Category Bubbles

3-Years Since ChatGPT Launched… What’s Changed? 

The seismic shift triggered by ChatGPT 3 years ago reminds me of 1995 when Netscape hit our screen. But as we approach the year 2000 - several "greey swans" emerged. Could 2026 be similar. This post discusses some of the possible risks looming for next year. This AI revolution has many of the hallmarks we saw some 30 years ago; i.e., creating extreme capital concentration in giants like Nvidia. As we enter what I think is a late-cycle phase, our focus shifts to systemic risks—from AI disillusionment to credit volatility.

Nvidia’s (Expected) Beat Was Not Enough… 

Nvidia's earnings beat—reporting 62% YoY revenue growth and projecting $65B in Q4—solidified its position as the AI chip leader. Yet, the question for investors is not if the AI boom is real, it's how much you pay? The massive AI capital expenditure by giants like Google and Meta is transforming them from "asset light" platforms to "asset heavy" infrastructure owners. This shift, coupled with the unanswered question of whether the returns on this spending will justify the massive increase in assets, is fueling investor unease and market momentum loss

Market Correction Chorus Grows

Goldman Sachs are warning of a 10-20% correction within the next 12-24 months. And whilst saying this would be a healthy outcome - it aligns with stretched valuations seen only during the dot-com bubble, according to the Shiller CAPE Ratio. The market's risk is concentrated: returns are currently driven by a handful of mega-cap tech stocks. As Michael Burry's short of Palantir highlights, the issue isn't business quality, but the extended prices being paid. From mine, better opportunities exist outside the Mag 7.

Why You Should Avoid Paying Too Much

It’s very tempting to chase AI and "Mag 7" gains, but your long-term returns are ultimately determined by the price you pay. With the S&P 500 trading near 25x forward earnings and the Shiller CAPE ratio flashing warnings similar to the 2000 dot-com bubble, the market is lofty territory. History is clear: investing at such elevated valuations drastically lowers subsequent 5 and 25-year returns. While FOMO is powerful, be cautious. As a long-term investor, focus on the risk of what you could lose, not just what you might miss

How Far Will Multiple Expansion Take Us?

Earnings per share growth has averaged 8.3% pa from 2015 through today. However, capital appreciation in stocks (exclduing dividends) has seen a CAGR of 11.0% over the same 10-year period. This divergence is widening which indicates multiple expansion. From mine, investors should be braced for mean reversion.

Will Investors be Emboldened by Fed Easing?

Are stocks headed for a melt-up with the Fed set to ease rates over the next 12+ months? It could seem that way as stocks continue to print new highs as the 'soft landing' script firms. And whilst there might be further upside - the environment echoes a lot of what we experienced from the mid 1990's. For example, at the time we had expanding growth, low inflation with aggressive easing from the Fed. What's more, investors were very bullish on the promise of the internet - set to deliver powerful productivity gains. Stock multiples continued to expand as the S&P 500 delivered strong double-digit gains not seen in decades. Today conditions feel similar.

S&P 500 +10.1% for Q1 – Can it Continue?

If you asked me at the end of December whether I thought the S&P 500 would be up ~10% at the end of the first quarter this year - I would have said "unlikely". And yet here we are. With the promise of (coming) interest rate cuts and continued strong economic growth (implying growth in earnings) - US equities have arguably exceeded most analysts full year targets. For we have already exceeded all but 1 of 18 full year S&P500 forecasts "experts" made at the beginning of the year.

A Different Lens on the ‘AI Bubble’

25 years ago Cisco (CSCO) was the largest company on the S&P 500 by market cap. Its shares soared on the demand for networking equipment. But it didn't last. The stock lost 89% of its value in two years. Nvidia is not only charting a very similar technical pattern to CSCO - there are also similarities with valuation metrics. Both the price-to-earnings ratio and price-to-sales multiples have been very similar. What we don't know (or cannot know) is whether the same fate lies ahead for NVDA (as investors pay a staggering 35x sales for a slice of the AI pie)

Are Semi’s Set to Cool their Gen-AI Heels?

Whilst the technology sector is outperforming the benchmark index this year -- semiconductor stocks have done the bulk of the heavy lifting. And it's not difficult to explain investor FOMO. It's entirely due to the hype around "AI" and specifically something called "Generative AI". For example, in a report by Grand View Research, they valued Gen-AI at ~$13B last year. However, its anticipated CAGR is estimated to be ~36% - which puts the industry hitting $109B by 2030. That's a sharp ramp higher from basically zero two years ago. And today - there a very few chipmakers who produce the GPUs required to meet the insatiable demand. However, is the demand semis are seeing today (and revenue) sustainable long-term? That's unlikely.

Something Doesn’t Add Up… 

It's Nvidia's world and we're living in it (if you believe the stock market). The S&P 500 (and Nvidia) recorded all new highs post the AI chip maker's earnings. Be careful paying too much. The rapid rise in Nvidia's market cap has only seen the market narrow further. And from mine, that makes it more subject to both volatility and risk. Deutsche Bank’s Jim Reid dimensioned the risk another way. He shows how the Top 10% of stocks in the S&P 500 constitute ~75% of the total capitalization. We have not seen that since 1929! The only other time we saw something similar was the dot.com bubble...