This week, we witnessed the end of the greatest bull market in history.
It’s nothing to do with technology, house prices, gold. None of that fancy stuff.
I’m talking about fags.
For 118 years, tobacco stocks have been among the best-performing stocks on US and global stock markets. Research from Elroy Dimson and Paul Marsh at the London Business School has shown that tobacco is the best-performing industry in the history of the US stock market, going back as far as 1900.
The level of outperformance is just ridiculous. Their research showed $1 invested in the US market in 1900 would be worth $38,000 today. $1 invested in tobacco stocks would be worth $6,300,000.
But I wouldn’t be rushing to put your money into tobacco stocks if I were you. Tobacco’s run looks like it’s coming to an end.
This week Philip Morris reported dreadful quarterly numbers. Its effort to reposition itself as a nicotine-delivery-technology company isn’t paying off. Nobody’s buying Philip Morris’s fancy vape pens. The stock has dropped 20% on the news — or $23 billion dollars in value. Other tobacco stops dropped in sympathy.
I don’t think many people will feel sympathy for Philip Morris shareholders. Hopefully the next 118 years of stock market returns will produce a more worthy winner.
But why is it that tobacco stocks outperformed? How did they beat the market so well?
Tobacco stocks: small caps in disguise
Research has shown “sin stocks”, ie companies selling stuff like fags, booze, gambling or guns, tend to outperform the market significantly over time.
Why is that? Well it’s sometimes speculated that upright and ethical investors shunned those stocks, which pushed down their price and up their returns.
For example the yield in US tobacco and casino stocks has stayed steady at about 4% for a century, which is above the average yield in the market.
But things aren’t as simple as they appear. According to research by Andreas Hopener at Henley Business School, the outperformance of sin stocks like Philip Morris isn’t what it appears to be. And the story about how ethical investors lower the demand for sin stocks and raise their price is a load of nonsense.
I’m sure this isn’t the first time you’ve heard something like this.
Stocks in XXX industry have been proven to outperform!
When the price of copper goes down, stocks in YYY have been shown to go up!
You should be very careful with these claims.
Reason one: there’s masses of data on stock prices, going back to 1900 and earlier. And there are thousands of financial economists prodding, poking and testing that data. Given enough regressions, the economists will eventually find some statistically significant relationships by pure chance.
Did you know that Coca-Cola is tightly correlated with a water utility from southern Brazil?
Reason two: every individual company has a thousand different characteristics. It’ll have distinctive value, growth, capital structure, industry, governance, momentum, and on and on and on.
But in the 50 years or so that financial economists have been crunching the data, they’ve only find a handful of characteristics that persistently outperform the overall market.
1 Size. Investments in the stocks of companies dramatically outperform investments in the stocks of big ones.
2 Value. Investments in companies that are cheap relative to their assets outperform.
3 Momentum. The shares in certain companies seem to have “momentum” – they keep moving in whatever direction they’re going.
Very often, weird new stock price “anomalies” — like the outperformance of tobacco stocks — are size, value or momentum in disguise.
You see, tobacco stocks tend to be smaller than most stocks in the market. And as we’ve seen, small companies strongly outperform the market.
The research from Andreas Hopener at Henley Business School shows that when you control for their size, tobacco stocks show no tendency to outperform.
That’s good news! It means you haven’t missed the boat on tobacco-stock like returns. You can just buy small fast-growing companies instead.
Equally good returns, and an untroubled conscience.