Yesterday’s trading was all about Nvidia (NVDA) and its eye-popping positive guidance about anticipated spending on artificial intelligence (AI). Today’s winner, in a similar vein, is Marvell Technology (MRVL). They reported decent earnings per share (EPS) after yesterday’s close, $0.31 on a non-GAAP basis, versus a $0.295 consensus[i], but also mentioned that they expect AI revenues to “at least double” in 2024. That’s propelled MRVL +20% higher today.

For the contrarians and natural skeptics among us (including yours truly), it is understandable to question whether AI can live up to the hype that it’s currently receiving. We won’t know the answer for years. But in the meantime, it’s irrelevant. Regardless of whether AI might prove to be earth-shattering, as long as companies are willing to spend money on AI-related technology, then the companies that supply that technology will benefit.

It is well known that the most likely way to profit from a gold rush is to supply the goods and services required by the hopeful miners. In February, on the day after NVDA’s prior earnings report, we noted:

…many investors take a “gold rush” approach. During the California and Klondike gold rushes, selling supplies to prospectors was a safer and smarter to make money than the actual prospecting. It is quite reasonable to think of NVDA as selling picks and shovels, first to the cryptocurrency gold rush, now to the AI gold rush.

Levi Strauss, founder of the eponymous company (LEVI) that still posts profits today, is the best-known example. Since then, we have seen several other gold rushes – literal and figurative – with some of the figurative ones taking place in the same California region that came to prominence in 1849. 

NVDA benefitted tremendously from the recent rush into cryptocurrencies before stumbling and once again regaining its footing (and then some!). The verdict is still out about whether blockchain and cryptocurrencies will be a life-changing innovation, but we can look back at an innovation that did indeed prove to be life-changing – the internet.

It is hard to imagine life without the internet now. One can certainly assert that it indeed live up to the late ‘90s hype, even if the enthusiasm for related stocks proved to be wildly overblown. Earlier this week we outlined some of the parallels between the internet craze and the current AI enthusiasm. We’re far from the full-on mania that existed back then, but when we see companies benefit from simply announcing that they are considering AI investments – much as they did when they announced websites in the ‘90s – there are certainly echoes of that era.

Because that era had its shares of flameouts, some of which seemed sketchy even then (Pets.com, anyone?), there were many who preferred to invest in the companies that would make up the backbone of the internet – that era’s metaphorical picks and shovels. But those too proved to be susceptible to overvaluation as well. Two of the most prominent internet infrastructure companies of that era still trade today.One is Cisco (CSCO), the other is Akamai (AKAM). Both peaked in early 2000.CSCO finally topped its peak about 20 years later, while AKAM still has not:

25+ Years, Monthly Data, AKAM (white), CSCO (blue)

(Click on image to enlarge)

Source: Bloomberg

Oh wait, I almost forgot – Enron was also a huge player in bandwidth back then. For obvious reasons, an up-to-date chart is unavailable.

It is indisputable that NVDA and MRVL, along with others, are benefiting from their role as key providers of chips that will drive AI expenditures. They have been phenomenal investments and are very hard to bet against right now. They can reasonably be considered to be continuing winners if AI remains a key investment thesis. Whether or not they are solid long-term investments remains an open question, however, NVDA and MRVL sport nothing like the premiums attributed to AKAM and CSCO at their peaks, but the latter stocks demonstrate that even during a virtual gold rush it is possible to overpay for the metaphorical picks and shovels.

[i] MRVL reported that GAAP EPS was a net loss of -$0.20 on a diluted basis, but who’s counting?

More By This Author:

It’s Nvidia’s World, We’re Just Living (And Investing) In It
Lots Of Rhyming History Out There
Would A Debt Ceiling Resolution Be A “Sell-The-News” Event?