our guidance and representing a 31% contraction both sequentially and year-over-year.

在貸款融資方面，ABS 市場在夏末的短暫樂觀情緒已經消退，貸款融資總體上仍然充滿挑戰。自去年以來，我們證券化投資者的總體融資成本增加了約 500 個基點。這些較高的融資成本和可用資金的普遍稀缺性導致了業務量的壓力。在前面的背景下，現在是上一季度的一些財務亮點。最重要的是，1.79 億美元的費用收入基本符合我們的預期。然而，我們資產負債表上的貸款產生的負公允價值調整和銷售損失使整體淨收入下降至 1.57 億美元，低於我們的指導，環比和同比均收縮 31%。

The volume of loan transactions across our platform in Q3 was approximately 188,000 loans, down 48% year-over-year and representing over 125,000 new borrowers. Average loan size was up 14% versus last year. Our contribution margin, a non-GAAP metric, which we define as revenue from fees minus variable costs for borrower acquisition, verification and servicing, came in at 54% in Q3, up from 47% last quarter, but still behind our guidance. We have been successful in expanding our margins through higher take rates and more efficient marketing spend, and we expect this to continue in Q4.

第三季度我們平台上的貸款交易量約為 188,000 筆貸款，同比下降 48%，代表超過 125,000 名新借款人。與去年相比，平均貸款規模增加了 14%。我們的邊際貢獻率是一個非 GAAP 指標，我們將其定義為費用收入減去借款人獲取、驗證和服務的可變成本，在第三季度為 54%，高於上一季度的 47%，但仍低於我們的指導。我們通過更高的採用率和更有效的營銷支出成功地擴大了利潤率，我們預計這將在第四季度繼續。

Operating expenses were $215 million in Q3, down 17% sequentially. We reduced our sales and marketing by 46% sequentially to reflect a weakened conversion funnel, which has declined as a result of our higher offer rates. Engineering and product development grew 16% sequentially, and general administrative spend grew 2% sequentially. Across both areas, hiring has now largely been limited to only a few key strategic positions. Taken together, these components resulted in a Q3 GAAP net income of negative $56.2 million. Adjusted EBITDA was negative $14.4 million, and adjusted earnings per share was negative $0.24 based on a diluted weighted average share count of 81.7 million.

第三季度的運營費用為 2.15 億美元，環比下降 17%。我們連續減少了 46% 的銷售和營銷，以反映由於我們更高的報價率而下降的轉化漏斗。工程和產品開發環比增長 16%，一般行政支出環比增長 2%。在這兩個領域，招聘現在基本上僅限於幾個關鍵的戰略職位。總而言之，這些組成部分導致第三季度 GAAP 淨收入為負 5620 萬美元。調整後的 EBITDA 為負 1440 萬美元，調整後的每股收益為負 0.24 美元，基於 8170 萬股的稀釋加權平均股數。

We continue to be in a favorable liquidity position with $830 million of total cash and $431 million in net loan equity on our balance sheet. Our gross balance of loan assets at the end of the quarter was $700 million, up $76 million from last quarter. Of that total, loans made for the purposes of R&D represented $451 million, principally within the auto segment, and our balance of core personal loans stood at $249 million. The near-term outlook for our business remains tied to the direction of the macro economy. And while this has historically proven hard to predict, we are currently pricing our loans expecting a further degradation in the environment and in our macro index. The volume assumptions underpinning our revenue and earnings guidance are consistent with this outlook.

我們繼續處於有利的流動性狀況，資產負債表上有 8.3 億美元的總現金和 4.31 億美元的淨貸款權益。本季度末，我們的貸款資產總額為 7 億美元，比上一季度增加 7600 萬美元。其中，用於研發的貸款為 4.51 億美元，主要在汽車領域，我們的核心個人貸款餘額為 2.49 億美元。我們業務的近期前景仍然與宏觀經濟的方向相關。雖然這在歷史上被證明難以預測，但我們目前正在為我們的貸款定價，預計環境和我們的宏觀指數會進一步惡化。支撐我們的收入和盈利指引的數量假設與這一前景一致。

In order to provide some additional insight into revenue, we are splitting out our top line guidance between revenue from fees, which reflect our baseline volume and fee expectations; and net interest income, which includes impacts from fair value and gain on sale. With these specifics in mind, for Q4 of 2022, we will expect revenues of between $125 million and $145 million. Within that, we expect revenue from fees of approximately $160 million and net interest income of approximately negative $25 million; contribution margin of approximately 54%; net income of approximately negative $87 million; adjusted net income of approximately negative $40 million; adjusted EBITDA of approximately negative $35 million; and a diluted weighted average share count of approximately 89.3 million shares.

為了對收入提供一些額外的見解，我們將我們的收入指導分為費用收入，這反映了我們的基線數量和費用預期；淨利息收入，包括公允價值和銷售收益的影響。考慮到這些細節，我們預計 2022 年第四季度的收入將在 1.25 億美元至 1.45 億美元之間。其中，我們預計費用收入約為 1.6 億美元，淨利息收入約為負 2500 萬美元；貢獻率約為 54%；淨收入約為負8700萬美元；調整後的淨收入約為負 4000 萬美元；調整後的 EBITDA 約為負 3500 萬美元；稀釋加權平均股數約為 8930 萬股。

As ever, we will take this opportunity to extend our gratitude to all of the employees at Upstart who continue to make daily progress against our underlying business and technology goals in what continues to be a challenging external environment around us all.

與以往一樣，我們將藉此機會向 Upstart 的所有員工表示感謝，他們在我們所有人周圍仍然充滿挑戰的外部環境中繼續朝著我們的基本業務和技術目標不斷取得進步。

And with that, Dave and I are now happy to open the call to any questions. Operator, back to you.

有了這個，戴夫和我現在很高興打開電話詢問任何問題。接線員，還給你。

(Operator Instructions)

（操作員說明）

We will take our first question from David Scharf with JMP Securities.

我們將回答 JMP Securities 的 David Scharf 提出的第一個問題。

Dave or I guess for Sanjay as well, I wanted to maybe ask a little bit of a kind of longer-term strategic question as it relates to structure. Obviously, funding environment is going to go through dislocations here or there and ultimately resolve themselves. But I guess in terms of the structure of the business, I know last quarter, you talked about seeking some more kind of longer-term partners. And reflecting on some of your all-digital lending peers, there seem to be a lot of different ways to skin the cat in your industry.

