Tech stocks seem cheap on paper - based on valuations relative to the market using both P/E multiples and free cash flow yields they are the cheapest they have been in decades - but as JPMorgan's Michael Cembalest notes there are some obvious reasons for this (and some not-so-obvious). The challenge with technology companies: there are a lot of haves and have-nots; as sector-level valuations may be less useful when looking at technology companies than when looking at sectors with more homogenous revenue performance.
Via JPMorgan's Michael Cembalest,
Technology Stocks - A Sharp Divide Between The Haves And Have-Nots
Tech stocks seem cheap on paper: the first chart shows tech valuations relative to the market using both P/E multiples and free cash flow yields.
[4]
Tech stock P/Es are now roughly the same as the overall market, while their free cash flow yields are 2.5% higher. Both measures suggest that tech stocks are cheaper in relative terms than they have been in decades. This phenomenon is partly explained by investors piling into defensive stocks with higher dividends in a zero interest rate world.
[5]
It may also be partly explained by the chart above: outside of recessions, tech earnings growth is usually ~20%, and now it’s close to zero.
The challenge with technology companies: there are a lot of haves and have-nots. The chart below looks at rolling annual revenue growth since 2009. The blue lines are the fastest growers, and the red lines are the slowest. We added some bellweather names in green as well. We are obviously mixing a lot of things here: companies dependent on overall business capital spending trends; companies whose revenues depend on other companies’ product releases (game consoles, tablets, etc); companies negatively affected by the migration to cloud/mobile computing; and companies affected by the US government sequester.
But that’s partly the point: sector-level valuations may be less useful when looking at technology companies than when looking at sectors with more homogenous revenue performance. The specific products and services tech companies sell often overwhelm whatever macro/industry trends may be taking place. As a counter-example, almost all US companies targeting high-end consumers in S&P’s Luxury Index grew revenues over the same time frame.
[6]
Revenue dispersion between technology haves and have-nots since 2009 is wider than in most other sectors, with only energy being similar. Investment strategies designed to exploit these divergences are worth a look: long-short technology funds, or technology funds that have the ability to hold concentrated positions rather than having to track an index.
[7]
Given this high revenue dispersion, it may not be very informative to look at sector-level tech valuations. Instead, a closer look at individual S&P 500 tech stocks shows that their multiples may not be that mispriced, at least relative to each other. The chart below uses 9 variables related to revenues, earnings, margins, dividends, etc. to estimate current tech stock P/Es. The model does a pretty good job estimating P/Es (r^2=62.5%) given how company-specific P/E multiples are. In general, larger companies with lower gross margins and lower earnings growth are the ones with the lower P/E multiples.
[8]
We separately looked at “size”, and found that today’s large tech companies are twice as big relative to the overall economy as they were during the 1990’s. This may explain lower P/Es on megacap tech stocks: can they continue to grow at the same pace, either organically or via acquisition? Tech stocks may offer better long-term value than the defensive dividend-payers whose prices have been bid up, but are not as universally cheap as the sector-level chart suggests.