Regarding… Vol 2 no 13 – AI’s Dirty Little Secret

Dear Fellow Investors and Friends

Welcome! I do appreciate you taking the time to read this.

I’m Piet Viljoen, and today is Thursday, the 11th of April, the 102nd day of the year. There are only 264 days left until the end of the year. That’s right… 264.

This week, I’ve had the pleasure of accompanying some international investors to visits with the CEOs of some outstanding South African companies. The investors were all clients or prospective clients of our associate fund, Desert Lion, managed by Rudi van Niekerk.

We met with Johnny Copelyn of HCI, Zak Calisto of Karooooo, Marco van Niekerk of Outdoor Investment Holdings and Michael Jordaan of Montegray Capital. They were kind enough to take time out of their busy day to meet with these investors and share the ups and downs of doing business in South Africa. Nothing was sugar-coated, but the realisation that it wasn’t all sackcloth and ashes was clear.

Shock and awe were the order of the day – shock at how cheap the businesses were and awe at how good management was.

I’m pleased to report we also had the opportunity to dine at some excellent establishments. If you’re ever looking for a place for a special occasion, you could do a lot worse than Waterside in the Cape Town Waterfront.

Hats off to Rudi for getting this group of investors to our shores – we don’t see many around here anymore.

Today, I’d like to share the third part of my trilogy riff about the global energy situation. Two weeks ago, in “The Trolley Problem”, I wrote about my hopes that humankind could effect a third energy transition towards clean nuclear energy.

Last week, in “Malthusian Malady”,  I wrote about my hopes that positive-sum thinking would triumph over negative-sum thinking in this transition.

Today, I want to share my thoughts on why we need a proper transition sooner rather than later.

Say what you want about bubbles, but they provide cheap equity funding to companies building out the economy’s cutting edge. In turn, cheap funding can lead to overcapacity.

And so it was with the birth of the internet, culminating in the TMT bubble. Everybody wanted the internet, but you could only get it by routers and fibre – the “picks and shovels” of the internet. The installed capacity of routers and fibre rocketed because you could get free money to build out the “backbone” of the internet. Of course, actual demand fell short of the projections provided by the fever dreams of investment bankers selling stock to the public.

Remember “dark fibre”?

In the background, the oil price, which had been going sideways in a broad range between $10 and $30 for at least 20 years, broke to a new high in 2003. Surprise! All these “picks and shovels” used a lot of energy.

What was true then remains true today: Computing requires energy. Supercomputers need super amounts of energy. AI is the new big thing in computing, and the picks and shovels of AI are cutting-edge chips. And guess what – these chips are inveterate users of energy. This article in Scientific American sets out the issues.

In January, OpenAI’s CEO Sam Altman said in Davos that an energy breakthrough is necessary for future artificial intelligence, which will consume vastly more power than people expect. The International Energy Agency says that after consuming an estimated 460 terawatt-hours (TWh) in 2022, data centres’ total electricity consumption could reach over 1,000 TWh in 2026, equivalent to Japan’s.

This week, ARM CEO Rene Haas warned that AI data centers could consume as much as 20% – 25% of U.S. power requirements by the end of the decade, a significant increase from the current 4% or less.

Of course, we are installing a lot of megawatts of “green” energy”.

Aren’t we?

We are now halfway between 1997 – when delegates of nearly 200 nations first met in Kyoto to agree on commitments to limit greenhouse gas emissions – and 2050, the target date for decarbonising the global energy system. And the numbers are out. All we have managed to do so far is a slight decline in the share of fossil fuel in the world’s primary energy consumption – from 86% in 1997 to 82% in 2022.

Despite 15 years of subsidies, propaganda and “market interventions”, renewables comprise less than 10% of the global energy supply. And most of that is in the developed world.

Transitioning to “green energy” is a marathon, not a sprint. It took us a hundred years to build a global economy that runs on petrochemicals, and it will take us many decades to unwind it.

