Agora (ticker: $API) is a public company I’ve been following for a long time. Back in January, I wrote “What Does Agora Do?” to explain its core technology, founding story, and API-based business model. The post went minorly viral, because Agora powers Clubhouse (officially unconfirmed, but everyone who cares to know knows), and Clubhouse’s rise in popularity and valuation dominated the tech news cycle back then.

A lot has changed since for Agora, especially in the last two weeks, when additional regulatory details from both China and the US threw its stock price into a rollercoaster ride.

Agora's stock chart from July 19th to 30th

All public companies with a Chinese root, in fact, went on a similar rollercoaster ride, which once again proves in the unfortunate reality that when something happens in China, the reaction is all knee-jerking and no nuance.

Didi != Agora != Tencent != New Oriental.

So for Agora, what has changed and what remains the same? How is its long term prospect affected, if at all, while short term volatility continues? Let’s revisit this company from these two new regulatory dimensions:

For-Profit Tutoring May Go, Online Education Will Stay

Policy changes to China’s education industry is the most obvious factor triggering Agora’s stock volatility. They also have the least long term effect on Agora's business, in my opinion.

Agora draws a solid chunk of revenue from China’s EdTech -- about 35% based on its most recent Q1 2021 earnings call. Thus, when the new policies rendered every AST company into a non-profit overnight, Agora took a hit too. That hit only makes sense in the long term if you believe that not only will tutoring disappear, but so will online education.

Online education will stay. It’s not going anywhere. In fact, there may even be more of it in China, because COVID has transformed online education from a novelty to a necessity. This likely future bodes well for Agora, because all it does is sell the tools to build and enable online learning. More generally, all Agora does is enable human interaction in a digital setting.

If for-profit AST was a “gold rush”, then Agora provided the digital “picks and shovels”. When for-profit AST becomes a national non-profit, the online courses and virtual classrooms still need the same set of “picks and shovels” to operate. So while New Oriental and TAL Education’s stocks may go to zero, Agora’s won’t.

The one clause in the new policies that received the most coverage from Western English media is Clause 13, which transforms all AST companies into non-profits and forbids foreign investors from investing in them. That hurt the pocketbooks of Western investors, so that’s what gets the most attention. However, if you dig into the other sections of this 30-clause policy document, you will see that most of them address the more systemic set of problems in China’s education industry that is saddled with troubling, perverted incentives. AST is just one of the many symptoms.

China treats “education” as a strategic national resource. In this regard, it could not be more different from America, where the national education policies give wide latitude to state governments, private institutions, religious organizations, and individual families. But one challenge that is shared between these two rivaling superpowers is that both are having a hard time delivering more equal access to quality education for more citizens.

More online education is a crucial part of delivering education equality in China. This is specifically enumerated in Clause 12 and 16, which talk about adding more online educational materials and narrowing the urban-rural education gap, respectively. Thus, the tools that enable more online and virtual learning will be key as well.

Agora’s revenue from the education sector will no doubt take a short term dip. Analysts whose funds have a short term horizon should prudently adjust their models and projections. But as long as online learning is here to stay (and even grow) in the long run, the only thing that changes is a different set of organizations paying the bill.

APIs = “Easy” Accounting

Switching gears to US regulations, in particular, the SEC’s new guidance. To better protect American investors, the guidance states a stronger disclosure requirement and expectation of Chinese companies in three risk areas:

  • Variable Interest Entities (VIEs) ownership structure
  • Government approval or denial to go public abroad
  • Inspection by the Public Company Accounting Oversight Board (PCAOB)

Of these three risks, the “government approval or denial to go public abroad” part is very much a response to Didi’s IPO fiasco. It does not apply to Agora. Like I said, Didi != Agora.

As for the financial auditing by the PCAOB, this risk actually exposes a non-obvious advantage of Agora’s API-based technology and business model. Let me explain why.