Dave 或者我想 Sanjay 也是，我想問一些與結構相關的長期戰略問題。顯然，資金環境將在這里或那裡經歷錯位並最終自行解決。但我想就業務結構而言，我知道上個季度，您談到了尋找更多類型的長期合作夥伴。回顧您的一些全數字借貸同行，似乎有很多不同的方法可以在您的行業中剝皮。

LendingClub went out and got a bank charter. Pagaya's gone for pre-funding securitizations in investment vehicles exclusively. LendingPoint, they've always kind of opted for a 60-40 mix between loan retention and securitization. Obviously, as you noted, the macro environment is going to shift and, ultimately, will emerge on the other side. But in terms of strategically thinking about the types of dislocations that are happening right now, is it different longer-term funding structure something the company evaluates every now and then?

LendingClub 出去並獲得了銀行執照。 Pagaya 專門為投資工具的證券化預先融資。 LendingPoint，他們總是在貸款保留和證券化之間選擇 60-40 的組合。顯然，正如您所指出的，宏觀環境將發生變化，最終將出現在另一邊。但就目前正在發生的錯位類型的戰略思考而言，公司不時評估的長期融資結構是否有所不同？

This is Dave. That's a good question. We certainly think about funding on our platform pretty much constantly. But I will say this, we believe fundamentally in a marketplace structure in the sense that a lot of lenders making independent decisions over the long haul is going to get to the right answer. I mean marketplace -- market-based economies are historically far more efficient than centrally planned economies. That's a very -- I would just say a very basic truism.

這是戴夫。這是個好問題。我們當然會經常考慮在我們的平台上提供資金。但我要說的是，我們從根本上相信市場結構，因為從長遠來看，許多做出獨立決策的貸方將得到正確的答案。我的意思是市場——以市場為基礎的經濟在歷史上比中央計劃經濟更有效率。這是一個非常 - 我只想說一個非常基本的真理。

But having said that -- so that means we don't want to become a centrally planned economy. We don't believe us being a bank makes a lot of sense for what we hope to pursue for lots of reasons. But having said that, we can certainly do a better job of securing supply of funding on our platform. And that can really be through some of the things we talked about: getting longer-term funding agreements in place; being in more products -- a more diverse set of products, such as secured products like auto loans, mortgages, et cetera.

但是話雖如此——這意味著我們不想成為一個中央計劃經濟體。我們不相信我們成為一家銀行對於我們出於多種原因希望追求的東西有多大意義。但話雖如此，我們當然可以更好地確保我們平台上的資金供應。這真的可以通過我們談到的一些事情來實現：簽訂長期資助協議；涉足更多產品——更多樣化的產品組合，例如汽車貸款、抵押貸款等擔保產品。

So it is certainly something we have to think hard about and do more work on. But underneath it all, we do believe a market-based economy -- a marketplace where there's a lot of participants on both sides will ultimately have the greatest scale and the greatest opportunity. Albeit we're dealing with volatility today, but over the long haul, we're confident this will lead to the greatest outcome for Upstart.

所以這當然是我們必須認真思考並做更多工作的事情。但在這一切之下，我們確實相信以市場為基礎的經濟——一個雙方都有很多參與者的市場，最終將擁有最大的規模和最大的機會。儘管我們今天正在應對波動，但從長遠來看，我們相信這將為 Upstart 帶來最大的成果。

Got it. No, I appreciate the color, Dave. And maybe just as a follow-up, digging a little deeper on the funding side. Obviously, as Sanjay noted, I mean, the ABS markets remain a bit volatile. But at the same time, even though spreads are wider, we've actually seen in the last couple weeks, months, a number of non-prime deals start to get done: [Inova,] Opportune, Regional Management, so spreads remain wider, but investors are stepping up for the unsecured personal loan non-prime asset class. Any further updates you can provide based on -- either anecdotal, on discussions you have with existing bond investors or when you would think you might be able to return to the market?

知道了。不，我很欣賞這種顏色，戴夫。也許只是作為後續行動，在資金方面更深入地挖掘。顯然，正如桑傑所說，我的意思是，ABS 市場仍然有些波動。但與此同時，儘管價差更大，但我們實際上在過去幾週、幾個月中看到，一些非主要交易開始完成：[Inova，] Opportune，區域管理，因此價差仍然更大，但投資者正在加緊購買無抵押個人貸款非優質資產類別。您可以根據軼事，與現有債券投資者的討論或您認為何時可以重返市場提供任何進一步的更新？

David, this is Sanjay. As you said, it's volatile. We remain in the market. We completed a couple of deals in Q3, and we're going to be back in the market in Q4. Our cadence is generally every sort of 2 to 3 months or so, and I think we've been holding to that cadence. So like you said, cost of funds, spreads are all pretty volatile, and they will sort of dictate the economics in any given deal. But there's always deals to be done or at least until now, there's deals to be done. So we're going to continue with that cadence. And we have a stable of investors who are contributing to those securitizations who are -- continue to have interest in contributing the collateral into the securitization as well.

大衛，這是桑傑。正如你所說，它是易變的。我們留在市場上。我們在第三季度完成了幾筆交易，我們將在第四季度重返市場。我們的節奏通常是每一種 2 到 3 個月左右，我認為我們一直保持這種節奏。所以就像你說的，資金成本、點差都非常不穩定，它們會在某種程度上決定任何特定交易的經濟性。但總有交易要做，或者至少到現在為止，還有交易要做。所以我們將繼續保持這種節奏。我們有一個穩定的投資者，他們正在為那些證券化做出貢獻，他們仍然有興趣將抵押品貢獻給證券化。

We will take our next question from Ramsey El-Assal with Barclays.

我們將向巴克萊銀行的 Ramsey El-Assal 提出我們的下一個問題。

I wanted to ask about the on-balance sheet loans. It looks like that number went up about $70 million, $80 million this quarter to around $700 million. I'm just curious in terms of going forward what your plans are there. Do you intend to stabilize that number here? Or ship will go up, will go down? How should we kind of think about that for modeling purposes?

我想問一下表內貸款。看起來這個數字增加了大約 7000 萬美元，本季度從 8000 萬美元增加到 7 億美元左右。我只是好奇你的計劃是什麼。你打算在這裡穩定這個數字嗎？還是船會上升，會下降？出於建模目的，我們應該如何考慮這一點？

Ramsey, this is Sanjay. Yes, I don't think we've necessarily guided a specific guideline or a number with respect to our balance sheet. I think we gave some sort of high-level parameters last quarter, and I think we obviously operated within that. And I think that, that will continue to be the case. So I think that whether we draw it up or draw it down over the next quarter or so will continue to be an operating decision we sort of discuss and take, but I think it will be within the parameters of what you saw in this last quarter.

