February 28, 2023

What Makes a Tech Bubble

  1. What Fu Peng Said
  2. What Taleb Said

You probably cannot find better investment than tech company stocks for the past couple of decades. And being part of a unicorn company is the dream job for domestic or international students: It's arguably the easiest way for middle class to become financially independent. However, as the Fed started to raise interest rates, we witnessed a huge downturn for the tech industry. The once glamorous programmers suffer from massive layoffs. Investors mourn sharp losses.

As an employee working for a big tech company and worrying deeply about the job security, I start to wonder the nature of the rapid growth and even faster collapse of the tech industry. Here, I summarize my findings by reading articles written by Fu Peng (付鹏), a Chinese economist and Nassim Nicholas Taleb, one of the greatest "troll"s of our generation.

What Fu Peng Said

Fu Peng essentially argues that the tech bubble is a result of excessive capital derived from unusually low interest rates. And the bubble is further inflated by, ironically, the slow growth and even the decrease of productivity. Take the post-pandemic tech bubble as an example. The commence of China–United States trade war signifies the reverse of globalization and hence negatively impacts productivity of companies from both sides. In the meantime, investors, equipped with the extreme liquidity as a result of near-zero interest rates after the pandemic, were eager to spend their free money. Hence, they naturally focused on the fastest and easiest way to make money -- tech companies. After all, all you need is a good story for the IPO, after which it's party time. For big tech, their stock prices are easily lifted by share buybacks. So, it's still a pretty good deal.

So, the cycle begins. The management correspondingly shifted their focus from real innovation (i.e., technology that works for real) to those that can only make a bombshell. Their business model is no longer the classic: Devise something that works and make profit out of it. Instead, it's to lure investors with easy money. A side-effect is the spoiled programmers. The rapid expansion of tech companies demands more workforce (pun intended). So, we get ridiculously high compensation, benefits, and the privilege to work from home. Correspondingly, what we talk about daily is not technical issues as "hackers" used to do. Instead, we mostly chat about the stock market, compensation, and the housing market.

Now, tech companies do not have access to easy money and have to please the impatient investors with profitability. The moves are straightforward for the management: cut the expense as well as the projects that won't produce profit anytime soon. The good news is that people restart to make real innovation. The bad news? A new cycle begins sooner or later.

What Taleb Said

While I haven't got a chance to read Taleb's understanding (or teasing) of the tech bubble, the nonlinearity (i.e., winner takes all) he discussed in his book "Fooled by Randomness" is quite relevant in understanding the nature of the tech industry and hence the tech bubble. In a nutshell, Taleb attributed the success of not-the-best-but-dominating products such as the QWERTY keyboard layout and Microsoft office to path dependence. In other words, users simply get used to these technology even though they are far from being the best.

This reminds me of the notorious business model of Chinese tech companies. They spend tons of money on a low-quality product just to make the users depend on it. After that, they raise the price of the product to profit. A character of this business model is that the competition is not about technology but a game of capital. Put it differently, programmers are asked to deliver something as fast as they can and the technology used in it is normally a low priority. A US counterpart of the business model is Uber and Lyft, who still lose money years after their IPOs. So, Lyft's stock price is hit hard by the interest rate hike: it fails to make its users dependent and thus loses the investment value.