Comparing previous tech booms and bust cycles with today’s AI surge highlights some notable patterns. Historically, undervalued stocks with low price/earnings ratios have outperformed during market downturns that followed unsustainable rises in valuations. This underscores the importance of focusing on valuation to manage risk. Understanding historical parallels can help inform smarter investment decisions amid market uncertainty.
This is a shortened version of the paper A feeling of déjà vu: Parallels between the dotcom era and today’s AI surge by Carmine de Franco
As we stand on the cusp of a new technological revolution driven by artificial intelligence, many investors are drawing comparisons between the turn of the century’s dotcom boom and bust and today’s technology sector backdrop.
Perhaps now more than ever, it is vital to understand historical patterns to better navigate current uncertainties.
The late 20th century witnessed the rise of personal computing with Microsoft’s release of Windows 95. By 2000, Windows held approximately 90% of the market share, marking a pivotal point in the technology sector.1
This period ushered in the internet revolution, leading to immense growth in tech stocks and a market bubble that peaked in March 2000. Microsoft’s stock soared during this time, delivering an extraordinary 60% annualised return from the release until the peak.2
However, the subsequent crash in 2000 was severe. It took nearly 14 years for Microsoft’s stock to recover. From March 2000 to December 2014, major indices like the S&P 500 and Nasdaq experienced sluggish growth, with returns below risk-free assets like US Treasuries and on a par with cash.
Now and then
Fast forward to today, the market exhibits several similarities to that late 1990s period. The recent surge in AI technology, fuelled by firms like OpenAI and Google’s DeepMind, is comparable to the internet boom.
The valuation metrics, particularly the price-to-earnings (P/E) ratios of the S&P 500, have increased significantly, reaching levels dangerously close to those seen before the 2000 crash. Historically, such high valuation levels have been followed by market corrections, when the market tends to revert to its long-term average P/E ratio of around 15.
The market may take comfort from the fact that back in the 1996-2000 period, much of the growth was driven by companies with little or no profits, often in the internet sector. Today, instead, the AI boom is backed by profitable companies with solid earnings.
However, the earnings per share (EPS) of the S&P 500 in 1996 was like today’s figure, indicating that current growth, although more grounded in fundamental profits rather than speculative valuations, shares some similarities with that period.
Furthermore, today’s macroeconomic landscape looks, if anything, worse compared to 1996, when inflation was low, hovering around 2%, and the Federal Reserve had more room to manoeuvre with monetary policy than today.
Today, inflation is higher and more persistent, though long-term yields on US Treasuries are lower than in 1996, at around 4%. The US government’s deficit is also significantly larger now, exceeding 5% of GDP, compared to a surplus in the late 1990s.
Financial conditions, including credit spreads, tell a nuanced story. Spreads on investment-grade and high-yield bonds have been stable but show signs of stress similar to those in the late 1990s, especially amid concerns about debt levels used to fund AI infrastructure investments. These macro factors suggest while the environment today is different, vulnerabilities remain.
Emerging ‘value’
This analysis emphasises the importance of valuations in guiding investment decisions. Historically, after the 2000 crash, value-oriented stocks – those with lower P/E ratios – outperformed their more expensive counterparts for several years.
For example, during the period following the dotcom bust, low P/E or ‘value’ stocks consistently delivered superior returns, a trend that persisted until the market re-rated these companies higher.
Currently, valuation disparities between value and ‘growth’ stocks are evident. Value companies are trading at a fraction of the P/E ratios of their more expensive peers, similar to post-2000 conditions. This suggests that an emphasis on undervalued stocks could potentially be a prudent strategy, especially if history repeats itself.
Lessons from the past
The key takeaway from this historical perspective is that speculative excesses in the market – whether during the internet bubble or the current AI wave – could possibly lead to sharp corrections. The 1996-2000 period demonstrates that markets can take more than a decade to recover from such crashes, and the aftermath often favours value stocks that are undervalued relative to their growth prospects.
While the current AI surge is backed by profitable companies, the macroeconomic conditions are likely worse, hence the similarities in valuation levels and market behaviour warrant caution. Investors should be mindful of valuation metrics and consider the potential risks of investing at extreme highs – a selective, thoughtful approach is critical.
[2] Source Bloomberg in USD, from August 24th, 1995 to March 30th, 2000