拉姆齊，這是桑傑。是的，我認為我們不一定就資產負債表制定特定的指導方針或數字。我認為我們在上個季度給出了一些高級參數，我認為我們顯然是在這個範圍內運作的。我認為，情況將繼續如此。因此，我認為我們是否在下個季度左右制定或縮減將繼續是我們討論和採取的運營決策，但我認為這將在您在上個季度看到的參數範圍內.

Got it. Okay. And then a quick follow-up for me just on the conversion rate on the rate request. I think that, as you mentioned, it's a tougher environment. That trended, I think, down 300-plus basis points quarter-over-quarter. Also there, just curious from a modeling perspective, do we keep that sort of stable here? Or is that a metric that we could see deteriorate further? Or is the answer it's just contingent on the environment and how it evolves?

知道了。好的。然後對我快速跟進費率請求的轉換率。正如你所提到的，我認為這是一個更艱難的環境。我認為，這種趨勢環比下降了 300 多個基點。還有，只是從建模的角度好奇，我們在這裡保持那種穩定嗎？或者這是一個我們可以看到進一步惡化的指標？或者答案只是取決於環境及其演變方式？

Yes. Thanks, Ramsey. It's a good question. I mean, I think that I would bring it back to the vocabulary of this Upstart Macro Index, which Dave referenced, which, as we said, it's something we've started disclosing in our investor materials. It essentially is sort of an index to try and capture the external macro's impact on defaults. And the simple way to think about our conversion rate is that index went up about 20% versus last quarter. So that's sort of an expression of the fact that the macro is impacting defaults in our portfolio by that amount.

是的。謝謝，拉姆齊。這是個好問題。我的意思是，我認為我會將它帶回到 Dave 引用的 Upstart Macro Index 的詞彙中，正如我們所說，這是我們已經開始在我們的投資者材料中披露的內容。它本質上是一種嘗試捕捉外部宏對默認值的影響的索引。考慮我們的轉化率的簡單方法是，該指數與上一季度相比上漲了約 20%。因此，這在某種程度上表達了宏觀影響我們投資組合中的違約金額這一事實。

And when that happens, our models recalibrate, APRs go up, and essentially approval rates and acceptance rates both go down. So in terms of how to think about it on the go forward, it really kind of amounts to what you think about the macro conditions. And if defaults are going to continue to go up or normalize or stabilize or maybe even reverse course at some point, that will really dictate the offers that we're making and hence, the conversion funnel.

當這種情況發生時，我們的模型會重新校準，APR 會上升，基本上批准率和接受率都會下降。因此，就如何在未來進行思考而言，這實際上相當於您對宏觀條件的看法。如果默認值將繼續上升或正常化或穩定，甚至可能在某個時候逆轉，這將真正決定我們正在提供的報價，從而決定轉換漏斗。

We will take our next question from Peter Christiansen with Citi.

我們將向花旗銀行的 Peter Christiansen 提出我們的下一個問題。

I wanted to ask about, again, back to rate request, the previous question. It looks like they were down considerably sequentially, not even looking at the conversion rate yet, but at the top -- the very top of the funnel. Just wondering, are you taking a different go-to-market approach in terms of attracting new borrowers to the platform? And how should we think about that in context to this potential rise in like debt consolidation and those kind of themes? Just wondering how your go-to-market is changing their top-of-funnel new borrower adds.

我想再次詢問上一個問題的評分請求。看起來他們連續下降了很多，甚至還沒有看到轉化率，而是在頂部——漏斗的最頂端。只是想知道，在吸引新借款人到平台方面，您是否採取了不同的上市方式？在債務合併和此類主題的潛在增長背景下，我們應該如何考慮這一點？只是想知道您的市場營銷如何改變他們的漏斗頂部新借款人添加。

Pete, this is Sanjay. Go ahead, Dave.

皮特，這是桑傑。去吧，戴夫。

No. Go ahead. Go ahead, Sanjay.

沒有，繼續。去吧，桑傑。

So I was just -- I'll add, and I'm sure Dave can add to the response. But a lot of our marketing activities are in fact governed by the conversion funnel in the sense that when loss estimates go up, conversions go down. We target our marketing campaign size and our activities to some target for unit economics. And when the conversion funnel is weaker, we reduce marketing size accordingly. So as all of this sort of like this 2-step process, whereby conversion funnel improves, it converts more. And we expand marketing and then the reverse is true, of course, when the marketing funnel contracts.

所以我只是- 我會補充一下，我相信Dave 可以補充一下。但實際上，我們的許多營銷活動都受轉化漏斗的支配，因為當損失估計上升時，轉化率就會下降。我們將我們的營銷活動規模和我們的活動定位到單位經濟的某個目標。當轉化漏斗較弱時，我們會相應地減少營銷規模。因此，由於所有這些都類似於這個兩步過程，因此轉化漏斗得到改善，它轉化得更多。當然，當營銷漏斗收縮時，我們擴大營銷，然後反過來也是如此。

I think the genesis of the question is like, how are you thinking about spend on lead gen generally?

我認為問題的起源是，您一般如何考慮在潛在客戶上的支出？

The spend on the lead gen is sort of an -- it's sort of a function of how performing our conversion funnel is. So we will spend up until the point where the marginal return on the marketing dollar is zero in a sense. And so sort of what I mean is if our conversion funnel improves, we will spend more on lead gen because it will be more economical. We'll be able to spend more up until the point where the marginal cost is zero and vice versa. And so what's been happening recently is as our conversion funnel is impacted by higher losses in the portfolio, our target sort of unit economics contracts.

潛在客戶的支出有點像 - 它是我們轉換漏斗執行情況的函數。因此，我們將一直花費到營銷美元的邊際回報在某種意義上為零。所以我的意思是，如果我們的轉化漏斗得到改善，我們將在潛在客戶上花費更多，因為它會更經濟。我們將能夠花費更多，直到邊際成本為零，反之亦然。因此，最近發生的事情是我們的轉換漏斗受到投資組合中更高損失的影響，我們的目標類型單位經濟合同。

We will take our next question from Mike Ng with Goldman Sachs.

我們將向高盛的 Mike Ng 提出下一個問題。

I just had 2. First, I was just wondering if you could tell us what the transaction fee rate as a percentage of fund in principal was in the quarter, obviously, ex servicing fees and how we should think about the opportunity to continue to take pricing going forward? And then secondly, I was wondering if you could talk about how much of the principal in the quarter was self-funded off of the Upstart balance sheet versus just the core model?