But here’s an “energy truth” for you: the energy supply will inevitably move to whatever best satisfies the consumer needs at the lowest cost. The innate desire for progress – a better life – will drive this. PST in action (see my thoughts here last week on PST/ZST).

On the other hand – as discussed last week – Malthusians’ solution to the problem is to have less energy and fewer human beings. Less growth (i.e. degrowth) is more palatable to those of us lucky enough to be chosen to live by the ZST crowd.

Today, you have a two-pronged driver of energy demand, which overwhelms Malthusian reasoning:

• The increased demand for steady and cheap energy is driven by orders of magnitude growth in computing power in the West.
• The Global South requires access to cheap and efficient energy to drive productivity and high growth in per capita GDP.

I am an optimist. I believe in human ingenuity and positive-sum thinking and I believe most people also think that way. The solution to our problem is not forcing a transition to “clean energy” but supporting the developing world in expanding its access to stable and efficient energy sources. Natural gas and nuclear should make up a large proportion of the mix.

The deal to buy Mead Johnson in 2017 looks set to result in unquantifiable litigation risk for the consumer group. An Illinois court awarded $60mn in damages to a mother whose child died after consuming Enfamil baby formula. Reckitt believes there are strong grounds to overturn the verdict. On the other hand, Bayer’s acquisition of Monsanto has – so far – cost it $20bn in lawsuits.

The problem here is these consumer staple businesses have been struggling to grow organically for a long time. So, using their highly-priced paper, they have been making a series of acquisitions to boost growth. And, as we have seen so many times, many – most? – acquisitions fail to add value. And in some cases – like Bayern and Reckitt – it can destroy colossal value.

Here is what Reckitt shareholders have endured:

My take: I have avoided exposure to the “high quality” consumer staple sector for a long time. Now is not the time to change that view. I discussed the industry in my “Squeeze Play” note earlier this year.

Swatch not only makes those funky, brightly coloured watches. They also make the “movements” for many other brands of luxury watches. As such, it is the world’s largest and arguably most influential Swiss watch manufacturer. You might recognise some of these brands:

Luxury goods businesses like LVMH or Richemont have been on a tear over the past few years. Not the Swatch group, which is currently trading at prices approaching its Covid lows:

My take: The Swatch Group is the canary in the coal mine for the watch market specifically and luxury goods businesses generally. And I will continue to avoid exposure to this sector. There will come a day to invest in it, as it has desirable investment characteristics. But we are not there yet.

You’ve probably never heard of this company, nor had I before this week. However, it highlights an important factor when investing in other jurisdictions.

First, a bit of background: U.S. private equity house KKR chose it for its first mainland China investment in 2007 before helping to take it public four years later. Chinese media often describes its chair, Li Liufa, who, along with his wife, owns about 69% of Tianrui, as “the richest man in Henan province”. So, the company comes with good credentials.

Recently, the problematic property situation in China has caused businesses to report losses. Li has apparently pledged parts of his stake in Tianrui for loans to support his company’s operation. Hong Kong regulators allow substantial shareholders in listed firms to borrow against stock and not disclose their pledged position.

And this is what happens when collateral gets called:

My take: Two things. Firstly, China is far from being out of trouble. Secondly, how many other situations like this are there? As they say in the classics when the tide goes out… stay away from China.

Japanese monetary authorities can’t tighten monetary policy. With a debt:GDP ratio of 263%, the highest of any developed nation, any increase in interest rates would cause massive distress in the Japanese financial system.

In his research note, The Solid Ground (a subscription which is well worth it), dated 25 March, Russel Napier writes, “If we take the JGB maturing in June 2033, its coupon of 0,4% implies a decline in the current bond price of more than 60% if the yield were to rise even to match the current inflation rate.”

What can they do? Financial repression, as I have noted before. Force Japanese saving institutions to buy government bonds, keeping yields below inflation. That is exactly what South Africa had to do in the 1980s. We know how this playbook goes.

