Hello, YieldAlley readers! In this issue:

🚘 What do Disney+ and your local car wash have in common? (Sherwood)

👑 Gold's Role Reconsidered (Larry Swedroe)

📈 The eccentric investment strategy that beats the rest (The Economist)

🇺🇸 Long-term treasuries in diversified portfolios (Meketa)

🥇 Can the Gold Rush Continue? (Morningstar)

The stock market keeps hitting new highs, technology companies dominate the global economy, and everyone seems to be talking about artificial intelligence. Sound familiar? It should, it's reminiscent of past moments in history when markets soared on the promise of transformative technology, only to eventually crash back to earth.

But not every rally is a bubble, and not every bubble bursts the same way. Understanding what actually makes a bubble can help you make sense of today's markets without falling into either complacency or panic.

Financial bubbles aren't just about high prices. Markets can stay expensive for perfectly rational reasons, strong earnings growth, low interest rates, or genuine innovation. A true bubble requires three key ingredients working together:

Rapidly rising asset prices. This is the obvious part, valuations climbing faster than historical norms, often accelerating as the bubble matures.

Extreme valuations disconnected from fundamentals. Prices rise not because companies are earning more, but because investors expect them to earn dramatically more in the future. The aggregate value of companies exceeds what they could realistically generate in cash flows.

Significant systemic risk from leverage. Debt amplifies both gains and losses. When highly leveraged investments start failing, they create cascading problems throughout the financial system.

The crucial element is that these factors must combine. High prices alone don't make a bubble. The tech sector has posted impressive returns recently, but that doesn't automatically signal trouble, it depends on what's driving those returns and how they're being financed.

History shows that bubbles typically unfold in predictable stages, usually centered around a genuinely transformative technology or innovation.

The excitement starts with something real. Canals revolutionized transportation in the 1790s. Railways opened up continents in the 1840s. The internet fundamentally changed communication in the 1990s. Each innovation promised, and eventually delivered, tremendous economic value.

The problem comes in the middle chapters. Investors struggle with three impossible questions: How big will the market become? How quickly will it scale? And who will be the winners?

Unable to answer these questions, investors do something rational that creates irrational outcomes: they buy options on the future by investing in many companies, hoping to catch the eventual winners. As excitement builds, more investors pile in, and the aggregate value of all these "options" balloons beyond what the actual future market could support.

In the canal bubble of the 1790s, investors expected 50% returns. Within a decade, returns had fallen to 5%, and a quarter-century later, only one in four canals still paid dividends. Yet canals themselves were crucial infrastructure that reshaped the economy, the innovation was real even though most investments failed.

The same pattern repeated with railways in both the UK (1840s) and US (1870s), where stocks collapsed 85% after their peaks despite the transformative impact of rail transportation. Radio stocks in the 1920s, computer companies in the 1980s, and internet stocks in the late 1990s all followed similar trajectories.

So where does the current situation fit in this historical pattern? The answer is complicated, and different metrics tell different stories.

What looks concerning:

Market concentration has reached extreme levels. The five largest US tech companies are collectively worth more than the entire stock markets of the UK, Japan, India, Canada, and the Eurostoxx 50 combined. These five companies alone represent about 16% of the entire global public equity market. The top 10 US stocks account for nearly 25% of global equities.

Valuations are elevated. US market price-to-earnings ratios are above their 20-year highs. Companies are rapidly increasing capital expenditures on AI infrastructure, spending that has accelerated dramatically since ChatGPT's emergence. IPO activity has picked up, with first-day premiums averaging 30%, the highest since the late 1990s tech bubble.

There's also the emergence of vendor financing, where companies help customers finance their purchases, which has historically been a warning sign in other bubbles.

What looks different from past bubbles:

The price appreciation has been driven by actual earnings growth, not just speculation. The gap between tech sector earnings and the rest of the market has widened dramatically since the financial crisis. Unlike the late 1990s, when valuations soared ahead of profits, today's dominant companies have delivered extraordinary fundamental performance.

Current valuations are high but not at bubble extremes. The median forward price-to-earnings ratio for the "Magnificent 7" tech companies is roughly half what the biggest companies commanded at the peak of the dot-com bubble. The price-to-earnings-growth (PEG) ratio remains well below late-1990s levels.

Balance sheets are remarkably strong. Today's tech giants are financing most of their capital expenditures from free cash flow rather than debt. During the telecom bubble of the late 1990s, companies took on massive debt to build infrastructure, which amplified the damage when the bubble burst. Several major telecoms held $25-100 billion in debt when they collapsed. Today's situation looks very different, these companies are profitable with strong cash positions.