我剛買了 2 個。首先，我只是想知道您能否告訴我們本季度的交易費率佔本金的百分比是多少，顯然，前服務費以及我們應該如何考慮繼續採取的機會未來定價？其次，我想知道你是否可以談談本季度有多少本金是從 Upstart 資產負債表中自籌資金的，而不僅僅是核心模型？

Mike, this is Dave. I'll take the first part of the question and, Sanjay, perhaps answer the second. In the first, we have -- we're able to basically increase revenues by increasing fees on a per-loan basis. And then also, as Sanjay was suggesting earlier, our acquisition spend per loan at a time like this can go down a lot. So in effect, the unit economics on each loan is significantly better, much more sort of gross profit per loan, albeit at a lower loan volume.

邁克，這是戴夫。我會回答問題的第一部分，桑傑，也許會回答第二部分。首先，我們能夠通過增加每筆貸款的費用來增加收入。然後，正如桑傑早些時候建議的那樣，我們在這樣的時間每筆貸款的收購支出可能會下降很多。因此，實際上，每筆貸款的單位經濟學要好得多，每筆貸款的毛利潤要多得多，儘管貸款量較低。

So those are things within our control, which is why contribution margin has gotten higher during this time. And I think that's sort of a form of pricing power that means we can -- when we need to be in a little bit of a defensive mode can make sure that we're monetizing well enough to cover our expenses, et cetera. And we view that as a very positive part of our platform.

所以這些都是我們控制範圍內的事情，這就是為什麼這段時間邊際貢獻變得更高的原因。而且我認為這是一種定價能力的形式，這意味著我們可以——當我們需要處於一點防禦模式時，可以確保我們的盈利能力足以支付我們的費用等等。我們認為這是我們平台的一個非常積極的部分。

We will take our next question from Simon Clinch with Atlantic Equities.

我們將向大西洋股票的西蒙克林奇提出下一個問題。

I was wondering if we could just start with the contribution margin. I was just wondering if you could talk about, I guess, how -- because the guidance for 59% in the quarter, you've fallen quite short of that. But it also implies, I think you're tracking well above that as probably entering the quarter. So I'm just kind of curious about the dynamics that go into contribution margins that are within your control and not within your control? And how to think about that going forward?

我想知道我們是否可以從邊際貢獻開始。我只是想知道你是否可以談談，我猜，如何 - 因為本季度 59% 的指導，你已經遠遠沒有達到這個目標。但這也意味著，我認為您的跟踪可能遠高於該季度。所以我只是對在您的控制範圍內而不是在您的控制範圍內的邊際貢獻的動態感到好奇？以及如何考慮未來的發展？

This is Sanjay. Yes, that's a great question. So I guess the dynamics on the contribution margin are, it was up from last quarter, 47% to approximately 54%. It was sort of our guidance. Maybe one simple way to think about it is, when we are funding-constrained as a platform, we tend to expand contribution margins. And we do that by expanding take rates. And it makes sense to do that when you're funding-constrained.

這是桑傑。是的，這是一個很好的問題。所以我猜邊際貢獻的動態是，它從上一季度的 47% 上升到大約 54%。這是我們的指導。也許一種簡單的思考方式是，當我們作為一個平台資金有限時，我們傾向於擴大貢獻邊際。我們通過擴大收取率來做到這一點。當您資金有限時，這樣做是有意義的。

Now when you're borrower-constrained, you sort of do the opposite. You want to sell for volume, and you can take down your take rates a little bit and expand volume. And I think we probably assumed we'd be funding-constrained for all of this past quarter. In reality, we've sort of bounced back and forth a little bit. We've been at times funding-constrained and at times borrower-constrained. And at those times where we've been borrower-constrained, we've actually acted to reduce contribution margins a little bit. And so I think that we are sort of bouncing around between those 2 states. And as we go into Q4, to the extent we are funding-constrained in any given period of time, our contribution margins would be above the numbers that we produced and probably closer in line to what we had guided.

現在，當您受到借款人的限制時，您會做相反的事情。你想按量出售，你可以稍微降低你的收取率並擴大交易量。而且我認為我們可能假設我們在過去一個季度的所有時間裡都會受到資金限制。實際上，我們有點來回反彈。我們有時會受到資金限制，有時也會受到借款人的限制。在我們受到借款人限制的時候，我們實際上已經採取了一些行動來降低邊際貢獻。所以我認為我們在這兩個州之間來回擺動。當我們進入第四季度時，如果我們在任何給定時間段內都受到資金限制，我們的邊際貢獻將高於我們產生的數字，並且可能更接近我們的指導。

But to the extent we are borrower-constrained and, again, the borrower constraints really come from the fact that our macro index is so high that the approval rates are low, you'll see sort of lower conversion rates more in line with how we looked in Q2 probably. So the outcome is sort of a function of where we are between those 2 states.

但就我們受到借款人約束的程度而言，借款人約束確實來自於我們的宏觀指數如此之高以至於批准率很低的事實，你會看到較低的轉化率更符合我們的方式可能在第二季度看。所以結果有點像我們在這兩個狀態之間的位置的函數。

Okay. Appreciate that. And I was wondering if you could -- just jumping back to the structure of your funding and sort of the appetite of the investor base at the moment. I mean, could you go a little bit -- I mean, how far have you got in terms of exploring the idea of having -- of shifting the base towards more long-term investors? And maybe you could just update about your thinking on that.

好的。感謝。我想知道你是否可以 - 只是回到你的資金結構和目前投資者基礎的胃口。我的意思是，你能不能稍微——我的意思是，你在探索擁有的想法方面走了多遠——將基礎轉移到更長期的投資者身上？也許你可以更新一下你對此的想法。

Sorry, could you just repeat the very last part of your question? What is our thinking in terms of what?

對不起，你能重複你問題的最後一部分嗎？我們的想法是什麼？

About the shift towards finding more longer-term capital-type investors.

關於尋找更多長期資本型投資者的轉變。

I see. Well, just to describe it at a high level, I mean, historically, we've been sort of like 3/4 institutionally funded and about 1/4 bank-funded. That ratio has changed. And it's -- the percentage of bank funding on our platform has done quite a bit. It's been a more stable source of capital. The institutional world has obviously been a lot more volatile. And then within the institutional side, I think there's a desire, as we talked to last quarter to enter into some more strategic transactions, some more sort of committed sort of type partnerships.

