Fra Hermes:
It’s easy to understand why investors can’t stop talking about US tech stock froth. The tech-heavy NASDAQ has logged more all-time closing highs during 2017 than any other year on record and tech sector earnings grew by 22% in the third quarter – double the rate forecast. As fears of a tech bubble mount, we explain why US tech valuations are justified.
The composition of US indices has evolved rapidly. In 1990, the technology sector accounted for just 5% of the S&P 500 index by market weight. Today, it makes up about a fifth of the S&P 500’s $24tn value. But that weighting is only moderately above the long-term average and significantly below the dotcom boom era.
Figure 1: S&P 500: Historical sector weightings
Source: Bloomberg, Raymond James as at October 2017
It’s not 1999
US indices’ increasingly punchy valuations have been driven primarily by tech stocks this year. Indeed, when tech stocks soar, comparisons to the dotcom bubble of 1999 inevitably follow. But this isn’t 1999.
The so-called FAANG stocks – Facebook, Apple, Amazon, Netflix and Google (now Alphabet) – have all made gains of more than 30% this year. The NASDAQ has rallied by almost 30% so far this year, putting it on track for its best year since 2013 – its fourth best annual gain since 1999 when it soared 86%.
However, institutional investors are underweight technology stocks, according to a quarterly report from eVestment. It found large cap growth equity funds went underweight the broader tech sector at the end of the second quarter this year1. The average weight of tech stocks was 28% compared to 31% for the Russell 1000 Growth Index.
US tech stocks are not expensive
At Hermes, we believe tech valuations are not extreme. Here’s why:
1. Adjusting for today’s credit conditions, valuations are actually well within their historic band. Data from Renaissance Macro (see Figure 2) shows that the percentile rank of technology valuation is at an average reading versus the sector’s long-term history. This method examines earnings before interest, tax, depreciation and amortisation (EBITDA) to enterprise value (EV) for each individual stock, subtracts the BBB yield, normalises for valuation over time and aggregates it to the index level – that is, the lower the yield, the more expensive it is.
Figure 2: S&P 500 information technology valuation
Source: Renaissance Macro as at October 2017
2. There’s a case for buying tech stocks as a value investor. Apple and Microsoft have been widely held as large positions in both value and growth portfolios over the last year.
Furthermore, veteran investor Julian Robertson also outlines the value argument that Apple, Facebook, Alphabet and their ilk are “priced cheaper than they would have been in the 1960s, 1970s, and 1980s”. For example, the Nifty Fifty – the historic benchmarking index that was used in the 1960s to reflect the stock index – traded on P/E ratios of 50x to 100x in the late 1960s. This also included ‘winner-takes-all’ companies – businesses that cut prices and boost investment in order to build a wide moat through scale economies – like Amazon.
3. We believe technology spending as a percentage of GDP could rise for decades to come. In the past, tech spend has tended to coincide with periods of labour shortage, as companies try to plug the gap by investing in automation. In the next 10 years, growth in the 16-64 age bracket will lag the overall population growth rate, creating a strong tailwind for the sector.
4. Ignore the price-earnings ratio – Investors should use a more suitable valuation metric. Amazon’s P/E is at elevated levels because it purposefully does not make money today, instead it recycles profits to keep growing, so its earnings are negligible. Amazon faces the largest and least penetrated addressable markets: $20tn retail (10% online penetration), $1tn cloud (10% online penetration), $1tn advertising (30% online penetration), $5tn business supplies (10% online penetration) and there may be others.
Normalising Amazon’s P/E ratio would yield a multiple of roughly 25x. However, we think EV/ gross profit (GP) is a better metric to assess its true value. EV/GP tries to show what Amazon could earn if it wanted to on a true earnings basis. Using this measure, it trades between 5x and 10x – a more reasonable valuation given its gross profits are growing at more than 40%.
Figure 3: Using a different valuation metric: EV/Gross Profit
Source: Bloomberg as at November 2017
Admittedly, some FAANG names have an attractive P/E ratio, thanks to their strong growth rates. For example, Facebook has delivered 18 consecutive quarters of ad revenue growth near 50% and it has an operating margin of 50%. Furthermore, its future growth prospects look robust due to its low market share of global advertising and its ownership of Instagram, which is at the early stage of monetisation. Nevertheless, Facebook trades on only 25x GAAP P/E.
The Hermes Approach
Bubble talk is unlikely to go away soon. But we believe the US All Cap Strategy is well-positioned in the current environment. Technology stocks and Amazon – classified as a consumer discretionary stock – make up 23% of the Hermes US All Cap Strategy. We have a small number of other holdings in Consumer Discretionary and REITs, reflecting the risk of disruptive technological change. The Hermes US All Cap Strategy has achieved a net annualised performance of 10.63% since inception and 3.03% in October2.
Figure 4: Net performance of US All Cap Strategy, since inception
Source: Hermes as at 31 October 2017. Performance shown is the Hermes US All Cap Strategy in USD, net of all costs and a 50bps