The current leaders are established winners, not speculative upstarts. Most bubbles form during periods of intense competition as hundreds of new entrants fight for market share. Think of the 600 radio stations that opened between 1920-1922, or the numerous PC manufacturers in the 1980s, or the hundreds of internet companies in 1999. Today's AI investment is dominated by a handful of existing tech giants rather than a flood of new competitors.

The most significant risk isn't necessarily a bubble bursting, it's that these companies might not earn adequate returns on their massive capital investments. If the billions being poured into AI infrastructure don't generate proportional profits, stock prices could correct significantly even without a full bubble collapse.

There's also the historical precedent of technological displacement to consider. Of the Fortune 500 companies in 1955, only about 10% remain on the list today. None of the 10 largest S&P 500 companies in 1985 were still in the top 10 by 2020, and only one from the 2000 top 10 remained by 2020. Dominance tends to be temporary in the technology sector.

However, and this is crucial, even if today's leaders eventually lose their positions, that could happen through gradual competitive displacement rather than a sudden crash. IBM dominated mainframe computing but lost its lead to Microsoft in software, which later lost ground to Apple in consumer devices, which now competes with newer platforms. This succession happened over decades, not in a single catastrophic collapse.

The honest answer is that we won't know if we're in a bubble until after the fact. Markets can remain expensive for extended periods when supported by strong fundamentals. They can also correct sharply when sentiment shifts, regardless of underlying business strength.

The key distinction is between company concentration and systemic vulnerability. High concentration means that if these stocks stumble, broad market indices will fall significantly, they're simply too large not to drag the market with them. But systemic vulnerability requires widespread leverage and interconnected risks that can cascade through the economy.

Right now, concentration is extreme but systemic leverage appears limited. If tech valuations contract, it would likely hurt portfolios without triggering a broader economic crisis. Strong bank balance sheets and relatively low household debt levels provide buffers that didn't exist in past bubble episodes.

The implication for portfolio construction is straightforward even if the market outlook isn't: diversification matters more as concentration increases. That doesn't mean avoiding technology or betting against the leaders, it means not putting all your eggs in a basket that represents an unprecedented share of global markets.

Regional diversification has already started paying off, with international markets performing surprisingly well this year. Different sectors like energy, capital goods, and infrastructure could benefit from the physical demands of AI technology, data centers need electricity, which requires real-world investment in generation and distribution.

Bubbles are easier to identify in hindsight than in real-time. The current market shows some characteristics of past bubbles, high concentration, elevated valuations, surging investment, but lacks others, particularly the extreme leverage and speculative frenzy that amplified previous crashes.

What we can say with confidence is that valuations are stretched and concentration is historically high. Whether this represents the early stages of a bubble or simply the natural result of a few companies legitimately dominating the most important technology of our era remains an open question.

The safest assumption is probably somewhere between the extremes: these aren't the tulips of 1630s Holland, but neither are current valuations without risk. The technology is real, the companies are profitable, and the balance sheets are strong, but trees don't grow to the sky, and nothing stays dominant forever.

Watch what's driving returns. If prices keep rising on earnings growth and fundamental strength, that's different from prices rising purely on expanding valuations and future promises. And remember that even if this isn't a bubble about to burst, that doesn't make it risk-free. Markets can correct for many reasons, and concentrated positions amplify that risk regardless of the underlying cause.

Government Money Market Funds (7-Day Yields)

Brokered CD Rates (6-Month Rate)

ETFs (30-Day Yields)

American Airlines is offering AAdvantage members exclusive early access to FIFA World Cup 26 match tickets through miles redemption starting October 13, 2025, creating a tiered priority system that rewards elite status holders with first access before opening to all members. The redemption structure prices experiences (2 tickets per redemption) from 75,000 miles for group stage matches to 965,000 miles for the final, effectively converting loyalty currency into access for the world's largest sporting event across matches in the United States, Canada, and Mexico.

This offer transforms AAdvantage miles into FIFA World Cup access during the critical booking window before general public sales, with the 75,000-mile entry point for group stage matches representing compelling value when secondary market prices for World Cup tickets historically reach $500-2,000+ per seat, effectively valuing miles at 1.3-2.7 cents each, significantly above typical economy redemption rates. The tiered October 13-15 access windows create meaningful advantages for elite status holders who can secure premium matchups and seating locations before inventory depletes, while the two-ticket minimum redemption eliminates solo traveler flexibility but aligns with typical group attendance patterns. Elite members should prioritize October 13 access for marquee matchups (opening ceremony, semifinals, finals, popular team group stages) where seat selection creates compounding advantages, while general members face diminished inventory on October 15 but may find value in less-demanded matches. The parallel daily sweepstakes offers zero-cost lottery exposure for members unwilling to commit miles, though this potentially represents the highest-value AAdvantage redemption opportunity in 2025-2026 for sports enthusiasts with flexible mile portfolios, particularly when compared to standard flight redemptions hovering around 1.0-1.5 cents per mile.