【想看中文的读者请点击这里

Two weeks into the first Wall Street earnings season that includes COVID-19 data, most big tech companies that operate a large cloud unit have reported results. Judging from the growth rate and revenue numbers alone, the cloud seems to be weathering the coronavirus storm just fine. AWS grew 33% year-over-year and hit the $10 billion USD quarterly revenue mark for the first time. Azure grew by 59%, Google Cloud (includes both GCP and GSuite) grew by 52%, and IBM’s Cloud & Data Platforms category (includes Red Hat) grew by 32%. Other major players like Alibaba and Oracle have yet to report their earnings.

But as a full-fledged industry, the cloud is relatively new, even though cloud computing as a form factor and product has been around for more than two decades now. Loudcloud (later Opsware) and Rackspace were both founded before the dot com boom busted. AWS was launched in 2006 with S3, SQS (Simple Queue Service), and EC2.

The current COVID-19-induced recession, if not depression, will be the first time the cloud computing industry is a meaningful participant of a large-scale economic downturn.

So given its current, uninterrupted growth trajectory, is the cloud recession-proof?

(NOTE: in this post, “cloud” means either infrastructure or platform-as-a-service, IaaS or PaaS. It does not include cloud-based SaaS products, where there are too many types and variations.)

Definitions and Framework

Before we dive into the analysis, let’s establish some common definitions and a framework first.

What makes a business or industry recession-proof?

Wall Street's somewhat technical financial definition is anything that has a negative beta. In plain terms, it means any security whose price movement is opposite of the stock market’s; if the market goes up, that security goes down, and vice-versa. Gold and U.S. Treasury bonds are the classic examples. Bitcoin is the more edgy, less understood example.

There’s also a more Main Street way to think about what is recession-proof. It’s anything that people must spend money on, regardless of how bad the economy is and how bad their personal financial situation is. Typically, that means things like utilities (electricity and water), car insurance, and essential consumer products like food and toothpaste.

Of course, these are all heuristics that do not reflect reality completely. One common heuristic is that alcohol consumption is recession-proof. (You’d want to get drunk when things are good and when things are bad.) However, certain academic study has shown that alcohol consumption actually decreased during the 2008 Great Recession.

So how should we evaluate the cloud’s recession-proof-ness? Investopedia’s five characteristics of a recession-resistant company seems to be a reasonable framework for this exercise, if we imagine each cloud platform to be an independent company. You are a recession-resistant company, if you have one of the following characteristics:

  • provides critical repair services
  • sells consumer essentials
  • serves customers insulated from downturns
  • provides mandated products or services
  • sells proprietary or specialized products

Let’s see if a cloud platform would have one or more of these characteristics.

Provides Critical Repair Services?

Not really. If anything, cloud platforms are more the consumer of these critical repair services. Given the increasing criticality of the cloud infrastructure to the essential functions of our society, the businesses that do serve the cloud infrastructure, e.g. server and network maintenance, physical security of data centers, etc., may be more recession proof than the cloud itself. Regardless of which cloud is more in demand, and even if the whole industry experiences a categorical downturn, the services that keep them in shape and functioning can’t stop.

Sells Consumer Essentials?

No. Certain cloud vendors may be the infrastructure that support the stores that do sell consumer essentials. Obviously, AWS itself is now the foundation to Amazon.com that is selling and delivering many of the consumer essentials. But most consumers don’t know, or care, if their essential shopping is happening in the cloud or not. And in an apocalyptic scenario where the entire Internet economy collapses, people will adap and buy essential goods from brick and mortar options.

Serves Customers Insulated from Downturns?

Probably yes. The customer base of the cloud is broad and diverse, from startups, to larger enterprises from all industries, to governments of all levels and sizes. While some of those customers will suffer during a recession, others will be more in demand for different reasons. Projecting into the future of our current recession, travel companies and airlines won’t be using as much cloud, but Zoom, Slack, and other products used for remote work will use more. Many startups and SMBs will shutter, but healthcare providers, vaccine research, and certain government services like unemployment insurance applications will need more capacity than ever before. The shift may not be completely proportional, but continuous consumption of cloud resources is the constant.

Provides Mandated Products and Services?

Not really. Using the cloud may be a good idea for many businesses, but it’s not border or custom control, financial audits of public companies, or oil pipelines inspection. This category of recession proof services are typically attached to specific laws or regulations that require they happen regularly, no matter what’s happening to the economy. Using the cloud is a business choice, by no means a requirement or mandate.

