A Data-to-Everything Platform Powered by AI
“The most arrogant thing an investor can do is to stock-pick,” was the title of an article by MarketWatch which suggested that purchasing any one stock implies the price is too low and every other person on the face of the planet thinks otherwise. According to the efficient market hypothesis, all available information should already be priced into any given stock, and it’s only additional information – bad or good – that makes the stock price move up or down. Then you have Warren Buffet claiming that it’s “hard to argue that the market always prices rationally. In fact, market prices are frequently nonsensical.”
Maybe the arrogance of stock-picking is ignoring that fact that there are thousands of professionals out there whose job it is to stock pick and 80% of them get it wrong. So don’t think spending a Saturday afternoon reading a few articles about a stock, and then deciding to invest in it, will somehow put you at an advantage to the rest of the stock-pickers out there. That doesn’t seem likely. Instead of trying to stock-pick, what about buying a basket of stocks? That’s why Exchange Traded Funds (ETFs) were created. They’re a way for retail investors to get exposure to a particular theme while enjoying the benefits of diversification. The problem is, many of these ETFs provide very diluted exposure to the themes they’re trying to target. That’s because there aren’t many pure-play stocks out there for many disruptive technology themes like artificial intelligence, so the ETF creators need to get creative when selecting constituents.
Take, for example, some of the ETFs from ARK Invest, a company that builds thematic ETFs. We recently talked about how the ARK Invest 3D Printing ETF (PRNT) contains some names that would hardly be considered 3D printing pure plays. That’s not their fault though, as they’ve simply created that ETF using someone else’s index. But when we look at their “active ETFs” – this means ETFs where they’re actively stock picking in an attempt to beat the market – we notice a few common themes. Firstly, they love Tesla, and secondly, they’re overweight on a few names we haven’t written about before. One of those is Splunk.
Founded in 2003, San Francisco-based Splunk achieved profitability in 2009 before taking their company public in 2012. Since then, the $26 billion company has performed well for shareholders returning +357% since April 2012 compared to a Nasdaq return of +218% over the same time frame. A self-described “data-to-everything platform,” what Splunk does is to process something called “machine data” which is basically the data exhaust that happens as a result of running loads of enterprise apps in an organization. (For all you original web gangsters out there, this would be the equivalent of checking the web server logs for errors – and then telling the user to “refresh their browser” after your developer fixes the problem – thus allowing your customer to not only pay your bills, but also provide some free quality assurance testing for your resource-constrained development team.) Using an organization’s real-time data exhaust provides “operational intelligence,” which can be used to improve service levels, reduce operational costs, mitigate security risks, demonstrate and maintain compliance, and drive better business decisions. Most of this magic happens in their flagship product, Splunk Enterprise.
Splunk Enterprise and AI
Splunk Enterprise can collect and index petabytes of machine data daily, irrespective of format or source. The product is priced primarily on the amount of data indexed, “the maximum aggregate volume of uncompressed data indexed on a daily basis, expressed in gigabytes, terabytes or petabytes per day.” That’s a great business model to have as 5G heralds in a massive increase in the amount of data companies can expect to ingest as we move towards “The Internet of Everything” where everything has a digital twin that’s being measured in real-time.
The above chart produced by IDC predicts that the amount of data in the world – the global datasphere – is expected to triple in just five years. By 2025, 30% of all data is expected to be real-time. Making sense of all this data seems like the perfect mandate for some machine learning algorithms.
The Splunk platform uses embedded machine learning algorithms to better predict and help prevent IT, security, and IoT incidents, and to forecast key business indicators. Those capabilities seem to be in demand. As of their last 10-K filing, the company had over 17,500 customers in more than 130 countries, including 92 of the Fortune 100 companies, with no single customer accounting for more than 10% of total revenues and just over 70% of their revenues come from the United States. Regardless of where revenues are coming from, they’re growing fast.
As with many high-growth Software-as-a–Service (SaaS) companies we look at, Splunk is focused on growing their revenues as quickly as possible with profitability taking a back seat for now. At some point revenue growth will slow, pundits will say bad things about the company, and the stock price will fall temporarily as investors panic. This transition from focusing on growth to focusing on profitability will be welcomed by investors who know about the fat margins that software companies like this can eventually realize. When growth slows, retention will come into focus since higher cancellations could actually result in negative growth. That’s not a problem today though, as the company enjoys an extremely low cancellation rate of around 95%.
Acquiror or Acquiree?
There are plenty of possible suitors who might be interested in a business like Splunk’s. Said an article by Forbes around the time of Splunk’s IPO, “valuation is likely to be kept aloft by speculation that incumbent tech players like Oracle, HP, IBM and EMC are going to be shopping for cloud and big data plays to spruce up their own product lines.” That’s probably still the case today, though all the acquiring we’ve seen so far involves Splunk eating up smaller companies. One company that might be hoping Splunk keeps that trend going is Moogsoft.
This isn’t the first time we’ve read about how data exhaust from enterprise software and hardware is being used to improve operations. A few years ago, we wrote a piece titled “Moogsoft – Machine Learning Takes Over IT Operations,” which talked about “algorithmic IT operations”:
Enterprises today use anywhere from 10-25 different tools provided by vendors which help them monitor their production stack of applications, networks, and infrastructure which generate millions of events and alerts every day. We call this “operational noise” because the majority of those events and alerts are not actual issues, but need to be analyzed anyway to make sure they’re benign. What Moogsoft has built is called “algorithmic IT operations” or AIOps for short.
Splunk and Moogsoft are actually partners. That’s because Moogsoft works at a very granular level to provide context around problems, and then feeds this output into Splunk. It’s the sort of company that Splunk might look to acquire, especially considering they’ve dedicated an entire page of their website to list out companies they’ve been acquiring. Any company that gets acquired by Splunk gets to immediately benefit from their broad distribution channels. When you’re selling a product to 92% of Fortune 100 companies, that means you also have access to their decision makers. Selling additional products on top of that becomes much easier.
If you’re actively looking to get some exposure to technology stocks these days, you have some serious cojones. That’s because we’ve just entered the longest bull market in the history of mankind. All it takes is for one big disaster to strike and the market could go into a recession. It’s not if, but when. While coronavirus was made in China, and consequently isn’t expected to last very long, it just might be the tipping point. If the market goes into a recession, there will be a demand for solutions that save companies money and create efficiencies. That’s what Splunk does, and they’re also a way to play the AI and big data themes as well.
“There really are few — if any — public, pure-play artificial intelligence stocks,” says an article by Investors’ Business Daily with a click-bait title that suggests anything but. Based on the research we’ve performed over the past five years, we believe there are a fair number of stocks in computer vision and healthcare that could be considered pure-play stocks on the AI theme. You just have to know where to look for them. We’re in the process of putting together some subscriber-only reports that detail these stocks, many of which are only available on foreign stock exchanges. Stay tuned.
Pure-play disruptive tech stocks are not only hard to find, but investing in them is risky business. That's why we created “The Nanalyze Disruptive Tech Portfolio Report,” which lists 20 disruptive tech stocks we love so much we’ve invested in them ourselves. Find out which tech stocks we love, like, and avoid in this special report, now available for all Nanalyze Premium annual subscribers.