According to Zhitong Finance APP, Societe Generale strategist Albert Edwards, long known for his extremely bearish market views and who even calls himself a "super bear," has compared the current artificial intelligence (AI) investment boom to the economic bubble that preceded the 1997 Asian financial crisis. He warned that massive capital expenditure may be masking weak productivity growth, raising doubts about whether the AI boom can deliver on investors' lofty expectations.
In a research report published on October 8, Edwards pointed out that U.S. total factor productivity is slowing down. Total factor productivity measures how efficiently an economy uses labor and capital. The weakening of this indicator stands in sharp contrast to the surge in AI infrastructure spending and the market optimism that has driven the S&P 500 to record highs.
Edwards said this divergence reminds him of the period of East Asian economic expansion in the 1990s, when rapid investment and strong growth fueled widespread optimism, only for the region to later suffer a financial collapse. The strategist cited research by Apollo Global Management chief economist Torsten Slok, noting that despite the raging AI investment boom, U.S. total factor productivity growth is declining. This indicator is designed to strip out growth driven purely by adding labor and equipment and to measure improvements in economic efficiency on their own.
The Warning From the Asian Financial Crisis
Edwards reviewed how, before the 1997 crisis erupted, economists and policymakers had praised Asia's rapid economic development. The World Bank published a report titled "The East Asian Miracle" in 1993, followed by a study in August 1996 on Thailand's economic success, yet less than a year later Thailand fell into financial collapse. Economist Paul Krugman challenged the prevailing optimism of the time in his 1994 article "The Myth of Asia's Miracle." He argued that most of the region's economic growth reflected increases in investment and labor rather than substantial productivity gains.
Edwards said he reached a similar conclusion at the time, warning that cheap and abundant capital was being allocated inefficiently. The subsequent financial crisis exposed vulnerabilities that had been masked by rapid economic expansion. The strategist also recalled his own experience of comparing the late-1990s U.S. tech bubble with Thailand's earlier investment boom. He believes that confidence in AI today may be overlooking similar warning signs.
AI Spending Fails to Boost Net Investment
Edwards highlighted the divergence between gross and net corporate investment in the United States. As companies spend heavily on AI-related equipment and infrastructure, gross investment has increased. However, corporate net investment as a share of GDP, after subtracting depreciation, has remained broadly flat. This distinction is crucial because higher input does not necessarily translate into an equivalent increase in productive capital. Companies must constantly replace aging equipment, especially in the technology sector where assets can quickly become obsolete.
Edwards cited observations by economist Rob Parenteau, noting that growth in gross investment has not brought a corresponding improvement in net investment. The report noted that in recent years real gross corporate investment as a share of GDP has climbed, while net investment has been largely flat. The strategist also questioned the composition of corporate spending. Equipment investment has accelerated in nominal terms, but after adjusting for inflation, the increase is not significant. Meanwhile, despite the continued expansion of data centers, nonresidential building investment has contracted. Edwards pointed out that corporate fixed investment's contribution to annual GDP growth remains within historically normal ranges.
The Productivity Dividend Remains Uncertain
Edwards acknowledged that AI may ultimately deliver the productivity gains investors are hoping for. The technology is still being deployed across industries, and its economic benefits may take time to show up in official statistics. Slok also said it may be too early to expect AI investment to produce measurable improvements in total factor productivity. Still, Edwards believes the reality of weak productivity growth is enough to warrant skepticism about the scale of the investment boom.
The analogy with the Asian financial crisis does not mean AI investment will necessarily end in a similar collapse. But Edwards' analysis raises the possibility that spending driven by expectations of future technological breakthroughs may exceed the economic returns that are ultimately realized.
The Bond Market Poses Another Risk
Edwards also warned that the global bond bear market remains a major concern for investors, and government borrowing costs may face a new round of pressure. He noted that bond investors have recently focused on France, but as they assess fiscal weaknesses in major economies, attention could shift to Japan, the UK or the US. The strategist believes that as investors test governments' ability to sustain borrowing, bond market pressure could move quickly from country to country.
For stock investors, the broader concern is that enthusiasm for AI has supported record share prices even though evidence of a sustained productivity boom remains limited. Edwards' analysis suggests that whether the AI-driven market rally can continue ultimately depends on whether the technology can deliver measurable gains in economic efficiency, rather than merely setting off another wave of capital expenditure.
Comments