Nvidia Corporation (NASDAQ:NVDA) is one of the best performers in the S&P 500; NVDA stock is up 42% year-to-date and 224% over the past year.
Under CEO Jensen Huang, who has been at the helm since founding the company in 1993, Nvidia has become the leader in graphics chips. Now the firm is in the race to develop chips specialized for artificial intelligence (AI), which will be needed in next-generation technologies such as self-driving cars.
At Nvidia’s developer conference in May, Mr. Huang spoke at length about machine learning and deep learning, two fast-growing branches of AI that deal with finding patterns in data.
As Google Trends shows, interest in these fields has soared in the past few years. Businesses are excited about these technologies, which can help detect fraudulent transactions, predict the movement of stock prices, and speed up drug discovery.
As adoption of “smart” devices takes off, so will demand for special AI chips, which NVIDIA produces. This makes the stock a bet on the future of AI, a “picks-and-shovels play” as one observer described it.
Yet some think the run-up in NVDA stock is a case of irrational exuberance.
Nvidia is not a cheap stock, trading at 51 times earnings, 43 times forward earnings, 2 times sales and 14 times book value. One Citi analyst wrote that the run-up in price estimates for NVDA and Tesla Inc (NASDAQ:TSLA) reminds him of the tech bubble.
But even the tech bubble didn’t kill tech, just some tech stocks. Artificial intelligence is here to stay. But are there some cheaper ways to bet on this new technology?
Bearish on NVDA Stock
Nvidia’s soaring valuation makes it difficult for me to view the stock positively. Cheap stocks tend to beat expensive stocks in the long-run, and I’m not convinced NVDA is an exception.
NVDA stock is up a whopping 660% since August 2015, and it is one of only five stocks in the S&P 500 with price-to-sales and price-to-book multiples both greater than 10.
When a stock trades at a high valuation, expected returns in the years that follow tend to be low. This happens to the best of companies; they may perform well but can’t help falling off the expectations treadmill.
And as McKinsey notes, Home Depot Inc (NYSE:HD) grew revenues from 1999 to 2007 by 11% a year, yet underperformed the S&P 500 by 7% a year. Why? Investors got too excited, pushing Home Depot’s P/E to 47 in 1999. They expected too much, and it took Home Depot nearly 13 years to recover the price it traded at in early 1999.
Maybe Nvidia will prove the skeptics wrong, and maybe the current bull market hasn’t peaked yet, but I wouldn’t bet on it.
Analysts sometimes create a feedback loop through their herd behavior. When one analyst raises the price target on Nvidia, others follow, and this pushes the stock up.
Fund managers aren’t immune to the herd instinct either, and may be prone to following their peers or trying to copy the moves of top managers.
The momentum can continue for a while, and some traders know how to play this. However, this is a poor time for long-term investors to be adding to their positions in NVDA stock. The odds suggest low expected returns in the years that follow.
An Alternative to Nvidia: IBM?
Nvidia may be a great company, but it’s an expensive stock.
Nvidia is loved by analysts and fund managers today, but might stumble in the years that follow. Other AI-related stocks, such as International Business Machines Corp. (NYSE:IBM), find themselves spurned by this crowd.
After all, they need to turn out high quarterly returns or their investors will leave. This creates an opening for investors with a longer-term focus who are willing to buy and hold a stock for a few years.
I recommend a competitor trading at lower multiples whose stock price isn’t shooting up.
IBM looks interesting. IBM’s Watson supercomputer beat the top human contestants on Jeopardy! in 2011 and is being used in cancer research. IBM is also making neurosynaptic chips inspired by the human brain.
IBM stock trades at a lower enterprise value to free cash flow multiple than Nvidia — a multiple not that far from the multi-year median. According to GuruFocus, IBM also trades at a Shiller P/E of 12.14, much lower than the average of 30 for the S&P 500. The Shiller P/E attempts to adjust for cyclicality by dividing price by the inflation-adjusted average of the last 10 years of earnings.
IBM stock looks like a bargain compared to the broad market, which appears expensive pretty much any way you slice it.
As of writing, Lucas Hahn did not hold a position in any of the aforementioned securities.