Tesla Inc. (TSLA): Tesla begins robotaxi operations - initial takeaways 特斯拉公司 (TSLA)：特斯拉开始自动驾驶出租车运营——初步结论
Tesla began robotaxi operations on 6/22 in Austin, Texas for a small group of early access individuals (generally those that report regularly on Tesla on X with podcasts and/or as FSD Beta testers). Recall Tesla had said the initial fleet would likely be 10-20 Model Y vehicles. 特斯拉于 6 月 22 日在德克萨斯州奥斯汀启动了 Robotaxi 的运营，面向一小部分早期用户（通常是那些定期通过播客报道 Tesla on X 动态和/或 FSD Beta 测试的用户）。记得特斯拉曾表示，最初的车队规模可能是 10 到 20 辆 Model Y。
Based on available data on Tesla’s robotaxi launch from news reports and the social media platform XX, our initial view is that the launch is both a sign of the progress that Tesla has made with its FSD technology but also that scaling in the near term will be slow. We discuss our views on three key items from the launch, and valuation/stock implications in this note. Specifically: 1) Our initial thoughts on performance of the early rides, which generally show a good degree of drive smoothness in our opinion but with one user posting a navigation issue with improper use of a left turn lane and the vehicle driving temporarily in the turn lane for traffic going the other direction; 2) How Tesla’s Austin service currently compares to Waymo, including the operating area, 3) Our views on how fast Tesla will scale (we continue to expect it to be measured in the near-term). 根据新闻报道和社交媒体平台 XX 上有关特斯拉 Robotaxi 发布的现有数据，我们初步认为，此次发布既标志着特斯拉在其 FSD 技术方面取得的进步，也表明短期内其扩展规模将会比较缓慢。我们将在本报告中讨论对此次发布中三个关键事项的看法以及估值/股票影响。具体而言：1) 我们对早期试驾表现的初步看法，我们认为早期试驾总体上表现出良好的驾驶平稳性，但有一位用户报告了导航问题，左转车道使用不当，车辆暂时在对向行驶的车道上行驶；2) 特斯拉在奥斯汀的服务目前与 Waymo 相比如何，包括运营区域；3) 我们对特斯拉扩展速度的看法（我们仍然预计将在短期内衡量）。
We will monitor for additional disclosures from Tesla going forward to better assess the technical progress and financial impact. 我们将密切关注特斯拉未来的更多披露，以更好地评估技术进步和财务影响。
Separately, please see our recent AV deep dive report for more on our industry expectations and AV market forecast. 另外，请参阅我们最近的 AV 深度报告，了解有关我们的行业预期和 AV 市场预测的更多信息。
Initial takeaways on the launch 发布会的初步收获
AV Ride Performance - Initial user reactions to robotaxi rides they experienced have largely been positive, with riders generally noting a smooth drive. However, one user showed a video with an incorrect use of a left turn only lane - Reviews from users on X generally noted that the vehicles were able to navigate very smoothly including handling turns and speed bumps, and this was evident based on videos posted on XX. Users stated that the vehicle also had some robotaxi specific features on the rear screen, such as a customer support button, and the robotaxi remembered user preferences from their Tesla account like their favorite music. Rides were handled via a robotaxi specific app from Tesla, and there was a Tesla employee in the front passenger side seat (we assume to help with some aspects like potentially aiding the vehicle in difficult situations that are more 自动驾驶出租车驾驶性能 - 用户最初体验的自动驾驶出租车体验普遍好评，乘客普遍表示驾驶过程平稳。然而，一位用户展示了一段视频，视频中左转专用车道的使用存在错误。X 用户的评论普遍认为，这些车辆能够非常平稳地行驶，包括处理转弯和减速带，这一点从 XX 上发布的视频中可以看出。用户表示，车辆的后屏幕上还配备了一些自动驾驶出租车特有的功能，例如客户支持按钮，而且自动驾驶出租车还能记住用户 Tesla 账户中的偏好设置，例如他们喜欢的音乐。行程是通过 Tesla 的一款自动驾驶出租车专用应用程序进行的，副驾驶座位上还有一位 Tesla 员工（我们推测是为了帮助解决一些问题，例如在车辆更难以应对的困难情况下提供潜在帮助）。
Goldman Sachs does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. For Reg AC certification and other important disclosures, see the Disclosure Appendix, or go to www.gs.com/research/hedge.html. Analysts employed by non-US affiliates are not registered/qualified as research analysts with FINRA in the U.S. 高盛目前并寻求与其研究报告中涵盖的公司开展业务。因此，投资者应注意，公司可能存在利益冲突，这可能会影响本报告的客观性。投资者应将本报告视为其投资决策的唯一因素。有关 Reg AC 认证和其他重要披露，请参阅披露附录，或访问 www.gs.com/research/hedge.html 。受雇于非美国关联公司的分析师未在美国金融业监管局 (FINRA) 注册/获得研究分析师资格。
