What the Panic of 1873 Can Teach Us About the AI Boom
One of the clearest lessons from the Panic of 1873 is that debt can help build the boom, then end it. That is worth keeping in mind as capital pours into AI at a remarkable pace, with companies racing for dominance and investors happy to fund the effort. The important question is not whether debt will matter. It will. The question is where we are in that cycle, and whether this is still the build phase or the point where leverage starts to do damage.
The Railroad Parallel
After the Civil War, America entered a period of enormous optimism and expansion. Railroads were the defining technology of the era. They connected distant cities, cut shipping costs, opened new markets, and pushed economic activity westward. This was not some frivolous sideshow. It was the buildout of essential national infrastructure.
That promise drew in a great deal of capital. Railroad companies borrowed heavily. Investment banks packaged and sold railroad bonds to investors in the United States and Europe. As long as money remained available and confidence stayed high, the system worked. The problem was that many of these projects were built on the assumption that financing would stay easy and demand would show up right on time. Markets have always been very good at assuming the best right before they stop doing that.
Europe helped turn a domestic excess into a broader panic. In the spring of 1873, the Vienna stock market cracked after its own speculative boom, and that shook confidence across the continent. European investors, already getting more cautious, began pulling back from American securities, especially railroad bonds. That mattered because the U.S. railroad buildout had become dependent on a steady flow of overseas capital. Once that flow slowed, some very ambitious projects suddenly looked a lot less financeable.
When investor appetite cooled, the structure began to crack. The failure of Jay Cooke & Co., a major financier of railroad expansion, helped trigger the panic in September 1873. Credit tightened, banks failed, businesses shut down, and the New York Stock Exchange closed for ten days. The boom had simply gone too far.
Even so, the bust did not make railroads irrelevant. The overbuilding led to painful losses, bankruptcies, and years of consolidation. Plenty of speculative lines failed. But the core infrastructure remained, and it helped power America’s long-term growth. The technology was real. The financing had just gotten ahead of the economics.
Echoes in AI
Now replace iron rails with data centers, semiconductors, power infrastructure, and cloud platforms. Replace railroad bonds with the massive pools of capital now being committed to AI. The pattern is familiar. A transformative technology attracts serious money, then more money, and eventually the kind of money that starts assuming every project connected to the theme will work out just fine.
We saw something similar in the late 1990s. The internet was real. The capital spending was real. A lot of the business models were much less so. Investors were right about the direction and sloppy about the details, which is usually where trouble begins. That combination created a few enormous winners and a lot of financial roadkill.
Important Differences
History rhymes, but it does not repeat line for line. A few things make today’s environment meaningfully different.
First, the earnings are real. Unlike many railroad ventures that were waiting for traffic to arrive, or dot-com companies that were held together by optimism and PowerPoint, today’s leading AI companies are generating real revenue, strong margins, and substantial cash flow. This buildout is being funded by businesses that are already producing.
Second, valuations are elevated, but nowhere near late-1990s territory. At the peak of the dot-com bubble, the Nasdaq traded at roughly 200 times earnings. We are not close to that kind of excess today. Is the market cheap? No. Is it behaving like it has completely lost adult supervision? Also no.
That does not make risk disappear. Capital spending can outrun demand. Overcapacity can show up. Debt and financing terms still matter. At some point, leverage can shift from fuel to friction. We just do not think that point has arrived yet.
Investment Approach
Our stance remains constructive. We want to participate in the opportunity, take profits along the way, manage our exposure, and keep a close eye on the point where the fundamentals begin to soften or debt starts to matter more than the story.
The lesson of 1873 is not that every boom should be feared from the start. The lesson is that you need to know what stage of the boom you are in. Getting out too late hurts. Getting out too early can feel smart and still cost you a great deal.
Right now, the evidence suggests we are still closer to the middle than the end. The infrastructure being built matters. The leading companies have real economics behind them. That is enough for us to remain bullish, while still keeping one eye on the exits like any sensible person at a crowded theater.
If you’d like to talk about how much of this theme belongs in your portfolio, give me a call at 661-302-4531 or email Jeremiah.Bauman@LPL.com. We’re happy to help you stay exposed to the upside while keeping the portfolio grounded in reality, which is still a useful discipline in markets where enthusiasm occasionally gets a much higher valuation than earnings.
The opinions voiced in this material are for general information only and are not intended to provide specific advice or recommendations for any individual. All performance referenced is historical and is no guarantee of future results. All indices are unmanaged and may not be invested into directly.
All investing involves risk including loss of principal. No strategy assures success or protects against loss.