The one thing any value investor comes to appreciate over a long time of practising their faith is that value investing is an extremely uncomfortable style.
This is because the largest component of returns often occurs in fairly short bursts.
However, one of the biggest mental errors value investors make is in observing what is happening somewhere else and looking for it to blow up — just because it is expensive. Unfortunately, investors such as Netscape’s Jim Clark made that even more uncomfortable by bringing tech companies to market at the mezzanine stage, when they have virtually no cash flows or earnings, and making it fashionable.
The truth, of course, is that these scary new listings can work because of the sheer velocity of revolutionary change in the tech sector and at these companies.
This is the position they find themselves in right now, with fairly high ratings on big tech (particularly AI-themed stocks) and not a whiff of bad news about the pace of development in sight. Of course, sooner or later adversity does arrive; it just doesn’t seem very constructive to hang around wringing one’s hands and waiting for the balloon to go up when one of the most seismic technological shifts in decades is playing through.
Of all listed equity sectors, tech is now easily the largest, with about 2,500 listed companies adding up to a market cap of about $36-trillion.
The 2022/2023 rate-hiking cycle would normally have severely pricked a bubble. This is because tech stocks’ growth stock-style expected cash flow profiles mean that the biggest component of their valuation is derived from long-term outcomes. These are inevitably hardest hit by rate hikes, in the same way that long-dated bonds are harder hit than short-dated bonds by rate hikes.
Now, the carnage in the market from February 2022 to July 2023 was messy, but it was far from terminal. Clearly there was lots of juice left in many lemons, as evidenced by the recovery in share prices and earnings. Interest rate cut expectations are certainly being firmly pushed to the flat side of things, but resumption of the pain of two years ago seems remote.
Moreover, if rate hikes do re-emerge, they should do so off a much higher base and they will likely lack the leveraged venom of those first equity smacks. Because valuations are a discussion point (whether one is about “value” or “growth at a reasonable price”) I note that on a measure of enterprise value to revenue tech indices are at half the levels of the dot-com and pandemic peaks, but, say, 30% above a stable(ish) long-term average. So, high, but certainly not peak, in an environment of sustained positive news.
The firm points out that growth cycles, when they get going, can be stronger and last longer than anyone expects
Once again I found myself at Citywire’s annual conference of offshore fund managers, held at Fancourt. Specialist fund industry follower Citywire played host to 15 offshore managers meeting South African multimanagers. Sometimes these were first-time exposures to the local market, making the meetings relatively fresh.
At this event Guinness showcased its Global Innovators Fund. This company was founded in 2003 by Tim Guinness (who founded the original Guinness Flight Global Asset Management in 1987, which was sold to Investec in 1998). Both fund managers are MSc-plus physicists — something the engineer in me appreciates.
The portfolio tends to have 30 equal weighted stockholdings on entry, with a global spread quite similar to the MSCI all country world index (ACWI). This fund has seven stocks in common with the Guinness Equity Income Fund I covered last year. The fund has a beta of 1.1 times against the ACWI, which means it is more volatile and sensitive to macro factors than the index; however, the relative performance seems to be worth it.
In terms of the holdings, the average five-year sales compound annual growth rate is triple that of the ACWI; forward earnings growth and return on capital are both 2½ times higher than the ACWI; and debt equity is only a third. The practice of equal weighting tends to lead to a position-sizing risk management benefit.
The fund managers pick from a 1,000-strong universe of large-cap stocks in these defined growth areas: advanced health care; AI and big data; clean energy and sustainability; cloud computing; internet and media and entertainment; mobile technology and the internet of things; next-gen consumer (e-commerce, everything-as-a-service, healthy living); payments and fintech; and robotics and automation. They claim to have held Nvidia since inception, and indeed held six stocks that form part of the magnificent seven (Mag7).
The fund managers observed that a significant outperformance run in 2023 in the Mag7 was sparked by the release of ChatGPT, but not all was AI-related; there were different drivers of performance across the group. The common attributes of the successful six (excluding Tesla, in other words) were: improving top-line growth; high and (improving) margin profile; and compounding earnings growth. The fund managers argue that the valuations of the six appear reasonable and near-term competitive threats appear low.
As an aside, I note comments from Bridgewater that the big five tech shares (Amazon, Alphabet, Meta, Apple and Microsoft) are sitting on a cash pile of $400bn, or 1.5% of US GDP. However frothy markets may seem, these certainly aren’t the kind of gunge that blew up the internet boom in 1998 and 2000. Anyway, the firm passed some useful Mag7 observations.
Nvidia had huge growth in revenues from AI/data centres, which has higher margins than the graphics segment and so led to faster earnings growth. Demand remains strong and competition far behind, and the high earnings reduced the p:e. The growth of Amazon Web Services improved already high group margins and profitability, and e-commerce returned to profitability.
Valuation derated on strong earnings growth. Alphabet is still dominated by Google search and advertising, a very high-margin business, while Google is No 3 in cloud and growing quickly. Though well placed to benefit from AI, it recognises that AI is a potential threat to the search business. For Meta, 2023 was announced as a “year of efficiency” and focused on improving financials.
There are “compelling investment opportunities” within AI to improve core businesses in the near term (better-served advertising) and the use of generative AI in the longer term. Apple has not been rewarded as a big AI winner, but the App Store is well placed to capture whichever app “wins” in AI, without the risk of trying to pick the winner. China iPhone sales remain a key driver.
The most controversial Mag7 stock at the conference was Tesla. The Guinness team was concerned about the deterioration in several financial ratios. Other teams argued that the Tesla AI could still be the “one to rule them all”!
I recall Peter Lynch of Fidelity’s Magellan Fund fame writing in one of his books that the users of technology were ultimately bigger beneficiaries than the vendors. Perhaps that was true once; these days the tech giants seem like standalone oligarchies rather than companies.
Nevertheless, the adoption of steam, electricity and the internet was widely beneficial, and there is no reason that should not be so in the case of AI. So I think the wisdom of Lynch will eventually come to pass.
Other teams argued that the Tesla AI could still be the ‘one to rule them all’
I raised the question of these trickle-down benefits at the conference; the most satisfactory answers came from Sands Capital. Another participant nudged my memory to the effect that in previous years, Sands had banged the drum very loudly on the potential offered by Nvidia. Another spot-on comment from the firm last year was the observation that ChatGPT is an “iPhone moment” for AI, suggestive of a vast secular growth story in its applications.
Fairly obviously, you can distinguish between those that make AI, those that sell AI and those that use (or repackage) AI. The firm points out that growth cycles, when they get going, can be stronger and last longer than anyone expects. This leads me to point out that a possible way to play this if you are a skittish investor is by using the trend-following approach that I described in April.
Sands offered a fairly short-term scenario where global AI spend could get to $1-trillion in fairly short order. Part of this is predicated on computing availability being up by a factor of 75 by 2030. The PC and internet eras were already very good for corporate profitability, but the firm expects even more from AI. Smarter apps are going to drive shorter upgrade cycles for the devices that will run them.
It was Sands’s commentary that prompted me to investigate the total factor productivity implications of AI, covered in this month’s Middle Road column, and also its example of the law-and-order benefits of Axon camera sets that piqued my interest in the policing and general productivity benefits of AI.
Accordingly, they have earned the last word. Long-term quality investing works over time — not all the time.
The author works at Vunani Securities. Opinions are strictly personal
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.