Sells Proprietary or Specialized Products

Quite possibly. This is a tricky one. What happens on a cloud platform is powered by a mixture of open source and proprietary technology. Some of the data center designs are open sourced as a part of the Open Compute Project. Most of the hardwares is acquired from vendors selling proprietary technologies. Most of the fundamental software components are open sourced at their core, but tied together with proprietary code that works differently in one cloud versus another. The core value that a cloud delivers for its customers is multi-dimensional: some technological, some operational, some financial, and each dimension plays out differently in a recession. So is the cloud, as a product, proprietary or specialized enough that its customers would maintain, if not increase, its spending on it during a downturn? Quite possibly.

Not Satisfying? I Agree

If you don’t find this analysis unsatisfying, you are not alone. I’m not either. (And if you came to this post looking for recession proof stock picks, I’m sorry to disappoint you.) There isn’t an established category of recession proof businesses that fully captures all dimensions of the cloud industry.

The cloud is part utilities, part commercial real estate, part railroad, and part something new that warrants its own framework. (I welcome your input to develop a better analytical framework together.)

There are also few data points that can help us understand and model how cloud platforms will perform during a recession. This COVID-19 recession will be the first one.

Lastly, even if we do have a framework and enough information to establish that cloud platforms are indeed recession proof, it’s hard to invest in them in a targeted way. All the big cloud vendors are parts of bigger tech conglomerates. And the financials around their cloud units are inconsistent and convoluted -- Microsoft, Google, IBM, Oracle all have their own definition and “spin” on what should be considered “cloud”, which is a point of frustration I’ve written about many times before.

Until Amazon decides to break AWS away into a separate entity, or when a cloud startup like Digital Ocean eventually go public, just investing in cloud computing platforms in the public market is difficult. The best you can do is invest in one of the big tech companies with a cloud or a data center colocation company like Equinix.

All the dissatisfactions aside, how the entire cloud industry will fare during this current crisis is worth watching and learning from, if only to prepare for the next recession. And there will be another recession.

If you like what you've read, please SUBSCRIBE to the Interconnected email list. New posts will be delivered to your inbox (twice per week). Follow and interact with me on: Twitter, LinkedIn.


Chinese Version Below

云计算是否能抵御经济衰退?

包括疫情数据在内的第一个华尔街财报季已经进行两周了。大多数运营大规模云计算业务的科技巨头都公布了财报。从增长率和收入数字来看,云似乎正在安然度过冠状疫情导致的经济风暴。AWS年同比增长33%,首次达到100亿美元的季度收入大关。Azure增长了59%,Google Cloud(包括GCP和GSuite)增长了52%,IBM的Cloud & Data Platforms类别(包括Red Hat)增长了32%。阿里巴巴和Oracle尚未公布盈利情况。

但作为一个行业,云计算相对还是个新生事物,没有那么成熟,尽管云计算作为一种形式和产品已经存在20多年了。Loudcloud(后来是Opsware)Rackspace都是在互联网泡沫破灭之前成立的。AWS是2006年正是启动的,当时只有包括S3、SQS和EC2。

当前由COVID-19引发的经济衰退,甚至是大萧条,将是云计算行业第一次“参与”一个大规模的经济衰退。

既然目前还在不断增长云行业是否能够抵御衰退?

(声明一下:本文中的“云”指的是IaaS或PaaS,而不包括基于云上的SaaS产品,因为其中种类太多。)

定义和框架

在开始分析之前,我们先建立一些通用的定义和思考框架。

什么样的企业能避免经济衰退的影响?

华尔街有个偏技术性的定义:任何有负beta的公司。简单地说就是任何价格变动与股市相反的东西;如果股市上涨,它就会下跌,反之亦然。黄金和美国国债就是典型的例子。比特币是一个更前卫、更难理解的例子。

还有一种更贴近老百姓日常生活的定义。那就是不管经济有多糟糕,个人财务状况有多糟糕,人们都必须买的东西,通常包括水电、汽车保险以及食品和像牙膏这种基本消费品。

当然,这些定义和例子都是各种“估算”,不能完全反映现实。一个常被提到的不受经济衰退的消费品就是酒。(日子好过要喝酒庆祝,日子不好过要喝酒解愁。)然而,一些学术研究发现,在2008年经济大萧条期间,在酒水上的消费其实减少了

那么,我们应该如何评估云的“抗经济衰退”性呢?如果我们把每个云平台都想象为是一家独立的公司,Investopedia 的抗衰退公司的五大特征算是个合理的框架。只要有以下特征的其中一个,你就是一家能抗经济衰退的公司:

  • 提供关键维修服务
  • 销售消费必需品
  • 为不受衰退影响的客户服务
  • 提供制定必须服务
  • 销售专有或专用产品

让我们看看一个云平台有没有这些特性。

提供关键维修服务?

