For those people who have never read Frank Herbert's epic science fiction saga, Dune, seen the trailer for the impending film, Dune 2, or have a working knowledge of the Bene Gesserit – these words form the start of a powerful incantation used to provide clarity and perspective when facing mortal danger.
Now fear is clearly not the most obvious way to characterize technology markets currently. Nasdaq – a US technology exchange – is up almost 40% this year. However, amid lingering cyclical fears, the sudden arrival (in the consumer space at least) of artificial intelligence (AI) adds another layer of complexity to analysis. That's notwithstanding the political, regulatory, and societal concerns around the broader impact of this new technology. Clarity – induced by an ancient incantation or otherwise – is most welcome right now.
The investor’s incantation
It's time to introduce another, far more powerful (for our purposes), incantation. Riffing this time on Gavin Baker’s popular technology blog (another Dune fan), it's this:
Earnings revisions drive performance over the short term. Duration of growth and changes in return on invested capital (ROIC) drive relative performance over longer time frames.
How does tech shape up (Part 1)?
To address the first part of our Baker-inspired mantra., the reasons behind short-term performance: We see that technology-sector fundamentals have been very strong so far this year. Some 75% of technology, media and telecom (TMT) companies have beaten their sales and earnings forecasts. We entered 2023 with concerns for almost the entire sector, but what we've seen has been resilient performance from software to digital advertising, to internet and payments.
One company has become the poster child for the sector's earnings bonanza. Nvidia, a firm with a quasi-monopoly on the microchips essential to training AI models, delivered a historic quarter. Analysts' forecasts weren't just wrong, they weren't even in the right ZIP code. As a result, Nvidia's share price has seen a meteoric rise this year.
Nvidia sits at the heart of this year's AI boom. Companies, from consumer internet to manufacturing firms, are trying to get their hands on the kit needed to accelerate their AI strategies. This is happening fast and there simply isn't enough to go around.
Consumer adoption of AI technology has been stunning. In 2006, it took Twitter two years to get to one million users. It took Instagram 2.5 months to achieve one million users in 2010. However, ChatGPT took only five days to reach this milestone. The question now is whether businesses, in addition to consumers, will adopt this technology as quickly.
Black and white?
There has also been plenty of concerns around an AI-driven technology bubble and investors have scrambled to group tech companies into winners and losers. Financial markets love binary dynamics as they simplify something that's likely to be far more complex and nuanced. AI is no different. Before we decide whether this is a bubble or otherwise, we should return to the fundamentals and repeat again: duration of growth and changes in ROIC drive relative performance over longer time frames.
How does tech shape up (Part 2)?
This second part of our self-labelled investors incantation will be the key determinant of longer-term equity market returns. Areas of semiconductors (not just Nvidia) have been extremely strong this year despite weakness in key end markets. ROICs in the semiconductor sector have been rising over the last decade driven by two broad forces:
- Industry consolidation
AI is more relevant to the second factor. Over time, semiconductor-market growth has diversified away from highly-concentrated sources – personal computers and then mobiles – to a more diverse range of chip users, including automobiles, data centers, and industrials. If AI is likely to be accretive to demand patterns across a number of areas – and data so far would suggest that it will be – then it will represent another very material driver of returns.
Anatomy of a bubble
According to James Montier, an expert in behavioral finance, the anatomy of a bubble has three key indicators:1
- An asset class or technology that’s new and poorly understood (think cryptocurrencies)
- Something that’s viewed as the next best thing
- A period in which equity returns are driven more by speculative behavior than changes in earnings prospects
AI is certainly a new asset class and, at the moment at least, there seems to be a general consensus that it might be that next big thing. Significantly, however, share price multiples haven't universally expanded beyond earnings estimates. For example, Nvidia's earnings estimates are up more than 75% this year. Meanwhile, some of the biggest technology stocks are generating sales growth at some five times that of the rest of the market and margins that are twice as high.
The technologies critical to AI are also far more mature than for corresponding technologies at the center of previous bubbles. These technologies provide the foundation that will allow some companies to scale up and become profitable much quicker, so that ROICs will likely be higher for these firms versus previous disruptors. It’s also likely that, in some parts of the market, the winners will keep on winning. Clayton Christensen, the guru of disruptive innovation, grouped technology change into sustaining innovation (incumbents become stronger) and disruptive innovation (where incumbents become weaker). This AI moment likely represents a sustaining innovation given its heavy reliance on data and advanced processing, which puts the advantage in the hands of those companies with the breadth, engineering capacity, and capital to design the silicon wafers and harness the data required to execute at scale. That’s perhaps also what separates this from an iPhone moment, or rather, a disruptive force which made it very challenging for incumbents, for example Nokia, to respond despite its 53% market share in mobile handsets at the time. Whisper it, but perhaps this possible bubble may be more fundamental in nature than speculative.
Ultimately, any investment strategy that relies simply on buying and holding the largest companies is likely to underperform over time. Some of those companies will, of course, thrive. But they will be few and far between. Investors will be successful if they can access and invest in those rare companies that can harness this technology to drive scale and returns. There will likely be more Nvidia moments, as the pace of technological change accelerates and, with growth an increasingly scarce phenomenon, there will to be a premium for those firms able to capitalize on this technology shift.
For investors, both incantations may yet prove to be invaluable.
1Montier, James (2007). Behavioural Investing: A Practitioner’s Guide To Applying Behavioural Finance. West Sussex: John Wiley & Sons, Ltd.