This requirement, enshrined by the Holding Foreign Companies Accountable Act (HFCAA), forces Chinese companies to meet the auditing standards of the PCAOB or risk being delisted. The PCAOB's auditing requirement of foreign companies has already existed for two decades, ever since the board was founded in the aftermath of Enron and Worldcom’s accounting scandals. But the rules have been poorly enforced when it comes to China, because Chinese companies’ IPOs have been so lucrative for Wall Street, until the Luckin Coffee fraud made non-enforcement untenable.

What makes this audit stringent is that the PCAOB needs to see all the paper trails, effectively receipts, of a foreign company’s historical financial transactions dating back three years. It won’t accept the manicured, sanitized reports produced by your Ernest & Young’s or other Big Four’s of the world for their clients; its mission is, in fact, to be the auditor of auditors.

Complying with PCAOB’s inspection will no doubt incur high compliance cost and a lot of boring accounting gruntwork. But for Agora, compliance could be much easier and cheaper. One of the hallmarks of the API-based business model is “pay as you use” -- a topic I explored in-depth in Part II of my “Global by Nature” series on APIs. For a technology platform to enable this “pay as you use” experience, it needs to keep granular records of every compute resource consumed, every API call made, and every interaction with the user, in order to bill customers correctly. The industry term for this is “metering”.

Building a rock-solid metering system is hard engineering work, but also table stakes for any successful API product. Agora has built it and now has an advantage in leveraging it to comply with PCAOB inspections more easily than other types of businesses. Tracking the different variables for an API product (e.g. how many times a user called a certain API, how many minutes an account used a video streaming service, how much storage a developer used) are pretty straightforward, which makes pricing an API product straightforward, which is one of the model’s core selling points to developers -- price transparency.

Unlike selling coffee, hailing cabs, delivering takeouts, or tutoring students after a full day of work as a poorly-paid public school teacher, the financial history of an API product has less room for human error and even less room for human corruption.

So Agora has some built-in technical advantage to become one of the first China-born companies to pass the PCAOB gauntlet. The question is: does it want to be?

Becoming a Global API Company

When I discussed the PCAOB in a previous post “Wall Street Blunts the ‘Delisting’ Hype”, I wrote the following:

Rules are rules, and they should be applied evenly across the board. I also don’t think such a change would force all Chinese companies to delist. There are ethical, ambitious entrepreneurs from China (and everywhere) who would welcome that scrutiny, because their mission is to build a global, world-changing business and would crave the trustworthiness that comes with a higher regulatory standard. And the ones who do have problematic financial practices and shady connections will delist. It’s a self-selecting process.”

Thus far, Agora appears to be one of the “ethical, ambitious” ones. Its earnings presentations have painstakingly noted revenue scenarios, with and without COVID-induced growth in its China market, to paint an accurate picture. Despite its international expansion effort being quite new, Agora’s CFO revealed that the company’s “revenue contribution from U.S. and other international markets was about 27% to 28% of total revenues” -- a fairly significant proportion already. Developers are everywhere, but technology spending still happens mostly in the US and Western Europe for the time being. Agora appears to be putting in the work to make inroads in these markets, and keeping itself publicly-listed in the US is important to growing its global brand. That’s why I feel comfortable holding onto my Agora shares.

Of course, I would dump them in a heartbeat if there is indication that the company would not follow SEC guidance to further disclose its VIE ownership structure, would not comply with PCAOB inspections, and stop investing in growing a global developer community.

Only time will tell how the Agora story will unfold, but how long can we wait?

According to the new SEC statement, “Holding Foreign Companies Accountable Act...requires that the Public Company Accounting Oversight Board (PCAOB) be permitted to inspect the issuer's public accounting firm within three years...” or be delisted. The HFCAA was signed into law in December 2020. Assuming this law is properly enforced (a big “if” given the PCAOB’s track record), then we will all know some time in the next 2.5 years whether Agora will self-select to become an ethical technology company with global potential or get delisted.

Luckily, I don’t run a hedge fund or have LPs to answer to. I can wait -- 2.5 years is not that long.

