Easy Bubbles. Hard 100 Baggers. Useless AI | 6 Things We Learned This Week

The 'Easy Bubble' Concept

Ben Inker defines an 'easy bubble' as one like the 2000 internet bubble, where investors could avoid pain by shifting from overvalued large-cap growth to other risk assets like small caps, REITs, and emerging markets while maintaining a normal portfolio structure.

Hard Bubbles Force Extreme Positioning

In contrast, the 2008 financial crisis was a 'hard bubble' where nearly all risk assets were overvalued, forcing investors into extreme portfolios that made no sense in normal times and risked massive underperformance if the scenario was wrong.

Reading the Risk-Return Line

The slope of the risk-return line indicates bubble difficulty; 2008 featured a downward-sloping line requiring counterintuitive positioning, while easier bubbles allow investors to find positive expected returns outside the bubble assets.

Don't Let Labels Do the Work

Chris Mayer warns against relying on labels like 'SpaceX' or 'AI,' stressing that investors must dissect companies into distinct segments—such as space launch, Starlink, and data centers—to assess specific competitive positions and capital requirements.

The Vague 'AI' Moniker

The term 'AI' currently acts as a catch-all covering vastly different technologies, from large language models to trivial software features, meaning investors must parse whether companies are driving genuine productivity or merely attaching buzzwords for marketing.

The Useless AI Phenomenon

Early evidence suggests many AI integrations are value-destructive gimmicks, with expert network sources citing enterprise customers calling new AI features a 'waste of time' that add no functional value to existing software subscriptions.

Bubbles Without High Valuations

Bubbles can form through inflated earnings rather than just high multiples, as seen in 2007-08 Europe and 2012 Emerging Markets where index earnings doubled then stagnated for over a decade, trapping investors who relied on P/E ratios alone.

Investment Spending Mechanics

Massive capital investments—like the current data center buildout—boost reported earnings mechanically because the spending becomes revenue for vendors immediately while depreciation expenses are delayed until multi-year construction projects are completed.

Temporary Earnings Peaks

With depreciation schedules not yet started on rapidly ramping investments, current corporate profits may be unsustainably high, creating an 'earnings bubble' even when headline valuations on stocks like Microsoft appear reasonable relative to current earnings.