The technology industry is in a bad way. To pick one of the big ones at random – Alphabet is down by 33 per cent.
The Ark Innovation ETF (ARKK) is more representative of the industry as a whole. It’s full of speculative money-losing technology companies in areas like energy storage, molecular diagnostics and e-commerce. ARKK’s share price has fallen by 76 per cent in the last two years which tells you a lot about the market’s changed view on speculative technology companies.
Compared to the mega-caps like Google, ARKK's companies are small and risky. But in the scheme of the entire technology industry, they must be among the biggest and least risky.
They're public companies, for one, which means they are mature enough and big enough to subject themselves to public market scrutiny. 96 per cent of them are in North America, where publicly traded companies tend to be much bigger than elsewhere in the world. The median company in ARKK's portfolio is worth between $2-10 billion. ARKK's performance would make you worry for the ecosystem of small, illiquid, private tech.
Last week I came across Asimilar. It's an investment company listed on Aim, and in many ways a catastrophic company. Like ARKK, it invested in a random collection of buzzy technology companies. In Asimilar's case, the target companies were unlisted. They were smaller and riskier. And Asimilar's share price is down by not 76 per cent, as ARKK is, but 96 per cent.
It's hard to get a clear view of how messy things are among private technology companies. As I've written before, they and their backers have every incentive not to mark their values to market. The best strategy for a private technology company is to hunker down, reduce its cash burn, and hope financial conditions improve by the time they need to raise more money. A recent survey by January Ventures, a VC fund, found 81 per cent of startups have enough cash to cover less than 12 months of expenses.
Digging through the data from Dealroom gives some sense of what's happening. Dealroom keeps meticulous data on transactions among venture capital investors and technology companies. It doesn't show anything about valuations, or about how the companies are performing. But it does show the deals: how much funding is flowing, in what regions, to what size companies.
Here are three of the trends from the 2022 Dealroom data.
Startups peaked a year before the market did
The number of startups — defined as a company designed to grow quickly — grew steadily since 2000. The US led the charge but other countries followed. The growth of startups wasn't dented by the aftermath of the Dotcom crash, the 2001 recession, or the global financial crisis. But in the US and in Europe too, the growth in startups began to flatten out after 2017. In 2020, the total number of startups peaked. The number fell slightly in 2021 and 2022.
Given that the US had a big head start, and the most mature ecosystem, you might expect the number of startups in other countries to have kept growing after US growth plateaued. But that's not what the data show. The number of startups peaked all over the world at the same time, in 2020.
Whatever happened in 2020, it took another year for investors to catch on. In Europe, venture capital investment in startups peaked in 2021. It fell by 24 per cent in 2022.
In Ireland, the music did not stop for another year. Venture investment was higher in 2022 than in 2021.
The chart showing the total number of startups reminds me of an old one from 2008. 2008 was the year of the Great Recession and there was much debate over what was causing it. One popular theory was that it had been caused by the bursting of the housing bubble. But the data showed that employment in construction in the US had peaked 20 months before the recession started. The connection between an industry's cycles and the overall economy is unclear, and may be non-existent, or may work with a very long lag.
US tech is far ahead — but Europe is reeling it in
For starters, America has many more startups than Europe. But they're also much better funded.
Dealroom doesn't go into specifics about company valuations. But you can infer them by looking at the total amount of funding in a country, and the total number of startups. This number tells you, not what an average startup is worth, but the average amount of funding raised by the average startup.
There's a huge gap between the amount of funding that was raised last year by American startups and their European cousins. Companies and investors think much bigger over there.
It's not all doom and gloom for European technology. Though it lags America in overall funding, it's striking how steadily Europe has closed the gap with the US in terms of overall VC funding per year. The proportion of VC dollars going to Europe has grown steadily, in good years and bad, since 2000. In 2022, European companies attracted 40 per cent of the funding of US companies — compared to four per cent in 2000.
A sad story is the decline of the Chinese technology industry. Chinese technology funding grew extraordinarily quickly between 2013 and 2018 — from 7.7 per cent of US funding to 68 per cent of it.
In hindsight, Europe's slow and steady approach is probably the better one. Returns on Chinese VC have been terrible, and funding has collapsed by nearly half since 2018. Prequin, a data provider, found that Chinese CV deals had fallen by 44 per cent in the year to October 2022. The crackdown by the Chinese government on technology firms, which started in 2020, won't have helped.
Per capita performance
I've tried to figure out how well Ireland is doing, but it's hard given the population difference, so I've divided a few important metrics by population.
When you rank them like this, Ireland's performance per capita is pretty good. On average it ranked 6th in these metrics out of the 46 countries in Europe (plus the US).
Ranking highly per head of population is good, of course. But it has to be said that operating out of a small country is a handicap. It's harder to hire people, raise money, and sell products. Hopefully, Europe's economy keeps integrating and the problem gets solved that way.