我懂了。好吧，只是從高層次上描述它，我的意思是，從歷史上看，我們有點像 3/4 的機構資助和大約 1/4 的銀行資助。這個比例發生了變化。它是 - 我們平台上的銀行資金百分比已經做了很多。這是一個更穩定的資金來源。制度世界顯然更加動盪。然後在機構方面，我認為有一種願望，正如我們上個季度所說的那樣，我們希望進行一些更具戰略性的交易，一些更加承諾的類型的合作夥伴關係。

I would just say we're in a number of encouraging conversations, but they're all quite preliminary. And I think we view that as being something that's not going to happen overnight. It's something that -- to the extent we get into those types of partnerships, they're not out of a sense of urgency. It's more about doing the right thing for the future of the platform. And I think those partnerships are available, but they may take some time to put into place because they're important and large and strategic. So nothing more concrete than that to report on that right now, but I think we're pretty encouraged at the opportunities that are out there.

我只想說我們正在進行一些令人鼓舞的對話，但它們都是初步的。我認為我們認為這不會在一夜之間發生。就我們進入這些類型的伙伴關係而言，它們並不是出於緊迫感。更多的是為平台的未來做正確的事情。我認為這些合作夥伴關係是可用的，但它們可能需要一些時間才能落實到位，因為它們很重要、規模龐大且具有戰略意義。所以沒有比現在更具體的報告了，但我認為我們對那裡的機會感到非常鼓舞。

We will go next to James Faucette with Morgan Stanley.

我們將與摩根士丹利一起前往 James Faucette。

This is Sandy Beatty on for James. Question on the forward flow funding process. How volatile is that month to month? So you mentioned some summer optimism. And I'm trying to get a sense of how volatile is that, what factors determine that volatility. Is it cost of capital, opinions about future credit performance? How do those conversations typically go? And is there any insight that you can provide us into that process?

這是詹姆斯的桑迪比蒂。關於正向流動資金流程的問題。那個月的波動性如何？所以你提到了一些夏天的樂觀情緒。我試圖了解這種波動性有多大，是什麼因素決定了這種波動性。是資本成本，還是對未來信用表現的看法？這些對話通常如何進行？您是否可以為我們提供有關該過程的任何見解？

Sure. Yes. This is Sanjay, again. I wouldn't necessarily characterize it as volatile insofar as volatile as ebbs and flows. I really would characterize it as a level of funding that degraded pretty steadily, let's call it, between, I'd say, March and August or July/August. And since then, it's been at a pretty stable level although obviously, 1 month that it was much lower than earlier in the year.

當然。是的。這又是桑傑。我不一定將它描述為易變的，只要像潮起潮落一樣易變。我真的會將其描述為相當穩定地下降的資金水平，我們稱之為，在我想說的 3 月和 8 月或 7 月/8 月之間。從那以後，它一直處於一個相當穩定的水平，儘管很明顯，1 個月它比今年早些時候要低得多。

So since then, it's not like there's been a bunch of comings and goings. So I think there's -- you could sort of characterize it as there's been a number of partners of ours -- funding partners of ours, predominantly with those who have worked with us for a longer period of time, who have been steady and stable in their activity. And there's been a lot of people who -- maybe those who have been working with us for a lesser period of time and are a little bit more sort of dependent on the ABS markets. And they've largely sat on the sidelines as they're sort of waiting to see how the world plays out.

所以從那以後，就好像沒有一堆來來去去的。所以我認為有——你可以把它描述為我們的許多合作夥伴——我們的資助合作夥伴，主要是那些與我們合作了更長時間的人，他們在他們的活動。並且有很多人 - 也許那些與我們合作的時間較短並且更依賴於 ABS 市場的人。他們基本上都在觀望，因為他們有點等著看世界如何發展。

Got it. And then just a follow-up on profitability in terms of the quarter and also the guidance. How are you thinking about managing that cost structure going forward? And how should we think about impact from the recent workforce reductions and how that might reduce pressures or reduce costs on a run rate basis?

知道了。然後只是就本季度的盈利能力和指導進行跟進。您如何看待未來管理該成本結構？我們應該如何考慮最近裁員的影響，以及這可能如何減少壓力或降低運行率的成本？

Sure. Yes. I mean, I think of the components of our cost structure, we think about the contribution margin, obviously, and we sort of guided that at a level next quarter, that's, I'd say, comparable to where we are now, so something in the mid-50s. The reduction in force that we did really will have -- as that runs its course, a positive impact on contribution margins because really, that was about rightsizing the size of the onboarding team that's processing the incoming loans to be a bit more in line with the volumes that we have.

當然。是的。我的意思是，我想到了我們成本結構的組成部分，我們考慮了邊際貢獻，很明顯，我們在下個季度的某個水平上指導了這一點，我想說，與我們現在的水平相當，所以有些東西在 50 年代中期。我們確實會減少力量——隨著它的發展，對邊際貢獻產生積極影響，因為實際上，這是關於調整處理傳入貸款的入職團隊的規模，使其更符合我們擁有的捲。

With respect to the sort of what we call the fixed sort of payroll between our engineering and technology teams and our general and administrative teams, there, we've pretty much paused hiring except for a couple of very strategic roles that are important to sell. And so I think you could expect to see a stable sort of fixed OpEx base. And that's something that we continue to sort of evaluate every quarter.

關於我們所謂的工程和技術團隊以及我們的綜合和行政團隊之間的固定工資單，除了幾個非常重要的戰略角色外，我們幾乎已經暫停了招聘。因此，我認為您可以期望看到一種穩定的固定 OpEx 基礎。這是我們每個季度都會繼續評估的東西。

And I think we like the size of that OpEx base given where we are now. And obviously, the world can take a number of different directions. And if we start to recover, we'll be in good shape. If we continue to degrade, we'll continue to sort of evaluate as we go. But beyond what we've done with the existing reduction in force, there's no plans in place to go any further at this time.

鑑於我們現在所處的位置，我認為我們喜歡運營支出基數的規模。顯然，世界可以採取許多不同的方向。如果我們開始恢復，我們將處於良好狀態。如果我們繼續退化，我們將繼續進行評估。但除了我們在現有的武力削減方面所做的工作之外，目前還沒有進一步的計劃。

We will take our next question from Vincent Caintic with Stephens.

