Technology has changed the world, but investment history is littered with the likes of Blackberry and Geocities. The balance between realism and evangelism lies in the combination of valuation and earnings. Find out whether the valuations of Facebook and Adobe are supported...
Just as it’s dangerous for anyone to argue “it’s different this time”, it’s similarly dangerous for technology sceptics to argue against even the most seemingly absurd ideas. Is it really risk-free to laugh at Elon Musk’s latest argument that Teslas get better with age as software upgrades* improve the car’s functionality?
This battle between reality and hope, with the realists often looking as silly as the evangelists, is a constant in technology and technology investing. Technology has changed the world. The value created is substantial. Yet, tech history remains littered with the likes of Blackberry, Nokia and Geocities. The latest sell-off in technology (to date, the smallest of blips) is a good time to reflect, again, on the value created by technology and where we currently stand.
Technology has supported much of the improvement in equities from 2009. The S&P 500 Information Technology index has risen 301% since March 2009, compared to 202% for the index as a whole. In the last twelve months, returns in technology of 30% were double those in the index to the end of May.
Chart 1: Technology stocks have supported improvement in equities post 2009
Weekly data, index = 100 @ 31 Mar 97, to 19 Jun 2017
The Evangelist and the Realist
The balance between realism and evangelism is the combination of valuation and earnings. Perhaps fortunately, we have a benchmark for extreme evangelisation, the Dot Com Boom of the late nineties. From a valuation perspective, today in no way matches the experience of 1999. Yes, valuations have risen recently, but for both Price to Earnings and Price to Sales, the rise is barely noticeable by comparison. For reference, Price to Earnings for the S&P 500 Information Technology peaked at 80x and the Price to Sales peaked at a little over 7x. By contrast, today, the index is at 24x earnings and 4x sales.
Chart 2: This time, valuations of tech stocks are relatively supported
Valuations to Mar 2000 peak vs. valuations to 31 May 2017
Source: Bloomberg, CFSGAM
So it’s not a bubble. Indeed, it’s possible, depending on your earnings view, to say that valuations in the expansion from 2005-07 were slightly higher (with earnings higher than in 2000) than they are today.
But not being a bubble doesn’t mean valuations are cheap. Much more important are earnings, and their trajectory. Of course, the Dot Com bubble’s narrative was “eyeballs” and the growth in “pets.com” but as with most narratives there was more than a little truth. Technology was driving substantial and real earnings growth. So rather than looking at the non-businesses^, it might be useful to look at four leaders from 2000 against two from today’s FANG.
It’s not clear, from this analysis, that the big constituents of 2000 weren’t as profitable as Facebook and Google are today. Gross margins were a little volatile around 2000 but future margins, what matters to long-term investors, have maintained high levels and greater stability. The ongoing profitability of Intel, Oracle, Microsoft and Cisco is remarkable.
The evangelism of 2000 was mis-placed yet the scepticism of 2010, 2011, when valuations were at their lowest, was similarly mis-placed. Oh, to have bought Facebook in September 2012!
Chart 3: Gross margins of the major tech companies seem sustainable
MCOI 4 vs Facebook & Google, quarterly data (%), Mar 1997 - Present
The Fundamental Evangelist: Facebook, Adobe and Nvidia / Micron
The profitability of the industry as a whole reflects two consistent themes across a variety of sub-sectors; strong demand for products and services with increasing utility and high margin business models with improving industrial economics. To illustrate this, I’ve highlighted four companies across three sectors; Facebook in the internet, Adobe for enterprise software and Nvidia/Micron in semiconductors.
As much as the internet enables cat videos, it also contributes to important productivity improvements generated from the sharing of information and, increasingly, commerce. Facebook is a dominant player in this.
“But the kids don’t like Facebook anymore”. This common view misses the essential value of Facebook to its users, shareholders and the broader economy. To go all evangelist, Facebook is the social network or, alternatively, social infrastructure. There are, arguably, more entertaining platforms; Instagram, Snapchat and Twitter, are examples. Facebook, however, creates value as users’ own personal internet bulletin board. These personalised bulletin boards provide Facebook with the ability to offer advertising that is extremely well targeted, with businesses receiving high, measurable Returns on Investment. Interestingly, many of these advertisers are new advertisers looking to advertise to a customer base that is often geographically near rather than far. Facebook, might seem to be changing the world, more importantly, it’s changing your neighbourhood.
