Bubble Trouble?

Amid recent speculation that an AI bubble is building and will subsequently burst, potentially wreaking havoc on the global economy, we highlight some key differences between the circumstances then and the current situation. Some are favorable, others less so.

We look at the deterioration in economic conditions following the corrections in equity markets — referred to as the “dotcom correction” — highlighting variations in the national and regional impacts and across different types of expenditure. We also put the effects into perspective by comparing them with the fallout from the subsequent global financial crisis (GFC).

1. The US growth boom from the mid-1990s came to a sudden halt from mid-2000.

From the mid-1990s to mid-2000, US growth was exceptionally strong and broad-based across expenditure components. Annual real GDP growth averaged 4.3% from 1996 to 2000, peaking at 4.8% in 1999. US real domestic demand growth was stronger still: From 1996 to 2000, annual real growth rates in private consumption and investment averaged 4.6% and 7.8%, respectively. Export growth was also very robust.  

As the dotcom correction intensified from the second quarter of 2000, US real GDP growth ground to a halt.  A “technical” recession, defined as consecutive quarter-over-quarter real GDP contractions, was averted, although only just. 

2. The contraction in US real GDP was modest compared with the subsequent global financial crisis.

While the downturn in US growth from mid-2000 was sudden and sharp, the contractions in US real GDP during 2001 were relatively small compared with other financial stress-related recessions. From peak to trough, US real GDP fell by 0.3%. By comparison, following the GFC, US real GDP dropped by 4% from peak to trough.   

The recovery in the US after the dotcom correction was also comparatively swift. The level of real GDP sustainably surpassed its pre-correction peak within five quarters, compared to 10 quarters following the GFC. Three years on from the pre-dotcom correction peak, US real GDP had risen by a net 7%. Three years on from the pre-GFC peak, US real GDP was only 0.5% higher.

3. US investment contracted, but private consumption was remarkably resilient.   

Strength in investment was a key element of the US growth boom from the mid-1990s. Investment in real terms expanded at a high single-digit or double-digit year-over-year rate for most of the period from mid-1996 to early 2000. It then fell off a cliff: An 8% year-over-year real growth rate in the first quarter of 2000 became a contraction by the third quarter of 2001. 

However, following the GFC, the contraction in US investment from peak to trough exceeded 16%. While it took 11 quarters for US real investment to exceed its pre-dotcom correction peak, after the GFC it took more than twice as long—23 quarters. 

In contrast, annual real growth in US private consumption averaged 2.5% in 2001–02. Although this was only around half the 5%-plus annual growth rates from 1998 to 2000, private consumption was nonetheless remarkably resilient in the wake of the dotcom correction. In 2008–09, private consumption contracted by an average of 0.6% annually in real terms.

Monetary and fiscal policy stimulus played a key role in supporting the US economy in the early 2000s. The increase in the US unemployment rate was relatively small following the dotcom correction as a result, peaking at a little over 6%. Following the GFC, it rose by over 5 percentage points, to peak at 10%.

4. The drop in US exports was similar to the post-GFC decline, although the recovery was much weaker, reflecting divergent dollar trends.

Over just four quarters, the level of US exports dropped by 12% in real terms from peak to trough. US imports also tumbled, though less precipitously, falling by around 7% in real terms from peak to trough.

Unlike the other expenditure components of GDP, the scale and speed of the drop in US exports following the dotcom correction were broadly similar to those during the GFC. In the latter, real US exports dropped by around 14% from peak to trough.

The subsequent recoveries in US exports over the two periods were very different, though not in the expected way. In the early 2000s, it took 13 quarters for the level of US real exports to exceed its pre-dotcom correction peak. Following the GFC, it took nine quarters.   

Differences in currency dynamics likely contributed to the divergent export recoveries. From the mid-1990s, and beyond the dotcom correction, the US dollar was on a strong appreciating trend.    

In contrast, in the run-up to the GFC, as the problems in US subprime lending became increasingly apparent, the US dollar was on a sharply depreciating trend. The dollar was particularly weak against the euro during this period, with the eurozone initially perceived — incorrectly, ultimately — to be less susceptible to a crisis.  

