Thoughts on Artificial Intelligence and Asset-Price Bubbles

The stratospheric rise of the stock price for graphics chipset pioneer NVIDIA (NVDA) and other artificial intelligence (AI) participants invokes more than a few thoughts of asset-price bubbles. As Wharton Professor Jeremy Siegel recently noted in commentary on the subject, “momentum can carry stocks far higher than their fundamental value, and no one can predict how high they might go.” We argue that price appreciation far above a company’s fundamental value hardly warrants being labeled a “bubble.” Appreciation of 160% in less than six months, however, warrants heightened scrutiny.

We concede that several fundamental reasons exist for reputable companies to move significantly higher in compressed periods of time:

  • Takeout bids
  • Product or service innovation that opens up significant new markets
  • Removal or reduction of litigation risk
  • Removal or reduction of credit or “going concern” risk
  • Discoveries of life-altering drug treatments
  • Discoveries of substantial resource endowments like gold, oil, or other precious minerals

Unfortunately, only time will tell if this move across AI-driven stocks was a bubble or not. A potential test for those looking to deploy fresh capital to these companies (an investing subset in which Solyco Wealth is not currently participating) being driven higher by AI aspirations:

How and to what degree will the target companies’ offerings aid their customers in either driving revenue and/or cash flow higher from existing customers or in expanding the total addressable market for their products/services?

In order for this move higher for AI-levered stocks not to be a bubble – it’s added $100s of billions in cumulative market cap in a relatively compressed period of time – AI will need to create substantial incremental sales or result in vast expense reduction (or both) for customers.The replacement of existing legacy chipsets in hyperscale data centers with AI-capable chipsets alone, in our view, will not be sufficient to support this recent move higher as this action alone would solely represent a potential pull-forward of an existing upgrade that likely would have happened over the next 1-3 years anyway. Instead, these newly AI-capable data centers will need to help their customers do a whole host of new activities or complete their past activities far more efficiently. Fundamentally, without these ultimate AI customers benefiting this move higher probably represents little more than an interim phase of substantial multiple inflation for a very limited subset of Tech equities. Indicative of the potential to create value, however, NVDA’s founder noted that the company struck a deal with advertiser WPP (WPP) for AI-generated advertising. Not exactly the value creation we hope for…but maybe it’s a start.

While we at Solyco Wealth are not chasing the AI move higher, we also managed not to completely miss it as well. Several of our semiconductor and equipment (AMD, ANET), communication services (GOOGL), consumer (AMZN), and Industrials (WCC), companies obviously benefited from an NVDA-led pull-through to higher equity valuations. Largely, however, we have looked to “fade” this move higher by booking gains and/or selling covered calls in these select positions. Again, only time will tell how right/wrong this action will be, but we do know two things at this point:

  1. It’s better to have participated in the move higher than not, and
  2. We can always redeploy the cash generated from selling shares or covered calls later if the fundamental conditions, in our view, make buying attractive.