I was listening to a finance-related podcast last week (as one does), and they said they thought that the best-performing industries in the US were beer and tobacco.

The podcasters chuckled because they pointed out that market returns reflect what society wants, and companies respond by providing those goods and services. Beer and smokes – that’s what America has valued the most, as the story went.

I thought it was interesting and agreed with the point, but then I decided it would be even more interesting to see whether it was true. For that, I went to the Ken French data library at the Dartmouth School of Business website.

French is one of the most prominent academics in finance, as the co-creator of the Fama-French three-factor model and many, many important papers.

His library is considered the gold standard for market data. I’d seen before that he broke the returns down into 49 industries dating back to 1926 (when applicable; the software doesn’t go back that far for obvious reasons).

Based on my reading of the data from July 1926 to March 2024, the story the podcasters told wasn’t totally accurate, but they weren’t far off either.

The beer industry earned 12.0 percent over the full span, and tobacco earned 12.1 percent. Those are about two percentage points better than the overall market during that period, and compounding that much over that long led to massive wealth from those industries.

However, the top spots are aircraft stocks and firearms makers, which earned 12.4 percent per year. I’ve heard it said that no one has made money in the airline business, but it appears that making planes and parts is a good business.

What were the worst-performing industries? Gold stocks earned 4.9 percent per year, and real estate stocks gained 4.7 percent per year. Those returns aren’t much higher than Treasury bill returns, and the volatility was very high on both – about twice the overall market.

Gold stocks are ore producers, and real estate stocks are a smattering of things from apartment building operators, developers, and railroad lessors.

There were some oddities that I noticed while looking at the data. Hardware companies made 12.1 percent, but software made 5.6 percent. That doesn’t seem right to me, but I suspect the ubiquities phones in our pockets are hardware.

Of course, it goes without saying that while these industries were the best over the last 90+ years, it doesn’t mean they will be in the future. Furthermore, there will continue to be new industries, like healthcare, which was first made into the data in the 1960s.

I am barely scratching the surface of this data, and I plan to take a much more detailed look at the coming months and share what I find along the way.