What learnings from the last tech-run can you apply to today's market?

When I think back to that period in the late 1990s, what stands out isn’t just the momentum. Every cycle feels different when you’re living through it, but if you step back, you begin to see familiar fault lines. And that’s where the real lessons are.

1. Narratives change, but the risks often rhyme

In 1999 it was the internet. Today it is AI. The language sounds different, but the common risk patterns are familiar: aggressive accounting, off balance sheet financing, circular capital flows, and vendors effectively financing customers.

There could be real long-term winners. But periods of froth inevitably pull in speculative companies that may not survive when expectations reset. The lesson is not to dismiss the technology. It is to separate durable business models from narrative momentum.

2. Investors tend to underestimate secondary and tertiary impacts

Near cycle peaks, markets often do a reasonable job identifying the primary names at risk. Where they struggle is managing portfolios for what happens next: funding stress, correlation spikes, and pressure on adjacent sectors. Historically, these second order effects have often been more painful than the initial drawdown itself.

During the Global Financial Crisis, many asset managers identified subprime risk early and reduced direct exposure. That was not the failure. The failure was underestimating the domino effect. Once subprime cracked, it triggered funding stress, correlation spikes, and a synchronized global equity decline. Assets assumed to be diversified suddenly moved together, exposing vulnerabilities in businesses tied to market guarantees, including products such as Guaranteed Minimum Withdrawal Benefit.

Today, AI enthusiasm is concentrated in services such as professional technology and finance, where early adoption, task level speedups, and inexpensive no-code tools dominate the conversation. That is the noisy first order battlefield. The more interesting opportunities may lie in the second and third order effects. Over time, productivity gains often accrue most powerfully to capital intensive industries through smart factories, predictive maintenance, robotics, autonomous systems, and supply chain optimization. If market rotation continues, cyclical and capital heavy sectors may become capital beneficiaries, lowering their cost of capital and increasing their index weight.

3. Traditional diversification and conventional risk monitoring could fail to capture this risk

Many portfolios appear diversified by number of holdings or sectors. But that framework may not necessarily account for shared drivers and knock on effects. When a dominant theme unwinds, assets that look unrelated can move together.

Diversification by labels is not enough. Diversification by underlying exposures and drivers is generally what matters most. Focus less on how many names or sectors you own, and more on what truly drives returns and risk. Stress test portfolios for what happens beyond the first headline. That is where real diversification, and real risk management, show up.