The resulting profitability, particularly in the US, is astounding. Each user in the US attracted $67 in advertising in the 12 months to March 2017.
Chart 4: Facebook Average Revenue Per User (ARPU)
Quarterly data, start 4Q 2010, rolling 4 qtrs, USD
Source: Facebook, Bloomberg, CFSGAM
This growth underpins Facebook’s profitability. The future of growth seems dependent upon two things; finding new use cases and building a sales infrastructure.
The monetisation of Facebook Messenger is a particularly good example of new use cases. The forced adoption of Facebook Messenger has been the fastest adoption cycle in internet history. It allows Facebook to drive transactions in ways similar to WeChat in China. For instance, the e-gaming company Activision used a chat bot to engage with users via Facebook Messenger. In the first 24 hours the chat bot had 6 million interactions. Similarly, tools that improve the quality of advertising enable smaller and smaller businesses to have a professional online presence.
One of the paradoxes of information technology, a service delivered remotely, is the importance of a strong sales model. In some cases, Sophos in London and Atlassian in Australia, it has been possible to deliver sales remotely but in the main the internet and software need an on premise sales model. Facebook faces just the same problem. At the Morgan Stanley TMT conference in March, Facebook CFO highlighted the importance of building its HR function to cope with a regular doubling in employment. In the United States, these sales people will increasingly be focused on engaging small business, while in South East Asia it’s a clean slate. Despite having more users in Asia than in any other region, the absence of a sales team means its revenue contribution is tiny. Facebook’s employee growth has been 39.7% compound annual growth rate from 2012.* (Source: Facebook, Bloomberg, CFSGAM).
Enterprise software is a key productivity theme in the technology sector. Firms are continually looking for ways to improve their business processes to focus on and invest in the things that really matter; product development, product marketing and customer service. Adobe is a good example; not only does it improve efficiency in many marketing functions, it also aids the value add part of the process.
Demand for Adobe’s product comes from increasing online content. As the internet and its importance increases, the demand for professional content rises similarly. Adobe is all about delivering that content. It uses a Software as a Service business model where it charges users an annual subscription for access to its Creative Cloud. Since 2014 revenue from the Creative Cloud has risen 75% on a gross margin of greater than 65% (source: Adobe, Bloomberg)
The use of the Cloud to drive growth in the Creative segment for Adobe highlights two key points; new use cases and the explosion in data creation. For instance, Adobe now uses Machine Learning to help improve searches of its stock photo gallery which in turn requires greater data storage and computing power.
Chart 5: Adobe Digital Creative Revenues continue to grow
Rolling 4 quarters (USD m), Nov 2014 - Jun 2017
Source: Adobe, Bloomberg
Semiconductors: Nvidia and Micron
Semiconductors are the “picks and shovels” of information technology. As the uses of the internet, software and computing power expand it drives an increasing requirement for data. At the heart of data processing is the chip.
Coming into 2016, the market saw stable to lower demand for semiconductors over the coming two years. The price of DRAM and 3D Nand had collapsed and smartphone demand seemed to be tailing. Instead, growth has been improved as new use cases for semiconductors have emerged. It started with Nvidia’s new chips for gaming but has morphed into cars, data centres, machine learning, a doubling in the memory on a standard smartphone, increasing use of sensors in industrial processes and payments. It’s also emerged in the demand for crypto-currency mining.
Importantly, it’s also occurred at a time when the industry in our view, is as under-invested as it’s been in some time. With a more stable industry structure and a two-decade long legacy of over-investment, semi-conductor companies have entered a period of demand with limited plans for capacity growth. For once, the industry seems capable of benefiting from a demand upswing.
On balance, being a tech evangelist has been a formidable strategy over time. Amazon was $107 at the peak of the dot com boom in March 2000 and it hit $1,000 in June 2017. That’s 13% CAGR for 17 years. And, if you bought the bottom tick (because that’s what we all do), it’s 39%. But every so often, a little realism doesn’t hurt.