我們將回答 Vincent Caintic 和 Stephens 的下一個問題。

First question actually on the Slide 12 in your deck, the Upstart platform performance versus target is recovering slide. Just wondering if you could talk about that in more detail. I'm seeing that that's improved, just wondering if -- essentially wondering when you think maybe investor demand can come back or volume demand can come back if the customer is being priced at an 800 basis points higher and your investors are seeing 500 basis points higher. Is there -- when do you think we can maybe see visibility into improving demand, whether it's on the credit demand side or in the borrower demand side?

第一個問題實際上是在您的套牌中的幻燈片 12 上，Upstart 平台性能與目標的對比正在恢復幻燈片。只是想知道您是否可以更詳細地討論一下。我看到情況有所改善，只是想知道如果客戶的定價高出 800 個基點而您的投資者看到 500 個基點，您是否認為投資者需求可能會恢復或數量需求何時會恢復更高。有沒有 - 你認為我們什麼時候可以看到需求改善的可見性，無論是在信貸需求方面還是在藉款人需求方面？

Vincent, this is Sanjay. It is a great question in the sense it's a million-dollar question. I mean, I think the return of sort of confidence and funding in the institutional side sort of requires a convergence of those 2 lines. And in a sense, we're chasing a trend that's on the prior slide -- or I guess it would be on Slide 10, the Upstart Macro Index, like that's the thing that is the moving target for us. And our models are calibrating as that is evolving.

文森特，這是桑傑。從某種意義上說，這是一個價值百萬美元的問題，這是一個很好的問題。我的意思是，我認為在機構方面恢復信心和資金需要這兩條線的融合。從某種意義上說，我們正在追逐上一張幻燈片上的趨勢——或者我猜它會出現在幻燈片 10 上，即新貴宏觀指數，就像這對我們來說是移動目標一樣。我們的模型正在校準，因為它正在發展。

And so compared to where we thought -- in terms of those 2 lines converging where we thought we were last quarter, it turns out we're on the lower side of our confidence interval now because the defaults in the world has continued to rise. And so it really is a question of when will that default trend stabilize. And as and when it does, you will see those models converge quite quickly.

因此，與我們認為的情況相比——就這兩條線在我們認為上個季度的匯合點而言，事實證明我們現在處於置信區間的較低端，因為世界上的違約率一直在上升。因此，這確實是一個默認趨勢何時會穩定下來的問題。當它發生時，你會看到這些模型很快收斂。

And in fact, the target returns themselves have gone up. I should clarify, this is obviously something that blends the returns and the performance of all of our loans, whether they're on the bank channel side or the institutional side. But if you look at the return targets on the institutional side, that's really where that the 500 basis point sort of number has come in. So yes, it's a bit down to how the macro evolves from here and how conservatively we're pricing with their models. So we've signaled confidence in where we are now and where we're pricing loans now, but obviously, the world needs to play out a little bit so we can demonstrate that.

事實上，目標回報本身已經上升。我應該澄清一下，這顯然融合了我們所有貸款的回報和表現，無論它們是在銀行渠道還是機構方面。但是，如果您查看機構方面的回報目標，這確實是 500 個基點的數字所在。所以，是的，這有點取決於宏觀如何從這裡演變，以及我們定價的保守程度他們的模型。因此，我們對我們現在所處的位置以及我們現在為貸款定價的位置表示了信心，但顯然，世界需要發揮一點作用，這樣我們才能證明這一點。

Okay. Understood. And then a follow-up question just on the costs. I guess when we're thinking about the long term and thinking about some of the investments in new products that you're making versus maybe in the near term where maybe there's not a lot of volume or trying to be conservative on expenses, how do you kind of manage the two? Because it does sound like you're building out the small business portfolio, building out auto lending. Maybe there's some traction going on there. But maybe you could help us understand when do you -- when would you -- the balance between conservative on the investments and on the expenses versus long-term opportunities with these products?

好的。明白了。然後是關於成本的後續問題。我想當我們考慮長期並考慮對您正在製作的新產品的一些投資與可能在短期內可能沒有很多數量或試圖保守開支時，如何你管理這兩個？因為這聽起來確實像是您正在建立小型企業投資組合，建立汽車貸款。也許那裡有一些牽引力。但也許你可以幫助我們了解你什麼時候——你什麼時候——在投資和費用上的保守與這些產品的長期機會之間取得平衡？

Yes. This is Dave. Vincent, the way we think about that is we would like to, to the extent possible, continue to invest or even increase investment in the future products because that's obviously what our franchise is built on and what will lead to significant growth in the future. So what we've been able to do is maintain that growth and actually continue to invest in the products. And a lot of that, we can see internally. I've shared some of the metrics with it in terms of actual improvements made to each of the products, but we don't actually benefit from them until really the funding and the economic situation is on a better footing.

是的。這是戴夫。文森特，我們的想法是，我們希望在可能的範圍內繼續投資甚至增加對未來產品的投資，因為這顯然是我們的特許經營權的基礎，並且將導致未來的顯著增長。因此，我們能夠做的是保持這種增長，並實際上繼續投資於產品。其中很多，我們可以在內部看到。我已經與它分享了一些關於對每種產品進行實際改進的指標，但在資金和經濟狀況真正處於更好的基礎上之前，我們實際上並沒有從中受益。

So we're a little bit building toward the future. But I think the good news is we have not cut back on that investment in the future of our products. And when I think we're in a more normalized environment, we will very quickly see the benefit of things. Just by way of example, we have the highest-ever rate of automated loans. 75% of the loans on our platform in Q3 had no human intervention in them, and that's a record high for us. We're not really benefiting from that as a business until we get to a place where loan funding -- when loan prices aren't so high, loan funding is abundant, et cetera. And I think across the board, if you looked at each of our products, they're actually getting better very quickly. And the teams are making very good use of this time, though the payback won't be until some point in the future.

因此，我們正在為未來建設一點點。但我認為好消息是我們並沒有削減對我們產品未來的投資。當我認為我們處於一個更加規範化的環境中時，我們將很快看到事情的好處。僅舉例來說，我們擁有有史以來最高的自動貸款率。第三季度我們平台上 75% 的貸款沒有人為乾預，這對我們來說是歷史新高。作為一家企業，我們並沒有真正從中受益，直到我們獲得貸款融資的地方——當貸款價格不那麼高時，貸款資金充足，等等。而且我認為全面，如果您查看我們的每個產品，它們實際上會很快變得更好。團隊很好地利用了這段時間，儘管要等到未來某個時候才能得到回報。

We will take our next question from Arvind Ramnani with Piper Sandler.

