The cost of believing it’s different this time

Notes on technological change, investor enthusiasm, and the old rules that still matter. 

Every cycle produces its own rationale for why things are different this time. Several years ago, financial engineering created the illusion of safety in subprime loans. During the dotcom bubble, investors believed that traditional net income and cash flow metrics were outdated for internet startups. We are beginning to hear a “it’s different this time” narrative emerge around AI innovation.

With technological disruption and innovation come stories arguing that the old rules no longer apply. When the narrative takes hold, valuation discipline can feel outdated or like a failure of imagination. Behavioral biases then do the rest: recency bias extrapolates the latest gains far into the future, confirmation bias filters out evidence that challenges the prevailing narrative, and herding makes participation feel safer precisely as prices may be moving further from fundamentals. In that environment, the narrative refrain “it’s different this time” is not an analytical conclusion but a shortcut explaining away any and all dissenting facts or evidence. The risk to investors is that it confuses technological promise with investment merit.

Source: Loomis Sayles, Morningstar. As of 6/30/2026.

Much of the current excitement centers on the AI-driven data center buildout, which has created a temporary supply crunch in memory and storage chips. For chipmakers, the result has been windfall pricing, margins, and profits well above historical norms. The market has rewarded many of these lower-quality, highly cyclical companies as if this pricing power is permanent, not simply reflecting a favorable moment in the cycle. What we see is a supply-demand imbalance. Imbalances like this tend to contain the seeds of their own correction. 

Higher prices invite new capacity as well as innovation. As supply catches up to demand, pricing power fades, and margins normalize. And, if hyperscalers happen to pause AI capital spending as supply increases, these chip stocks could face an even sharper cyclical downturn.

Near-term market enthusiasm should not be confused with durable business quality. Navigating markets means distinguishing between: 

• price and value;
• volatility and impairment;
• and temporary popularity and sustainable business quality.

Most memory and storage chips are undifferentiated, drop-in replacements for one another. Because this makes switching costs low for buyers, most chipmakers do not have structural pricing power. We believe stocks of these chipmakers are riding a wave of temporary pricing power that will inevitably crest and break. We don’t believe this chip cycle is different.

Not all margin expansion is created equal.

In our view, elevated margins are structural and durable only when they’re supported by ongoing innovation that continues to improve the product in comparison to peers and earns a company an increasing share of industry profits. Look at Nvidia. Its Rubin Vera architecture coming to market in the second half of this year commands a price nearly double (2x) its prior-generation Blackwell stack by delivering roughly a tenfold (10x) improvement in compute power and energy efficiency. By delivering increased value to its customers, Nvidia earns its pricing power, reinforcing its competitive advantage and its ability to generate cash flow. That is a company widening its moat, not just riding a cyclical wave.

Nvidia’s advantage did not appear overnight. It reflects decades of focused investment, accumulated know-how, and a software ecosystem that has attracted millions of developers. As AI adoption broadens, we believe its foundation positions Nvidia to continue to benefit from the long-term shift toward accelerated computing.

What this can mean for you

Markets periodically convince themselves that technological disruption means the old rules of price and value no longer apply. Investors are right to recognize the scale of the disruption, but wrong to assume it changes the rules. Even as AI reshapes parts of the economy, the nature of successful companies remains the same: durable competitive advantages, pricing power, and sustainable profitable growth.

None of this is a prediction about when market rationality will return to the chip industry. It’s a statement about process. A portfolio built with valuation sensitivity, downside risk awareness, and patience may lag when markets are paying any price for the current favorite. 

Our long-term outperformance has never come from getting every quarter right. We haven’t, and we won’t. It comes from never abandoning our process simply because the market disagreed with us for a while. Right now, the market disagrees with us. Loudly. Periods of underperformance can be uncomfortable, even when they are an inevitable part of active investing.

We are not asking you to ignore the discomfort of the moment. We are asking you to understand it in the context of a process designed to distinguish temporary enthusiasm from durable value, and to recognize that long-term results come from adhering to that process – not by abandoning the principles that shaped the portfolio in the first place.

Until next month,

Loomis Sayles Growth Equity Strategies Team