likely to be handled fully with remote assistance in the long-term). However, one user posted a video where the Tesla robotaxi went straight from a left hand turn only lane and then drove briefly in the turn lane for vehicles going the opposite direction las reported on by Bloomberg). We consider this navigation error as somewhat concerning about the overall level of tech readiness, especially in light of the limited number of vehicles on the road, but also we appreciate that this behavior is not unprecedented to see from human drivers either. 从长远来看，远程协助很可能能够完全解决这一问题。然而，一位用户发布了一段视频，其中特斯拉自动驾驶出租车从左转专用车道直行，然后在对面车辆的转弯车道上短暂行驶（正如彭博社报道的那样）。我们认为这种导航错误在一定程度上令人担忧整体技术准备水平，尤其是在道路上车辆数量有限的情况下，但我们也意识到，这种行为在人类驾驶员身上并非史无前例。
2) Waymo Comparison - The areas and conditions in which the Tesla Robotaxi can operate are currently more limited relative to Waymo - Waymo currently operates 24//724 / 7 in ∼37\sim 37 sq miles of Austin, whereas Tesla is limited to 6 am to midnight and in a smaller area of Austin (Exhibit 1). Additionally, the Tesla Robotaxi service is only available to a select group of early access users, while Waymo is available to all users on the Uber app. On pricing, Tesla is currently pricing its rides at a flat fee of $4.20\$ 4.20, while Waymo (operating through the Uber app) is charged like a traditional Uber ride (with pricing shown at time of booking). 2) Waymo 对比 - 特斯拉 Robotaxi 目前可运营的区域和条件比 Waymo 更受限制 - Waymo 目前在奥斯汀 ∼37\sim 37 平方英里的范围内运营 24//724 / 7 ，而特斯拉的运营范围仅限于早上 6 点至午夜，且奥斯汀的区域较小（图表 1）。此外，特斯拉 Robotaxi 服务仅面向部分早期用户，而 Waymo 则面向所有 Uber 应用用户。在定价方面，特斯拉目前的乘车费用为 $4.20\$ 4.20 ，而 Waymo（通过 Uber 应用运营）的收费标准与传统 Uber 乘车费用相同（预订时会显示价格）。
3) Scaling - The use of an Austin-specific tech stack, a Tesla employee being present in the vehicle (albeit on the passenger side), and the navigation/lane issue reported in the first day of use suggests scaling will be slow in the near-term in our view - As we discussed in our note from 6/20, we hadn’t initially expected a Tesla employee to be in vehicle. Given this news, the navigation related error from one early consumer ride, and other factors like geofencing and a version of the software that at least for now is specific to Austin, suggests in our view that scaling in the near-term will be slow. We also believe this suggests that it will be some time before consumers can use FSD on their personal vehicles in a wide operating area. 3) 扩展——奥斯汀专属技术栈的使用、车内有特斯拉员工（尽管是在乘客侧）以及使用第一天报告的导航/车道问题，这些因素表明我们认为短期内扩展速度会比较缓慢。正如我们在 6 月 20 日的报告中所述，我们最初并没有预料到会有特斯拉员工在车内。考虑到这一消息、早期消费者试驾过程中出现的导航相关错误，以及地理围栏和至少目前仅针对奥斯汀的软件版本等其他因素，我们认为短期内扩展速度会比较缓慢。我们还认为，消费者在个人车辆上广泛使用 FSD（全自动驾驶）仍需一段时间。
Separately, the regulatory environment will be key to monitor and could determine the speed of the ramp as well. California currently requires permits for robotaxi deployments, and on 6/20 the Texas governor signed a bill that requires robotaxi/AV operator permits starting in September. However, new federal rules could help speed AV adoption. 此外，监管环境将是需要监控的关键，也可能决定自动驾驶汽车普及的速度。加州目前要求自动驾驶出租车部署必须获得许可，而德克萨斯州州长于6月20日签署了一项法案，要求从9月开始向自动驾驶出租车/自动驾驶汽车运营商颁发许可证。然而，新的联邦法规可能有助于加速自动驾驶汽车的普及。
Implications 启示
We see the commercial launch as a step toward Tesla addressing the AV rideshare market (we estimate the market for AV rideshare in 2030 in the US will be about $7\$ 7 bn). However, we were surprised by the magnitude of the move higher in shares on 6/23 (TSLA stock was up ~8% or $26.52 per share compared to the S&P 500 up 1%), as we believe from our investor discussions that many investors had expected Tesla to meet its June target (and some degree of AV related profit was already in the stock in our opinion). With the move higher on 6//23//256 / 23 / 25, Tesla added > $90>\$ 90 bn to its market cap or roughly double Waymo’s reported valuation of >$45 bn as of October 2024 per Bloomberg. 我们认为，此次商业发布是特斯拉进军自动驾驶汽车共享市场的一步（我们估计，到 2030 年，美国自动驾驶汽车共享市场的规模将约为 $7\$ 7 亿美元）。然而，6 月 23 日股价的大幅上涨令我们感到意外（TSLA 股价上涨约 8%，即每股 26.52 美元，而标准普尔 500 指数上涨 1%），因为我们从与投资者的讨论中得知，许多投资者原本预计特斯拉将实现其 6 月份的目标（而且我们认为，该股已经包含了一定程度的自动驾驶汽车相关利润）。随着 6//23//256 / 23 / 25 的上涨，特斯拉的市值增加了 > $90>\$ 90 亿美元，大约是 Waymo 截至 2024 年 10 月（彭博社报道）> 450 亿美元估值的两倍。
