Extract from ABC News
Analysis
The market valuations of AI companies have been bloated as international investment funds have poured into them. (Unsplash: Steve Johnson)
What, if anything, has the past financial year taught us?
Not much, apparently, as it draws to a close this afternoon in much the same way it began.
Global geopolitics is in the kind of disarray unseen in almost a century, there is social and political upheaval on almost every continent and the global economy is facing serious issues on all fronts.
Inflation is up, growth is stunted and an energy crisis that has been suppressed for most of this year is once again threatening chaos.
And yet, investors sail blithely on.
Overshadowing the tumult has been a financial boom beyond almost anything ever witnessed before, focused entirely on the perceived benefits wrought by a brave new world of AI that threatens to up-end our lives.
While it's possible the new technology will enhance productivity, it's equally possible that too much money has been spent chasing elusive gains.
In any case, it is a boom centred on a mere handful of corporations that have embarked on a race for domination that not all will win. Along the way, their market valuations have been bloated beyond any reasonable level as international investment funds have poured into them.
As Australia's big superannuation groups prepare to unveil yet another year of solid, if not spectacular gains, it's worth noting the returns have been driven by what many perceive as a rapidly inflating bubble centred around one key issue: AI spending.
Here's how it works.
Wall Street now accounts for about 65 per cent of the entire world's stock valuations. Smaller markets like ours barely get a look-in.
And on Wall Street, it is the top 10 titans that have attracted more than a third of all the cash thrown at the New York Stock Exchange.
In terms of market concentration, it is a top-heavy profile that has outdone the tech boom of the 1990s and the burning fever evident just prior to 1930. And we all know how those episodes ended.
A spanner in the AI works
When it comes to races, there's usually only one clear winner. And quite often, the odds-on favourite becomes an also-ran.
For all the talk and all the promise, the time is rapidly approaching for AI to start delivering.
Huge amounts of investment, trillions of dollars, have been thrown at AI, initially into model development, then semiconductor and cloud computing and now into hard asset build-outs with data centres.
They, in turn, require vast amounts of energy and water. And that's where the newest set of problems begin.
While the race to develop the technology has been a sprint, little thought has been given to the problems and constraints associated with the rollout.
Now, suddenly, the brakes are being applied.
While the race to develop AI has been a sprint, little thought has been given to the constraints associated with the rollout, such as controversies around data centres. (ABC News: Rhiannon Stevens)
Building vast industrial parks complete with cooling towers and back-up diesel power plants in suburban areas is no simple matter.
Already, residents in some of Australia's biggest cities are jacking up over the sudden incursion of data centres through metropolitan areas, while authorities are questioning how to meet a doubling in their electricity demand by 2030.
This, in a nation in the midst of a full-blown energy transition, where power constraints were already on the cards.
Those roadblocks are likely to slow the pace of AI integration into the broader economy, which will weigh on earnings for the companies that stretched themselves to develop the technology.
That, in turn, is likely to raise the costs for corporations wanting to employ the technology, all of which may undermine the economics associated with the entire frenetic race.
The merry-go-round is slowing
The AI boom, once a blanket bet over all the major tech players, is fragmenting.
And investors are beginning to question whether the huge outlays will ever pay off.
Software groups came under attack earlier this year when everyone suddenly twigged that AI might just undercut the need for software. Most are still struggling to recover.
Microsoft, Oracle and even Meta, the owner of Facebook and Instagram, have all shed more than 30 per cent since their peak last year.
Investors have since targeted chipmakers and memory development, key components for AI, and shortages of high-quality gear are boosting earnings in spectacular style.
But that's pushing up the price of semiconductors and raising costs for the big tech developers, undermining their expected future returns.
Then there's the financing.
The tech giants funded the early stages of AI development with the vast amounts of cash they were throwing off their existing operations. The more they spent, the more investors loved them.
But their vast capital requirements combined with rapidly rising costs have forced them to tap credit markets.
Instead of spare cash, they're now raising debt, which ramps up the risks dramatically.
And it's only likely to increase. Research firm Gartner estimates global AI spending will hit $US2.6 trillion this calendar year, while Goldman Sachs estimates a further $US7.3 trillion will be spent by the end of the decade, much of it on data centres.
And that's the problem, according to Swissquote's Ipek Ozkardeskaya.
"These huge investments are also draining big tech's free cashflow, obliging companies to take on more debt and putting their valuations under pressure," she says.
Apple increased its prices last week, attributing the rise to AI demand driving up the cost of components. (Unsplash: Rami Al-zayat)
Ozkardeskaya says investors are staying on the AI train, but they are jumping carriages.
As big tech stagnates, chipmakers have pumped prices to deliver stellar results.
But that is squeezing profits down the line at companies like Apple, which increased prices last week.
"Investors will eventually realise that revenues cannot continue growing at the current pace forever," Ozkardeskaya says.
In another ominous warning over the weekend, the Bank for International Settlements — often referred to as the Central Bank for Central Banks — highlighted the speed of the AI boom, questioned its sustainability and openly raised the prospect of a financial crisis.
In addition to the growing social anxiety around the risk AI posed to jobs, the BIS noted that the intense competition between the tech giants could lead to the kind of overinvestment seen in previous boom and bust cycles.
Buckle up
You don't need to delve too deep to discover why your super fund has focused offshore.
While there's still a day or so to go, the local market barely eked out a gain during the past financial year, up just 2.6 per cent.
Why? We don't do tech, or very little of it.
Compare that to the Nasdaq, which pushed through a 28 per cent lift for the year, helping the broader New York market to a healthy 19.7 per cent gain.
But it appears to be getting a dose of the wobbles. There were big losses on Friday night, even among the memory chip makers like Micron, which took a 6 per cent hit despite a huge lift in earnings.
Wall Street wasn't alone. The tech-heavy South Korean market dropped 10 per cent last Tuesday, backed up by further heavy falls on Thursday, as hedge funds dumped their tech stocks at the quickest pace on record.
Major corrections and even crashes usually are preceded by heightened volatility.
If we're entering just such a phase, it hasn't been without warning.
