The name World Online probably doesn't mean much to many people, but it's a blast from the past that came back to me following the recent failure of the planned initial public offering (IPO) of Aussie AI data centre company Firmus.
Back when I was a young financial journalist in London, the dotcom bubble, and its subsequent bursting, proved a valuable educational experience. The heady days of the speculative bubble period saw more and more companies pop up, seemingly out of nowhere, with all sorts of ideas about what they were going to do on the-then relatively new internet.
And, of course, many of their founders needed money to fund their dreams and turned to the sharemarket to find it.
With huge excitement about the internet, and what it could enable, investors were throwing money at anything with a name ending in dotcom, or something similar, like proverbial drunken sailors. Excitement, greed and the fear of missing out, saw a stampede of people seeking their own slice of the dotcom boom.
This meant a listed dotcom company could release its financial results, showing widening losses and little or no revenue, and its share price would rise. That went against what I'd been taught, in my fledgling journalism career, should happen in such a scenario.
It was late in the 20th century, with the much hyped new millennium looming. New Year 2000 passed without the Y2K bug wiping us all out and the dotcom boom continued, at least initially.
The worm turned, however, and in mid-March 2000 Dutch company World Online's IPO heralded a key turning point. I found a useful article from The Observer, looking back at the bursting of the dotcom bubble. It was written a decade later, in 2010. And of course, World Online features.
The gong that signalled the end of the business day on the Amsterdam stock exchange sounded particularly ominously when it rang out on Friday, 17 March, a decade ago. It closed the first day's trading in shares of World Online. While previous dotcom IPOs had rocketed skywards, shares in the pan-European internet service provider halted a mere €0.20 above their offer price. The firm was still valued at an eye-watering €12bn (£7.2bn) but in web terms the performance was a disaster. The internet bubble, which had created a slew of paper millionaires and a fair clutch of real ones, was beginning to burst. The so-called "new economy" was turning out to be nothing of the sort: it was just boom and bust on steroids.
After the weekend, World Online shares crashed to earth.
Later that year Italian company Tiscali completed a £3bn takeover of World Online. And a decade later World Online investors managed to extract about €110 million from investment banks ABN Amro and Goldman Sachs after a court ruled the lead managers had misled investors during the IPO.
Despite the tumultuous events of the dotcom crash, the internet survived. This article is published on it. And companies that emerged in that heady period such as Google in 1998 - although its IPO wasn't until 2004 - and Amazon, are among the biggest around today.
Neocloud provider Firmus didn't even get as far as World Online. That's because the Nvidia-backed firm pulled the plug on what had been touted as the biggest Australian IPO since Telstra's in 1997 before it even happened. A key reason was potential investors thought the price was too high. Firmus had been looking to raise about A$7 billion at A$11 a share, valuing the company at A$44 billion.
As Bloomberg put it;
The failed [Firmus] deal underscores growing concern over how much capital AI infrastructure companies are demanding from public markets at a time when borrowing costs are rising. Much of Firmus’ valuation was based on the company successfully building a pipeline of data centers across Asia serving customers such as Meta and OpenAI.
The listing was poised to be the next beat of Firmus co-founder Oliver Curtis’s comeback story after serving time in prison on insider trading charges. It would also have cemented his position among the AI boom’s ultra-wealthy. Now the question is where the company will find the capital to keep funding its data center rollout.
And on Friday The Financial Times ran an article with the headline: 'The data centre IPO debacle that sparked a market meltdown.' It opens like this;
The tremors rippling through global stock markets this week amid rising concern about the valuations of AI-related companies and the huge sums they are spending can be traced back to the decision to abandon an Australian blockbuster IPO.
The broader macro-economic picture featuring an oil shock, challenging inflation and rising interest rates, doesn't help.
Whether the failed Firmus IPO turns out to be a straw that helped break the camel's back, in terms of the AI sharemarket bubble, remains to be seen. History may show the worm had already turned, or perhaps something yet to happen will be regarded as the key turning point.
