Dissing Cathie, Man-U and Wall Street analysts in a single AI-focused tech-fund presentation

You have to congratulate a portfolio manager who disses Cathie Wood, Manchester United and Wall Street analysts in a single presentation.

Kudos, therefore, to Richard Clode , co-lead manager at the Janus Henderson Global Technology Leaders Strategy, who achieved this remarkable feat in under 30-minutes at the 2024 Glacier by Sanlam International Seminar, held recently in Pretoria, South Africa.

Anyone can make the starting roster

The first of his three infractions occurred during his opening remarks. The portfolio manager joked that given Manchester United’s current form, “anyone in the room could get onto the team”.

His football credentials are unclear; but Clode knows a bit about long-term performance given his fund’s 35-year track record, including a 16% annual compounded return over the decade ending 2023. This, he said, meant the fund had made around USD4?billion for its clients over the period.

Clode’s talk dealt with how investors should position for the generative artificial intelligence (AI) trend, which is still in its infancy. “Generative AI is the next major technology inflection or wave,” he said.

More specifically, it stands out as ‘the fourth wave of compute’ following on from the mainframe; personal computing (PC) and internet; and mobile and cloud waves. The presenter singled out the launch of the Apple iPhone as the starting point of the third wave, with Chat GPT flagged as the inflection or starting point for the current, fourth wave.

Identifying tech waves is important to asset managers, financial advisers and investors alike. According to Clode, financial advisers who correctly identified the last compute trend, and ensured their clients were exposed to the winning companies in that trend, would have made their clients incredibly wealthy.

One of the important lessons shared during the presentation was the need for investors to be active to succeed.

The winners of the last technology wave tended not to be the winners of the next one.

Countless Kodak moments in tech ‘wash’

His comment is fairly intuitive: after all, we watched IBM dominate the mainframe computing trend to make way for the likes of Compaq, Dell and HP as the PC and internet age took hold. And those names yielded to the likes of Apple and Samsung during the so-called third wave of mobile, smartphone and cloud.

You have to be active because benchmarks are backward looking, they reward past success ... you have got to identify the new winners in each technology wave.

The investment expert then set about schooling his audience on common errors in using price-earnings (PE) to pick shares. First off, he said those who baulk at high PE numbers stand to miss out on much of the return dividend that a technology wave brings.

“The ‘E’ or earnings number is a consensus estimate by some over-paid Wall Street analyst sitting in New York and San Francisco; they are economists who are not good at forecasting things,” Clode said, delivering the second ‘diss’ mentioned in today’s headline and opening paragraph.

Case in point, around 18-months back these analysts were forecasting USD6,00 per share for Nvidia’s 12-month earnings compared to the USD25,00 actual. These and other wide misses show why asset managers have to think carefully about the “E” in PE, especially during technology inflections.

They also serve as a lesson for financial advisers who must endure the “Oh, but the PE is too high” retort whenever they propose tech shares to clients.

No Naspers to monetise the generative AI wave

There was an urgent message for South Africa-based financial advisers and their clients, being that the local market offers zero exposure to the fourth ‘wave of compute’.

Clode said that local investors had struck it lucky during the mobile, smartphone and cloud wave thanks to a fortuitous investment by Naspers into Tencent. He then stressed that there was no similar ‘silver bullet’ moment waiting in the wings for exposure to generative AI.

South African investors will have to go offshore, and failing to do so will leave them ‘high and dry’ as firms leverage the advent of so-called transformer models. An article on the method, hosted on Generative AI-trend ‘winner’ Nvidia.com describes a transformer model as “a neural network that learns context and thus meaning by tracking relationships in sequential data like the words in this sentence”. It is worth noting that the final “T” in Chat GPT stands for “transformer”.

According to Clode, transformer models “learn for themselves” at a scale unimagined just a few years ago. “We have gone from models that assess a million parameters, three years ago, to almost two trillion parameters today; as you exponentially make the model bigger, it can do amazing things,” he said.

Most readers will have experienced the power of transformer models when using the likes of Chat GPT, Firefly and Sora which generate prompted text, image and video content respectively. Although already shaken by the speed of these innovations, attendees were warned to expect more.

You will encounter everything as a first draft, and be happy

Imagine a future where everything you encounter is in first draft. This is the reality being developed under the Microsoft 365 Copilot initiative. “Generative AI means that everything you encounter will be in first draft, and you will use that template and tweak for compliance etc,” Clode said.

The resulting productivity gains will filter through into corporate results and portfolio returns. The caveat is that for the ‘fourth wave of compute’ to truly take hold, firms will have to get their data into a single place, probably the cloud, and ensure this data is appropriately tagged.

The good news for financial advisers and investors is that you are not too late to benefit from generative AI. To illustrate, if you had dumped your Apple shares just 18-months into the third wave, you would have been very unhappy indeed.

Over the next decade, the worst-performing of the so-called FAANG shares outperformed global equities by almost 800%, and the best performing by 17300%. FAANG was an acronym for Facebook, Apple, Amazon, Netflix and Google. PS, Facebook has since been rechristened Meta, and Google, Alphabet.

Clode offered some valuable tips on building a portfolio to benefit from the generative AI wave, including exposure to the various layers that underpin the technology.

“There are still plenty of investment opportunities in AI; we have only seen the beginning of it [as evidenced by] Nvidia in the AI-training realm,” he said, before lambasting the mainstream investment media for being too narrowly focused on the Magnificent Seven. He was equally critical of AI-focused exchange traded funds (ETFs) which contain too many companies that have “absolutely nothing to do with AI”.

High “E” is not an investment leprosy

The clear warning to attendees and lay investors: you should avoid buying ETFs or other funds that are stacked full of companies that will not benefit from AI, and even worse, risk being net losers as the trend plays out.

It is also important to realise that the high “E” number in some of today’s favoured tech shares should be assessed in light of the chip appetite among the large ‘scalers’ of generative AI, such as Google, Meta and Microsoft, and governments.

The key to investing [in this technology wave] is understanding that that “E” can be wrong, both ways.

?He noted that Tesla’s “E” ended up being a lot lower than the market expected whereas Nvidia’s “E” number was way higher than anticipated. The presenter then set about smashing two myths. The first, manufactured by the likes of CNBC, is the “Magnificent Seven or bust” stance.

Clode dismissed this view as nonsensical, saying that more than 25 of the stocks in his 50-stock fund had gone up by 50% or more over 2023. And the second, was a myth perpetuated by Cathie Woods of Ark Funds fame, being that you had to buy super-expensive, unprofitable stocks to play in the tech trends space.

Yes, tech can be profitable

“Cathie Wood would love to tell you that that is true, it is not … our strategy has an average PE of about 23x and we own zero unprofitable tech companies as of 2024,” Clode said, completing the promised ‘diss’ hattrick. Clode’s fund only holds five shares with a PE over 40, making up just 6% of the fund.

“You can get more exposure to generative AI, innovation, and really exciting secular growth within the sweet-spot of mid- to high-teen PEs, up to a maximum 40x,” Clode concluded. His fund has delivered 12% compounded annually over almost 30-years, surviving the Dotcom crash, global financial crisis (GFC), COVID pandemic and more recent tech sector corrections.