Tech Bubbles: What You Need To Know

Tech Bubbles: What You Need To Know

Submitted by Thomas Kirchner of Camelot Portfolios


  • Today’s tech market shares similarities with prior bubbles.

  • Investors should reduce tech allocations.

  • Limited risk to broader market

March 10, 2000 marked the peak of the dot-com bubble. Almost to the day 21 years later, a new tech bubble is in the process of bursting. The peak of the current NASDAQ-bubble occurred on February 9, while the 2020 Corona-trough had its one-year anniversary on March 23rd. If we want to know whether we are in a bubble yet again, we need to examine its characteristics.

Analogies abound

While the specifics vary, there are many similarities between the tech bubble over 20 years ago and the one that we may be in today.

  • Unproven business models
    One of the reasons why SPACs have been so active in the M&A space is because they take public companies without profits that would find going public to be be quite difficult otherwise. For example, helicopter ride sharing may be cool. But it’s far from clear that it will ever be a viable business, if only due to the energy inefficiency of rotor-based levitation compared to the highly efficient lift of fixed-wing aircraft, whose efficiency beats even that of land-based transportation for longer distances.

  • Overly optimistic market size projections
    The top 5 car manufacturers (Toyota, Volkswagen, Ford, Honda, Nissan) have a combined market share of about 1/3 of the worldwide car market. Their combined market capitalization is around $200bn. We find it hard to imagine a scenario for market share and margins in which a valuation of Tesla of over $800 bn could be justified. There simply aren’t enough car buyers in the world, even when you account for additional business lines that Tesla may have. A similar calculation can be made for the market size and pricing power of Covid-vaccines, as we wrote in January.

  • Ignoring margin compression and commoditization
    Fiber optic cables used to be very profitable; until everyone laid them everywhere and the price of bandwidth crashed. The same phenomenon occurred with solar cells – their prices remained elevated until Chinese manufacturers decided to build not just one, but several giga factories. In today’s boom, Uber and Lyft are going after the legacy taxi and livery market and struggle to make a profit despite charging exorbitant commissions. As a result of their competition, taxi fares have dropped significantly, which in turn has depressed the pricing power of ride share apps. We can think of several me-too rideshare apps that disappeared as quickly as they popped up.

  • Winner-takes-it-all risk
    Facebook is a great business, yet not the first one of its kind. Just like Netscape was the first internet browser whose business died when Microsoft bundled its browser for free with Windows, MySpace became irrelevant when Facebook induced many kids to switch. We have been invited to join quite a few me-too social networks since then, but none has been able to steal substantial market share from Facebook. Of course, that can change any day when someone comes up with a better mouse trap, at which point we would not want to have bought Facebook shares at a market cap of $800bn .

  • Underestimating obsolescence
    If you were invested in Amazon many years ago, you’d be doing quite well today. But Amazon is an outlier. Survivorship has otherwise been very low and obsolescence is high. Palm Pilot and Motorola’s Razr are history. Similarly, many companies that are tech leaders today and that have driven most tech returns more recently didn’t even exist in 2000 – think Facebook and Twitter. Remember Angry Birds? Its maker Rovio went public in 2017 in the midst of the craze for €11.50 per share and now trades at half that, despite growing earnings.

  • Margin debt at new highs
    Charts showing record levels of margin debt circulate on the internet. The growth looks exponential and commentators claim that levels are unsustainable and represent bubble levels. However, by our calculation, margin debt is not particularly elevated when expressed as a percentage of S&P 500 market capitalization. In fact, when you look at past extremes in the market and the subsequent pullbacks, margin debt has tracked very closely the pullback in the market. Similarly, in strong rallies, margin debt expands at roughly the same speed as the market. This phenomenon has been persistent in the current expansion, irrespective of whether you put its starting point at the 2009 trough or the 2016 beginning of the latest move higher.

  • Novice investors lured by message boards
    Silicon Investor and the Yahoo message board seemed to be driving a lot of the stock market mania during the late 1990s. Today it seems to be the WallStreetBets forum on Reddit. We have learned from Congressional testimony that stock promoters in these forums have made millions, while reporters of the Wall Street Journal have found novice investors who paid $300 for GameStop with money borrowed on credit cards.

  • Regulatory risk
    Often overlooked is the risk of regulatory action, because the dot-com bubble was not followed by a major regulatory clampdown on technology companies. Yet, if we look back at the investment boom in radio stations in the 1920 and 1930s, which was state-of-the-art communications technology at the time, the government created the Federal Communications Commission (rather, its predecessor) in response to the growing might of radio monopolies. Along with the commission came a byzantine web of ownership rules that survive to this day. Social media companies are already under threat from regulation of hate speech (penalties of up to $50 million per violation in Europe) and it is not hard to foresee a scenario where even stronger regulatory requirements could regulate away the profits of these companies.

But tech works?

Sure, tech has worked very well indeed in recent years. It depends, however, on the time period you look at. Over the last 10 years, the NASDAQ has outperformed the S&P 500 by 263 percentage points. This strong performance helps it offset much of its underperformance of the 2000-2010 period. Since the peak of equity markets on March 10, 2000, the NASDAQ underperformed the S&P 500 by 46 percentage points over the next decade.

In other words: technology outperforms, but, like any other asset class only in the right environment. Since environments tend to change, outperformance can be mean-reverting. After a decade of technology-sector dominance the risks favor future technology-sector underperformance for an extended period. The trillion dollar question is: when does the tide turn?

Not many or possibly no one has the market timing skill to answer that question, but we can go back and look at the last turning of the tide.

After its 2000 peak, it took the NASDAQ about a year to give back its outperformance that had started in late 1996 and fall back to the level of the S&P 500. However, losses did not end there but continued until the S&P 500 started turning around in 2004. Overall, the bubble was a 5-year roundtrip. For comparison, we are now in at least year 10 of a technology rally (we won’t start debating where we should set the starting point). The quick pullback that started in 2000 suggests that once mean-reversion starts, investors will have little time to adjust their allocations. Therefore, investors with a slow reaction time, in particular those whose allocation decisions need to be sanctioned by a process that involves consultants, investment committees and boards, would be well advised to trim their technology sooner rather than later.

The broader market is, in our view, not as overvalued as it was in 2000 because valuations look less stretched than they did back then. However, passive index investors could still suffer due to the high allocation of market-weighted indices to highly valued tech companies. The S&P 500 in particular looks vulnerable from a tech pullback as more than 25% of its top holdings are tech companies.

Tyler Durden
Mon, 03/29/2021 – 14:00

Share DeepPol
Generated by Feedzy