5. Substantial US monetary and fiscal policy responses helped to limit the economic damage, assisted by moderating inflation.

A retightening of US monetary policy in 1999–2000 contributed to the dotcom correction. The Fed funds rate peaked at close to 7% in mid-2000, followed by rapid easing. By the end of 2002, the Fed funds rate was sub-2%, and by mid-2003 it was close to 1%, a reduction from peak to trough of around 600 basis points.   

A moderation in US inflation at the time, helped by low crude oil prices, was helpful. Fiscal policy stimulus was also substantial. In 2000, the US ran a primary budget surplus (i.e., excluding interest payments) of well over 5% of GDP, with an overall budget surplus of around 2% of GDP. Over the subsequent three years, there was an easing in the US primary balance of around 7 percentage points of GDP. By 2003, the US’ primary deficit was 1.6% of GDP.  

6. Global real GDP growth slowed sharply in 2001, but the subsequent recovery was swift.

Annual global real GDP growth averaged 3.4% from 1996 to 2000, peaking at 4.3% in the latter. In 2001, it slowed to 1.7%. Exports and investment were the key drivers of the downturn.   

Still, between the first and third quarters of 2001, when the US economy was at its weakest, quarter-over-quarter global real GDP growth remained positive, averaging 0.2%. While global real GDP including the US contracted quarter over quarter in the third quarter of 2001, it was only a marginal fall (negative 0.1%), and growth rates in the subsequent quarters were robust. 

9. The economic impact varied considerably across major economies and regions.  

Annual real GDP growth rates lost considerable momentum across almost all regions in 2021, with Sub-Saharan Africa the only exception. Weakness in exports and investment were generally the key drivers, although there were significant differences in the scale, duration and breadth of the economic weakness.

Key points to highlight include:

• Latin America suffered a relatively deep recession, amid broad-based weakness across all expenditure components. 
• The Middle East and North Africa also went into recession. Real GDP declined in three of the four quarters of 2001, driven primarily by weakness in exports. However, the fall from peak to trough was comparatively small, at around 0.5%, and followed by a rapid recovery.
• Emerging Europe avoided a recession, as did Asia-Pacific, although both regions experienced short-lived periods of export- and investment-driven weakness in mid-2001.

10. Growth in Western Europe slowed markedly, although there were significant national variations, including UK outperformance. 

Annual real GDP growth in the eurozone averaged 2.8% from 1996 to 2000, peaking at 4.1% in the latter. In 2021, annual growth slowed to 2.2%, again misleadingly strong due to favorable carryover effects.

The sharp slowdown in the eurozone aggregate masked divergence among its member states. In the late 1990s and early 2000s, some of the entrants to the newly formed eurozone were benefiting from a windfall in the form of convergence of interest rates to historically low levels. Germany, in contrast, was struggling, amid poor competitiveness and high unemployment prior to the reforms introduced in the mid-2000s. 

The UK’s relatively impressive growth performance over this period stands out. Although its annual real GDP growth rate slowed to 2.4% in 2001, down from a 3.7% average in 1999 and 2000, real GDP still rose by 0.5% quarter over quarter on average in 2001 and then picked up in 2002. As in the US, fiscal policy was a key support. In 2000, the UK’s gross debt-to-GDP ratio was just 30%, compared with around 100% currently.

While there were variations regionally, nationally and across expenditure types, the real GDP losses following the dotcom correction were in most cases relatively moderate and short-lived in comparison with the GFC.

The much larger output losses and more persistent economic weakness following the latter reflects various differences. In brief, these include the larger exposure of US households to the weakness in real estate and the more widespread spillovers to the banking sector, credit provision, etc. The dotcom correction did not put the stability of the global financial system, and major financial institutions in the US and elsewhere, at risk.

The bottom line? The likelihood of a correction in equity prices of a similar magnitude to those in the early 2000s is less elevated currently. At the same time, however, the ability of policymakers to respond effectively if a crash were to occur is likely to be constrained, indicative of a significant risk of larger and more widespread output losses compared with the early 2000s.

This article was published by S&P Global Market Intelligence and not by S&P Global Ratings, which is a separately managed division of S&P Global.