Either way, the sheer hype around AI certainly reminds me of the dotcom hype back in the day.
As I was writing this on Friday afternoon, an article from Kristalina Georgieva, CEO of the International Monetary Fund, arrived in my email inbox. It highlights the scale of this bubble.
"Investment is flowing into AI on a scale that will likely exceed the vast sums that went into building the railroads, the electricity grid, and the internet," Georgieva wrote.
And risks stemming from the circular financing of the AI sector, often featuring Nvidia, are regularly being highlighted, including by the very sober institution known as the Bank for International Settlements. Then there was an episode of Bloomberg's Odd Lots podcast last year, featuring Paul Kedrosky of US venture capital firm SK Ventures, who suggested the AI bubble was "like every previous bubble rolled into one." Kedrosky cited a real estate aspect, a technology aspect, "exotic financing structures" to fund it, plus talk of government bailouts and backstops.
Whatever happens with share prices and IPOs, AI - like the internet - will still be with us. Its presence is already huge and it's evolving so quickly. The questions about what its impact will be on productivity, jobs, depletion of key resources such as water, electricity demand and the broader environment, democracy, plus cyber-security and its potential to kill us all, to cite just a few areas, are hot topics of debate.
In terms of jobs and the labour market, Georgieva said IMF research suggests potential for massive disruption.
"According to our research, roughly 40 percent of jobs globally are exposed to AI. In advanced economies, exposure rises to about 60 percent. Within these bands of exposure, some jobs will likely be augmented, while others may be replaced by the technology. It is still too early to predict the toll on employment ... but the nature of work is changing, and this is likely to accelerate," she said.
Against this backdrop debate about regulating AI, or letting unfettered capitalism run its course, is raging.
On the subject of regulation, Morrison CEO Paul Newfield made some interesting comments at last week's Institute of Finance Professionals New Zealand (INFINZ) national conference in Auckland. Morrison is probably best known in New Zealand as the manager of Infratil. All up Morrison has more than US$30 billion of assets under management, including investments in AI data centre businesses.
"If you actually look at history, how does regulation happen? Regulation generally happens after many, many years of cumulative harm," he said.
Newfield had a chart of US road fatalities, and noted they rose through to the 1960s, through an era of industry self-regulation.
"It was only when they got to a really horrific peak, like tens of thousands of human deaths in the US, when the government said, 'actually, we have to regulate this," Newfield said.
"We'll bring in seatbelt laws, we'll bring in safety laws. And this is actually true of a bunch of other things. Think about how long it took for health and safety laws to come in after the Industrial Revolution, how much harm was done by social media before you've just now started to see some regulation."
"When I shared this analysis with an executive from one of the big [AI] frontier labs, I don't know if this is hopeful or not, but he said, 'oh, there's one exception to that rule, and that's the Patriot Act, which was brought in after 9/11."
"So 9/11 obviously happened on the 11th of September. [On] the 23rd of October, the Patriot Act was passed through the House and Senate on one day, and signed into law by the President the next day. So when you have one big horrific event, you can focus political activity. And so I said maybe it will happen faster. I don't know if that makes you feel good or not, but certainly showed you how they are thinking about the risks," Newfield, who wants data centre regulation, said.
There's an old saying that the sharemarket is driven by greed or fear. Time will tell whether the failed Firmus IPO will prove to be a key marker in a major switch in attitude towards AI stocks from greed to fear.
One of the key events in the bursting of the dotcom sharemarket bubble occurred in the, perhaps unlikely, location of Amsterdam. Could it be that one of the key events in the bursting of an AI sharemarket bubble has occurred in Australia?
*This article was first published in our email for paying subscribers. See here for more details and how to subscribe.
We welcome your comments below. If you are not already registered, please register to comment
Remember we welcome robust, respectful and insightful debate. We don't welcome abusive or defamatory comments and will de-register those repeatedly making such comments. Our current comment policy is here.