If you like what you've read, please SUBSCRIBE to the Interconnected email list. To read all previous posts, please check out the Archive section. New content will be delivered to your inbox once a week. Follow and interact with me on: Twitter, LinkedIn, Clubhouse (@kevinsxu).


声网(美股代码:$API)是一家我关注已久的上市公司。早在一月份,我写了《声网到底是做啥的?》 这篇文章,解析它的核心技术、创始故事和基于API的商业模式。文章当时有小小“走红”,因为Clubhouse用声网(官方从未证实,但每个想知道的人都知道),而Clubhouse人气和估值的飙升当时一直上各大科技媒体的头条。




滴滴 != 声网 != 腾讯 != 新东方。






如果说私营AST是一场 "淘金热",那声网则提供了数字化版的 "镐头和铁锹"。当这些AST变成国有化的非营利组织时,在线课程和虚拟教室仍然需要同样的 "镐头和铁锹" 来继续运作。因此,虽然新东方和好未来的股票可能会归零,但声网的不会。


中国其实在把 "教育" 当作一种国家战略资源。在这方面,它与美国截然不同。美国国家教育政策给予各州政府、私营机构、宗教组织和个人家庭广泛的自由。但这两大国却都面临着一个共同的挑战:两者都很难为更多自己的公民提供更平等的优质教育机会。



APIs = "简化" 算账


  • 披露可变利益实体(VIEs)内的拥有权结构
  • 中国政府批准或拒绝公司在国外上市
  • 上市公司会计监督委员会(PCAOB)的检查

在这三种风险中,"中国政府批准或拒绝公司在国外上市" 这一部分在很大程度上是针对滴滴的,而并不影响声网。就像我提到的,滴滴 != 声网。




遵守PCAOB的检查无疑会产生高额的合规成本和大量无聊的会计工作。但对于声网来说,合规可能会更容易和便宜些。基于API商业模式的标杆之一就是 "随用随付"——我在《生来全球化》第二篇关于API的文章中深入解释了这一点。对于一个技术平台来说,要实现这种 "随用随付" 的客户体验,它需要对每一个消耗的计算资源、每一次API的使用以及与用户的每一次“交换”(transaction)都有极为细致的记录,才能准确地向客户收费。行业术语是 "计量"(metering)。

搭建一个扎实的计量系统工程工作量很大,很辛苦,但也是任何成功的API产品必备的。声网已经搭建了这个系统,所以比其他类型的企业更容易使用自己的系统来合规PCAOB的标准。计量一套API产品内的不同变量(例如,一个用户调用某个API多少次,一个账户使用一个视频流服务多少分钟,一位开发者使用了多少存储空间)其实是相对简单明了的,这使得API产品的定价很清楚,也是该模型面向开发者的核心卖点之一 —— 价格透明。




我在《华尔街钝化 "退市 "》这篇文章中在讨论PCAOB时,曾表达了以下观点:


目前来看,声网是属于 "有道德、有抱负 "这一类公司的。它在每季度财报中都有刻意把因为疫情而导致的在中国市场的增长单独解释,给投资人提供一个更准确财务收入状况。尽管其对国际扩张的投入还很早期,但据声网的首席财务官透露,公司“来自美国和其他国际市场的收入贡献约占总收入的27%至28%” —— 这比例已经不小了。开发者无处不在,但科技产品购买力目前仍主要在美国和西欧市场。声网正在努力向这些市场进军,而保持其在美国上市的状态对于发展塑造一个全球品牌是极为重要的。这就是为什么我目前继续持有声网股票,并不担忧。



根据美国证券交易委员会的新指南,"《外国公司责任追究法》......要求 PCAOB 在三年内检查公司的公共会计师事务所提供的财务报告...... "(非官方翻译,仅供参考)或被要求从美国退市。《外国公司责任追究法》于2020年12月正式变成法律。假设美国政府好好履行及执行这项法律(鉴于PCAOB的历史,这是个很大的 "假设"),那么声网故事的“大结局”将在未来两年半内“播出”。它会是一家有道德,有全球野心和潜力的纯科技公司,还是会被逼退市?