我們將回答 Arvind Ramnani 和 Piper Sandler 的下一個問題。

I just had a couple of questions. One, just as you think about the next kind of 12 months, what are some of the downside scenarios? Like, I mean, if the macro gets a lot worse, would you expect like kind of further deterioration in your business just given sort of the strong exposure you'll have to the macro?

我只是有幾個問題。一，正如您考慮接下來的 12 個月一樣，有哪些不利情況？就像，我的意思是，如果宏觀變得更糟，你是否會期望你的業務會進一步惡化，只是考慮到你對宏觀的強烈敞口？

Sure. Arvind, well, look, no doubt, any business looking to the future of the economy, there are downside scenarios for everybody. We're not different than that. I mean we're a fairly simple business in many ways that we have fixed costs, and then we have contribution margin to offset those. And certainly, if macro continue to deteriorate significantly, that would probably translate into lower volumes in our platform. And at some point, we will look at our fixed costs and ask whether we can afford that.

當然。 Arvind，好吧，看，毫無疑問，任何展望經濟未來的企業，每個人都有不利的情景。我們和那沒有什麼不同。我的意思是我們在很多方面都是一個相當簡單的企業，我們有固定成本，然後我們有邊際貢獻來抵消這些成本。當然，如果宏觀繼續顯著惡化，那可能會轉化為我們平台的交易量減少。在某個時候，我們會查看我們的固定成本並詢問我們是否能負擔得起。

So our first goal is, of course, retain solvency in the sort of solid footing the company is on. We have a large cash balance. We have relatively low fixed costs, and that's really helped us all through our existence. But -- so we don't have any fear other than, look, the thing we want to keep doing and thus far, we're able to do so, is investing in the products.

因此，我們的首要目標當然是在公司所處的穩固基礎上保持償付能力。我們有大量現金餘額。我們的固定成本相對較低，這確實幫助了我們所有人的存在。但是 - 所以我們沒有任何恐懼，除了，看，我們想要繼續做的事情，到目前為止，我們能夠這樣做，就是投資產品。

Certainly, there are scenarios we could imagine that are so bad that we would have to cut back investment or pause products, et cetera. But we don't see that today. I think today, we have enough volume and enough contribution margin to keep optimistically investing for the future, and that's what we would hope.

當然，我們可以想像有些場景非常糟糕，以至於我們不得不削減投資或暫停產品等等。但我們今天看不到這一點。我認為今天，我們有足夠的數量和足夠的邊際貢獻來繼續樂觀地投資未來，這就是我們所希望的。

Right. And as you think of your kind of, I would say, kind of existing cash burn and sort of projected cash burn, in which country you're making adjustments with -- just from some of your expense line, when do you think you may need to go sort of raise kind of additional capital, whether in the form of equity or debt?

正確的。當你想到你的那種，我會說，現有的現金消耗和預計的現金消耗，你在哪個國家進行調整——只是從你的一些支出線來看，你認為什麼時候可以需要去籌集某種額外的資本，無論是以股權還是債務的形式？

We don't see any need to do that, Arvind. And honestly, our cash burn today is quite small even in the very constricted position we're in. I mean, I think our volumes are pretty dramatically lower than they were, yet our cash burn is fairly minimal. So we don't see a scenario where we have to raise cash. As Sanjay said, we have over $800 million in cash as well as loan assets on the balance sheet. So that's just not something we anticipate at this time.

我們認為沒有必要這樣做，Arvind。老實說，即使在我們所處的非常狹窄的位置，我們今天的現金消耗也很小。我的意思是，我認為我們的交易量比以前要低得多，但我們的現金消耗卻相當小。所以我們沒有看到我們必須籌集現金的情況。正如桑傑所說，我們在資產負債表上擁有超過 8 億美元的現金和貸款資產。所以這不是我們目前所期望的。

I was just going to maybe put some quick back-of-the-envelope numbers to that. Our cash sort of fixed expense burn across payroll and OpEx every month is about $30 million. And even at zero origination scenario, we're getting a servicing stream of revenue that's about $15 million. So there's sort of -- maybe the sort of $15 million delta every month that we have to rely on contribution margin for to cover. So that's sort of like on a downside scenario where the gap might be. And as you saw, we've got about $800 million in total cash on the balance sheet. So as Dave said, that can take us for quite some runway.

我只是可能會在其中添加一些快速的粗略數字。我們每月在工資單和運營支出中消耗的現金類固定費用約為 3000 萬美元。即使在零起源的情況下，我們也獲得了大約 1500 萬美元的服務收入流。因此，我們必須依靠邊際貢獻來支付每月 1500 萬美元的增量。所以這有點像在可能存在差距的不利情況下。正如你所看到的，我們的資產負債表上有大約 8 億美元的現金。正如戴夫所說，這可以讓我們走上一段跑道。

Right. Right. I guess you can be quite patient in that case. Yes, that's pretty much the questions I had.

正確的。正確的。我想在這種情況下你可以很有耐心。是的，這幾乎就是我的問題。

We will take our next question from David Chiaverini with Wedbush Securities.

我們將回答 Wedbush Securities 的 David Chiaverini 的下一個問題。

So I'm looking on Slide 11, the in-period losses versus expectations. Can you walk through what this is telling us? Is it basically saying that 25% -- that loss in-period defaults are 25% above what you're modeling it? And I guess marrying that with Slide 10 with the UMI at 1.7, should we expect this line on Page 11 to go up towards 70%? Just could you talk through that a little bit?

因此，我正在查看幻燈片 11，即期內虧損與預期的對比。你能解釋一下這告訴我們什麼嗎？基本上是說 25% - 期間違約損失比你建模的高 25%？我猜想把它與幻燈片 10 和 1.7 的 UMI 結合起來，我們是否應該期望第 11 頁上的這條線上升到 70%？你能不能稍微說一下？

David, this is Sanjay. Sure. These are important questions. So let me start with Slide 11. This is essentially in any given sort of calendar period along the X axis for all of the loans we have outstanding at that time. So it's not a cohort of view. It's just of all vintages that are existing in Q3 of 2022, what were the losses incurred in that period versus what had been modeled at the time of origination.

大衛，這是桑傑。當然。這些都是重要的問題。因此，讓我從幻燈片 11 開始。這基本上是在 X 軸上的任何給定日曆期間，我們當時擁有的所有未償還貸款。所以這不是一個群體的觀點。僅在 2022 年第三季度存在的所有年份中，該時期發生的損失與最初建模時的損失相比是多少。

And so that would say that of all the vintages that were still active or outstanding in that quarter, the losses were 25% higher. And obviously, a lot of those same vintages compose the populations in the prior quarters that were below or on target. So you're correct in that what is causing this in a sense is what you see on Slide 10, which is our sort of expression of the macro impact in the environment.