Not really. 一个云更像是这些关键维修服务的消费者。鉴于云基础设施对社会基本运作的重要性在日益增加,为云提供服务的业务(如服务器和网络维护、数据中心安全保护等)可能比云本身更能抵御经济衰退。不管那个云厂商对用户更有吸引力,即使整个云行业都经历了一次彻底的衰退, 维护和保护这些云设施的服务应该不会停止。

销售消费必需品?

No. 某些云厂商可能是某些卖消费者必需品的电商的基础设施。AWS本身就是Amazon电商的基础设施,也是许多消费者的购买必需品的地方。但大多数消费者不知道,也不关心他们购物的地方是否在用云。就算是世界末日,整个互联网经济崩盘,人们也会适应环境回到实体店购买必需品。

为不受衰退影响的客户服务?

Probably yes. 云计算的客户种类多种多样,从初创企业到各行业的大型企业,到各级政府部门。虽然这些客户中的一些公司会在经济衰退期间遭受损失,但其他公司也会有更多的市场需求。如果看看目前的经济衰退,旅游公司和航空公司将不会使用那么多云计算,但Zoom、Slack和其他用于远程办公的产品将使用更多云计算。许多初创企业和中小型企业将被关闭,但医疗行业、疫苗研究和某些政府服务,如申请失业保险,将需要更多的资源。这种转变可能不是完全成比例的,但云计算提供的资源的长期使用和消耗是不变的。

提供制定必须服务?

Not really. 对许多企业来说,使用云计算可能是个好主意,但云计算不能和边境或海关监控、对上市公司的财务审计或石油管道检查的必须性相提并论。这类抗经济衰退服务通常于某条法律或法规挂钩,无论经济状况怎么样,都必须继续。用不用云是个商业选择,不是个要求或指令。

销售专有或专用产品?

Quite possibly. 这类产品分析起来有点复杂。在一个云平台上发生的事情是由开源和专有技术的混合驱动的。某些数据中心设计是开源的,是Open Compute Project的一部分。大多数硬件都是从专有技术的供应商买来的。大多数基本的软件组件的核心都是开源的,而把它们链接并运作起来的代码是专有的,之所以每个云平台的运作都有些不同。云为客户提供的核心价值是多维的:有些是技术、有些是运营、有些和财务有关。在经济衰退期间,对每个维度的影响都不同。那么那作为一种产品,云是否具有足够的专有性或专业性,足以让客户在经济低迷时期维持,甚至增加,在云上的支出?很有可能。

不满意?我同意

如果你对这篇分析的结果不满意,我同意。我也不满意。(如果你看这篇文章的目的是寻找不受经济衰退影响的股票,很抱歉我让你失望了。)目前还没有一个不受衰退影响的业务的分析框架能够完全涵盖整个云行业的所有方面。

一个云,部分像水电,部分像商业房地产,部分像铁路,部分完全新的事物,需要自己独特的分析框架。(我欢迎大家的意见,共同想个更好的分析框架。)

目前也没有足够的数据能帮助我们理解和模拟云平台在衰退期间的业务表现。这次冠状疫情导致的衰退将是第一次。

即使我们有一个靠谱的框架和足够的信息来证明云确实能抵御经济衰退,也很难有针对性地对其进行投资。所有大型云厂商都是一个更大的巨头下的一部分。围绕其云计算部门的财务也不够一致,错综复杂。微软、谷歌、IBM、Oracle对云的定义都有自己“风格”。对这一点的不满我之前多次写过

除非Amazon那天决定将AWS拆分成一个独立的实体,或者Digital Ocean这样的云平台初创企业那天上市,否则在股市上投资云计算平台是很困难的。能做的顶多就是投资一家有云计算的巨头,或者像Equinix这样的数据中心托管公司。

撇开所有的不满不谈,整个云产业在当前危机中将会表现如何,还是非常值得关注和学习。就算是为下一次衰退做准备,因为绝对还会有另一场经济衰退。

如果您喜欢所读的内容,请用email订阅加入“互联”。新文章将会直接送到您的收箱(每周两次)。请在TwitterLinkedIn上给个follow,跟我互动交流!