Set up for euphoria?

As equity markets surge to new highs, particularly in AI and tech sectors, investors are increasingly asking the question: are we in the early stages of a bubble? To answer this, we first turn to two foundational frameworks by the academics Kindleberger and Minsky.

Bubble causes

Kindleberger wrote the original book on bubbles. His model outlines a five-stage progression: displacement, boom, euphoria, crisis and revulsion. He notes it often begins with disruptive innovation or a policy shift that creates new opportunities. As optimism builds, credit expands and asset prices tend to rise. Eventually, speculation overtakes fundamentals, leading to euphoric buying and inflated valuations. A trigger – often minor – can then spark panic, causing prices to collapse and confidence to evaporate.

Minsky’s hypothesis complements this view by focusing on borrower behaviour. As market stability persists, risk-taking tends to increase and the financial system becomes increasingly fragile,  until it inevitably breaks – the so-called ‘Minsky moment’. The basis of that model is that stability creates instability.

The internet bubble of the late 90s follows Kindleberger’s five-stage arc quite well: the displacement came from the rise of internet technologies, which catalysed a boom in venture capital and IPOs, as well as a belief that “this time was different”. Euphoria set in as valuations soared, often detached from earnings or business models.

I remember stocks being valued as “price per click”. The ensuing crisis was triggered by a series of earnings misses and accounting scandals, leading to a revulsion that wiped out trillions in market value. In the aftermath, central banks lowered interest rates and injected liquidity to stabilise markets – ironically sowing the seeds of the next crisis.

Minsky’s insight that “stability creates instability” played out vividly in the lead-up to 2008. The relative calm and recovery post-2000 encouraged excessive risk-taking, particularly in housing finance. Deregulation of the financial sector, led to a wave of financial engineering. Remember MBS, CDS, CDOs and CDO squared?

Borrowers migrated from hedge positions to speculative positions and Ponzi arrangement, enabled by lax lending standards and complex securitisation. The illusion of stability masked growing fragility, culminating in a Minsky moment when subprime defaults triggered a cascade of losses across the financial system.

Bubble history

While the internet bubble was largely equity-driven and centred on innovation, whereas the 2008 crisis was credit-driven and systemic, both followed the same psychological and structural dynamics outlined by Kindleberger and Minsky. Together, they illustrate how financial history often rhymes, with each cycle building on the unintended consequences of how we survived the previous one.

Today’s market dynamics appear to echo some of these patterns. The concentration of gains in a few mega-cap firms, the speculative fervour around AI, and elevated valuation metrics suggest that we may be transitioning from ‘boom’ to ‘euphoria’. While fundamentals remain strong in some areas, the psychological and financial conditions underpinning bubbles are becoming more visible in our view.

Bubble monitoring

As a self-confessed bubble aficionado, I’ve written extensively about the topic. Our proprietary ‘L&G Bubble Index’, developed from over three decades of research, draws on more than 35 indicators to assess bubble risks from a variety of angles, incorporating the models and learnings from, among others, Kindleberger and Minsky.

History suggests that a stock market bubble typically emerges when asset prices surge far beyond their intrinsic value, fuelled by speculative euphoria and easy access to credit. At present, our Bubble Index signals heightened risk, though it stops short of confirming that we’re in a full-blown bubble. As Robert Armstrong aptly observes in his recent FT opinion piece, “we have a pretty good set-up for one”.

Our index is anchored by five key components, each capturing a distinct facet of bubble formation:

• Extreme valuations
• Macroeconomic conditions and disruptive change
• Speculative behaviour and overconfidence
• Credit expansion
• Patterns in asset price performance

Bubble trouble?

Currently, macro conditions are the most influential driver of the index. We have seen innovation that is powerful enough to holster beliefs that the world is about to change significantly in terms of growth, productivity and earnings.

This plays out against a backdrop of one of the longest economic growth cycles on record, creating stability in economy and markets that have only been materially interrupted by something as exceptional as COVID. Finally, central banks seem poised to intervene at the first signs of market fragility.

The importance of extreme valuation and behavioural indicators is rising as well, though both components are not yet at historic extreme levels in our view. For this to get to the next level, we are watching the performance of nonprofitable tech companies (signing that valuations become irrelevant) and the emergence of fantasy valuation measures.

Interestingly, credit expansion and market-based factors remain relatively subdued. For instance, corporate and banking sector leverage remains well below the peaks observed in previous bubbles.  

The steady but gradual rise in the index merits close monitoring as a potential risk and a potential opportunity. Bubbles only become perilous when they burst; until then, they have the potential to deliver impressive returns. Should a bubble take shape in the coming years, we believe the technology sector will be the driver.

At the moment, investors are focusing on the build of AI infrastructure, with a high demand for computer and data centres. Capex is extremely strong but demand is still outpacing supply as it takes time to build out capacity. At some point we believe this trend will broaden to a widespread adoption of AI across industries. If so, valuations of the broad market could have further to rise.

We are currently overweight AI technology in our equity portfolios, but our overall risk position is slightly cautious. This aims to deliver a careful balance between downside risks – driven by valuations and the short-term growth / inflation outlook – and the longer-term upside scenario driven by widespread AI investment adoption that unleashes productivity gains and cost savings.”