As we detailed in our AV deep dive, to help frame the potential present value for Tesla’s robotaxi business, we constructed an illustrative DCF of the incremental value assuming a fleet of 100K-1mn100 \mathrm{~K}-1 \mathrm{mn} AVs on the road with an EBIT margin of 10-40%10-40 \% in 2040, which implies a wide range of outcomes in value per share (Exhibit 2). We believe this analysis suggests that the ultimate value of its rideshare business will depend on the degree to which it can scale and earn a differentiated margin. One downside risk in this dimension is what has happened with the ADAS market in China, with many local OEMs now including hands-free technology as a standard feature or at low cost even for mainstream vehicles (suggesting that if AI technology allows for many entrants in AVs, profits will be diminished). An accident or other impediment to scaling is also a downside risk. However, if Tesla is able to benefit from scale/early-mover benefits/low hardware costs, there are scenarios where it could have a very attractive margin, in our view. 正如我们在自动驾驶汽车深度分析中详述的那样，为了帮助构建特斯拉自动驾驶出租车业务的潜在现值，我们构建了一个增量价值的说明性直流现金流折现法 (DCF)，假设 2040 年道路上行驶的自动驾驶汽车数量为 100K-1mn100 \mathrm{~K}-1 \mathrm{mn} ，息税前利润率为 10-40%10-40 \% ，这意味着每股价值的结果范围很广（图表 2）。我们认为，这项分析表明，其共享出行业务的最终价值将取决于其规模化程度和获得差异化利润的能力。这方面的一个下行风险是中国 ADAS 市场的情况，许多本土原始设备制造商现在将免提技术作为标准配置，甚至以低成本将其应用于主流车型（这表明，如果人工智能技术允许众多厂商进入自动驾驶汽车领域，利润将会下降）。事故或其他规模化障碍也是一个下行风险。然而，我们认为，如果特斯拉能够受益于规模化/先发优势/低硬件成本，在某些情况下，它的利润率可能会非常可观。
Exhibit 2: Tesla illustrative robotaxi DCF per share of incremental value 图表 2：特斯拉自动驾驶出租车 DCF 每股增量价值说明
We assume a WACC of 12% in the near-term decreasing to 8.7% in 2040 and a terminal growth rate of 3% 我们假设近期的加权平均资本成本为 12%，到 2040 年下降至 8.7%，终端增长率为 3%
Source: Goldman Sachs Global Investment Research 资料来源：高盛全球投资研究
We are Neutral rated on the stock. Our 12-month price target is $285\$ 285, which is based on 120X (implying about 100X non-GAAP EPS, which is near the higher end of Tesla’s historical trading range) applied to our Q5-Q8E EPS estimate including SBC. A downside valuation scenario for the stock, assuming less volume growth and slower margin improvement, could be ∼$150\sim \$ 150 (assuming a ∼20%\sim 20 \% reduction to our 2027E EPS estimate including SBC and a multiple of ∼60X\sim 60 \mathrm{X} ). An upside valuation scenario could be ∼$400\sim \$ 400, assuming a 115X multiple applied to non-GAAP 2027E EPS. 我们对该股的评级为中性。我们的 12 个月目标价为 $285\$ 285 ，该价格基于我们对包含 SBC 的第五季度至第八季度每股收益预测应用 120 倍（意味着非 GAAP 每股收益约为 100 倍，接近特斯拉历史交易区间的上限）。假设销量增长放缓且利润率提升放缓，该股的下行估值情景可能为 ∼$150\sim \$ 150 （假设我们对包含 SBC 的 2027 年每股收益预测应用 ∼20%\sim 20 \% 下调，且乘数为 ∼60X\sim 60 \mathrm{X} ）。假设对包含 SBC 的 2027 年每股收益应用 115 倍乘数，该股的上行估值情景可能为 ∼$400\sim \$ 400 。
Key downside risks to our view relate to potentially larger vehicle price reductions than we expect, increased competition in EVs, a larger than expected tariff impact or negative effects from government policy changes more generally, slower EV demand, delays with products/capabilities such as FSD/4680, key person risk, the internal control environment, margins, and operational risks associated with Tesla’s high degree of 我们的观点面临的主要下行风险包括：汽车价格降幅可能大于我们的预期、电动汽车竞争加剧、关税影响大于预期或政府政策变化带来的负面影响、电动汽车需求放缓、FSD/4680 等产品/功能的延迟、关键人物风险、内部控制环境、利润率以及与特斯拉高度相关的运营风险。