因此，在該季度仍然活躍或傑出的所有年份中，損失要高出 25%。顯然，許多相同的年份構成了前幾個季度低於或達標的人口。所以你是對的，從某種意義上說，造成這種情況的原因是你在幻燈片 10 上看到的，這是我們對環境中宏觀影響的一種表達。

Now the fact that the macro index is at 1.7 doesn't suggest that there -- we're 70% higher than what we had modeled. I guess the other side of the equation is where are we pricing loans. So today, we're pricing loans at a 2.0 sort of equivalent macro index. So put another way, if that macro index stays at 1.7 and we're pricing new loans at a 2.0, they should in fact over-perform. They should come under losses by -- to the tune of 17%-20%.

現在，宏觀指數為 1.7 的事實並不意味著存在——我們比我們建模的高 70%。我想等式的另一面是我們在哪里為貸款定價。因此，今天，我們將貸款定價為 2.0 等價的宏觀指數。換句話說，如果該宏觀指數保持在 1.7，而我們將新貸款定價為 2.0，那麼它們實際上應該表現出色。他們應該會遭受 17%-20% 的損失。

So because we rapidly adjust the model to recalibrate to where the sort of UMI is trending, we are sort of able to, in a sense, price these trends into the loans. So to get back to your original question, what should we expect that line to do on Slide 11 going forward? A lot of the existing loans, to the extent the economy continues to degrade, they are already priced. And so yes, as the economy degrades, then those losses will increase. But then you also got fresh production of loans being put into the population that are priced at much higher UMIs. And so the answer will be similar in the balance of those 2.

因此，因為我們快速調整模型以重新校準 UMI 的趨勢，從某種意義上說，我們能夠將這些趨勢計入貸款中。所以回到你原來的問題，我們應該期待這條線在幻燈片 11 上做什麼？許多現有的貸款，在經濟繼續惡化的情況下，它們已經定價了。所以是的，隨著經濟衰退，這些損失將會增加。但隨後你也獲得了新的貸款，這些貸款被投入到以高得多的 UMI 定價的人群中。因此，在這 2 的平衡中，答案將是相似的。

Very helpful. And then my second question relates to promotional activity in the third quarter related to gift cards. Was this new? And are you able to say how much that contributed to originations in the third quarter? And what level should we expect in the fourth quarter, if that's going to continue?

非常有幫助。然後我的第二個問題與第三季度與禮品卡相關的促銷活動有關。這是新的嗎？你能說出這對第三季度的起源有多大貢獻嗎？如果這種情況繼續下去，我們應該在第四季度預期什麼水平？

Sure. This is Sanjay again. So I guess taking a step back, I think that the way to think about that, and I sort of alluded to this in one of the prior questions is, in Q3 itself, we had sort of gone back and forth between a funding-constrained environment and a borrower-constrained environment. So in a sense sort of the availability of funding or lack thereof is competing with the loss trends in the economy and our ability to improve.

當然。這又是桑傑。所以我想退後一步，我認為思考這個問題的方式，我在之前的一個問題中提到了這一點，在第三季度本身，我們有點在資金受限之間來回走動環境和借款人約束的環境。因此，從某種意義上說，資金的可用性或缺乏與經濟中的損失趨勢和我們改善的能力相競爭。

And so there was some period of time in Q3 where we were actually borrower-constrained, and we took that decision to sort of run a marketing campaign where we provided some incentives in order to get some of the origination numbers up a little bit. I think the overall impact on the numbers is pretty de minimis. It was de minimis within that month and certainly within the entire quarter. It wasn't a very big impact, but it just arose from the fact -- go ahead.

因此，在第三季度的一段時間內，我們實際上受到借款人的限制，我們決定開展一場營銷活動，我們提供一些激勵措施，以使一些發起人數增加一點。我認為對數字的總體影響是微乎其微的。在那個月內是最低限度的，當然在整個季度內也是如此。這不是一個很大的影響，但它只是源於事實 - 繼續。

Sorry, Sanjay, I was going to say, I mean, what you're seeing there is, we are pretty constantly trying to find the lowest-cost source of borrowers. And in that case, I believe it was really incenting people that were already on our platform that essentially had no other acquisition costs associated with them. But what we're generally doing in all periods is trying to acquire users at the lowest possible cost. And gift cards to promote someone who has no other associated costs with them, meaning from digital or from direct mail or from a partnership, et cetera, can be a very good way to do that.

對不起，桑傑，我想說，我的意思是，你所看到的是，我們一直在努力尋找成本最低的借款人來源。在那種情況下，我相信這真的是在激勵那些已經在我們平台上的人，這些人基本上沒有與他們相關的其他獲取成本。但我們通常在所有時期都在做的是試圖以盡可能低的成本獲得用戶。禮品卡可以用來宣傳沒有其他相關費用的人，這意味著來自數字或直接郵件或來自合作夥伴關係等，可能是一個非常好的方式來做到這一點。

And there are no further questions at this time. Mr. Girouard, I will turn the conference back to you for any additional or closing remarks.

目前沒有進一步的問題。 Girouard 先生，我將把會議轉回給您，以獲取任何補充或結束語。

Thanks all. Thanks for listening. We definitely appreciate. This has been -- it's a challenging time, particularly for the mission that we're on and the business that we've chosen. But we are confident in it. We're committed to it and pretty -- need to make sure we make all the right decisions now, particularly in terms of credit performance and as well as being sort of fiscally responsible. But we're extremely confident that all the investments we're making today, continuing to do, are going to lead us to a much stronger position, and we'll be in a growth mode again soon enough. So thanks all for listening today.

謝謝大家。謝謝收聽。我們絕對感激。這是一個充滿挑戰的時期，特別是對於我們所承擔的使命和我們選擇的業務而言。但我們對此充滿信心。我們致力於它並且漂亮 - 需要確保我們現在做出所有正確的決定，特別是在信貸表現和財政責任方面。但我們非常有信心，我們今天所做的所有投資，繼續做，將帶領我們走向更強大的地位，我們很快就會再次進入增長模式。所以感謝大家今天的收聽。

This concludes today's call. Thank you for your participation, and you may now disconnect.

今天的電話會議到此結束。感謝您的參與，您現在可以斷開連接。