The market, via the Yen, is testing Japanese monetary and fiscal authorities’ resolve:

My take: Speaking of canaries in the coal mine, the yen is it for global markets. If it weakens, risk assets should do well. The authorities’ reaction will change the environment in a big way.

I have been banging on about the energy sector for the past three weeks, probably boring you to tears by now. But things are happening here. In a week when expectations for interest rate cuts in the USA have been reduced, commodities in general and oil specifically have done well.

I always say there is information in dissonance.

Here is the S&P500 energy sector (ticker $XLE):

It looks a lot like the Yen!

This is a sector where global demand inexorably increases year after year, and this is happening despite the best efforts of the anti-humanity ZSTs to reduce our inclination and ability to use energy. With little capex being expended in the sector (“stranded assets” anyone?) the stage is set for a substantial price signal to kickstart a new wave of investment in developing carbon-emitting energy sources. And, hopefully, nuclear.

My take: The MWI Worldwide Flexible fund (aka the “cockroach”) has a significant allocation to oil-related securities in the “hard asset” section of the portfolio, which, in turn, makes up a quarter of the fund.

Here’s another chart that looks almost like the Yen chart:

With most FMCG retailers struggling globally, Tesco stands out. A stock in an unliked sector of an unliked economy hitting new highs? There is information in that.

My take: On a P/E of 12 for a stable stream of earnings, Tesco remains a bargain. I think the UK is one of the most undervalued markets globally, and Tesco is an interesting play in this regard. I have a significant allocation to the UK in the equity section of the cockroach – which also makes up a quarter of the fund.

Until the beginning of this year, all seven of the Magnificent 7 were hitting new highs regularly (with the exception of Tesla). They were acting in concert.

Now, Apple and Tesla are trending down, and Nvidia is not making new highs. There is increasing diversity in how these bellwether stocks are acting.

When you have a business model that prints money, you shouldn’t be surprised if it attracts competition. At first, the competition is barely noticeable, but it starts having a real impact after a while. Ask Tesla about BYD and the myriad of other Chinese EV makers.

When you have a business that is so dominant that it can exploit the market, antitrust regulators start getting involved. Ask Microsoft how long it took them to recover from the antitrust suit against them in 2001. Then, ask Apple about its current regulatory scrutiny.

In the past week, we have heard that both Meta and Google are starting to design their own chips. Nvidia’s margin of 50% to other tech companies is like throwing chum on shark-infested waters!

My take: The Magnificent 7 is down to the Fabulous Five. And the diversity breakdown is just getting started. I continue to watch from the sidelines.

Yay! The JSE got a new listing!

And WeBuyCars is an exciting listing.

Remember, when you list a company, you effectively sell stock to the public. And insiders will only do that if they get a more than fair price for the stock they sell (which is why I warn against “investing” in IPOs all the time). With the price of assets in South Africa being so low, insiders have no incentive to sell their businesses to the public. So WeBuyCars is not a classic IPO, but a spin-off from its troubled parent, Transaction Capital.

As opposed to IPOs, spin-offs often make good investments. In a spin-off, the parent company usually spins off an unwanted subsidiary to its shareholders. Those shareholders don’t want it either, so they sell it, lowering the share price. Which, in turn, sets it up as a potentially good investment.

But WeBuyCars is also not that kind of animal. Instead, it is a reasonably good business – that everyone likes – being spun off. It’s a reverse spin-off if you will.

My take: I think the value investor will be better off spending her time evaluating what remains of the rump of Transaction Capital rather than getting involved in WeBuyCars.

The Seven Virtues of Great Investors, by Jason Zweig.

I return to this piece occasionally for Jason’s excellent writing style and to remind myself what I’m trying to do here and how I should approach it.

A checklist of sanity in an insane environment, if you will. I hope you find as much value in it as I do.

That’s all for this week.

Eid Mubarak to my Muslim readers. And to all my readers, be careful out there.

Piet Viljoen
RECM