Since 1900 tobacco has been the best performing industry in the US and alcohol has been the best performing industry in Britain – when compared to a sample of other old-economy industries like chemicals, mining, and shipping that have traded continuously over the past 115 years.
That is the conclusion of Elroy Dimson, Paul Marsh, and Mike Staunton, three professors at London Business School. Their February 2015 report is their latest in a series that use long-term historical data to test assumptions about markets.
The report did not analyze the fundamental drivers of tobacco and alcohol companies’ profitability (though it cited other works, noting: “The rationale for ‘vice investing’ is that these companies have a steady demand for their goods and services regardless of economic conditions, they operate globally, they tend to be high-margin businesses, and they are in industries with high entry barriers.”)
Rather, the report compares and analyzes the long-term performance of old-economy industries. It treats the outperformance of tobacco and alcohol as a function of undervaluation driven by social qualms about the industries. ‘Sin stocks’ like tobacco have rewarded investors over the ultra-long term because it costs relatively little to own a piece of their future earnings, and they demonstrably continue to earn despite expectations to the contrary.
Another takeaway is the absolute range of industries’ performance over 115 years. The average annual spread between the best- and worst-performing industries was over 100%, the professors found. Over the course of a century the widely differing fortunes of industries can lead to dramatic returns, especially when compounding is taken into account. A dollar invested in the listed US tobacco industry in 1900 would have yielded $6.28m by 2015 after a century-plus of the industry averaging 14.6 per cent annual growth. That compares to the US market average of a $38,255 (9.6 per cent annualized) return on $1 – and $1,225 for shipping, the worst performing industry.
The report implicitly endorses old-economy industries from an investment perspective. Even when their market value shrinks and they are perceived as declining, these industries often outperform over the long term, the report demonstrates.
FUSE talked to Professor Elroy Dimson, professor emeritus of finance at London Business School and the chairman of the Newton Centre for Endow
FUSE: Why were alcohol and tobacco the best-performing industries in the UK and the US over the long term?
ED: Tobacco and alcohol were somewhat unpopular industries over much the past century, with health-related issues being just one part of their lack of appeal. Some investors avoid ‘sin stocks’ like these. We decided to examine whether excluding a class of stocks has an impact on investment performance.
One view is that investors can do well by doing good: that more virtuous companies give a superior stock return and that ‘sin stocks’ underperform. Another view is that widespread divestment of a category of shares depresses their prices. We wanted to know which view might be correct.
FUSE: Maybe both views could be correct?
ED: Not really. Over the long term you cannot do well financially by avoiding ‘sinful’ companies while at the same time depressing the stock prices of those companies. If prices are driven lower than they would otherwise have been, that must generate a higher long-term return. These higher long-term returns are not necessarily because of capital appreciation. Even if the undesirable stocks stay forever below the fundamental value that they might otherwise have, the investor is still getting more dividends per pound that he or she invests.
So the surprise here, if you can call it a surprise, is that you can’t have your cake and eat it too with these ‘sin stocks.’ If the stock market price of tobacco or alcohol companies is permanently depressed, they offer a higher expected return in financial terms.
FUSE: Why do you say “in financial terms?”
ED: Company shares have a financial value but potentially they can offer a psychological value. Some people may get a warm glow – a psychological dividend – from knowing they own shares in a responsibly managed business. The value of a tobacco company may be depressed because its non-financial dividend is negative. In much the same way that we might favor buying shares in a company with a high psychological value – because we emotionally believe in what it is trying to do – we might avoid those with a perverse psychological value.
Prices may also be depressed when people expect costly interference: when potential massive litigation hangs over an industry, for example, but never has the impact that people expect. These factors in some cases push the stock price down, such that the expected return goes up.
Of course, investors never realize superior returns smoothly. Our study reports the average performance of industries over many decades. And there are many factors that can have an impact which have nothing to do with being sinful. We cite Dan Ahrens’ book which talks about the barriers to entry for businesses like tobacco and alcohol. To the extent that they were not anticipated, these barriers also contribute to long-term returns.
FUSE: You and your colleagues are noted experts on the FTSE Index. When you look at FTSE and other data over the past 50 years, do you find examples of industries that came into being more recently than 1900 – but that register the same sorts of returns of tobacco?
ED: If we look at the industries plotted over time, we see that there are somewhat more industries that did worse than the markets over time and fewer that did better over the long haul. Of the US industries that survived for the full 115 years of our study, the poor performers include coal, steel, textiles and shipbuilding – industries that migrated from the developed world to developing economies. In our paper we report research over more recent intervals, and the findings there are consistent with our ultra-long-term findings.
One of the most remarkable findings of our research is that the long-term performance of old-economy, long-established businesses was systematically better than the performance of shorter-lived, less-established businesses. We find this pattern holds up across different bands of companies – for example on the main market of the London Stock Exchange and the AIM [junior] market – over different time periods. Rather bizarrely, the same holds true even for investment companies. The stock market performance of old-established, closed-end funds is superior to the performance of younger funds.
FUSE: You talk about the seeming paradox of returns increasing over time from shrinking or declining industries such as US railroads. The same appears to be true for tobacco. What is going on there?
ED: As I mentioned earlier, the sequence of returns is never a smooth process. For example, there were long periods when railways underperformed the other transportation industries. But eventually the rail industry pulled ahead. The important question here is not about the strength of an industry or sector. Rather, the important question is whether people are overly pessimistic about a dying industry. It appears that investors are sometimes too pessimistic about the tried and trusted, long-established industries. And on average this pessimism pushes prices down further than they might otherwise have gone.
FUSE: Your research shows a remarkably wide range of industrial performance over more than 100 years. Do you see that range narrowing in the present century?
ED: In our study the difference in annual return of the best-performing US industry and that of the worst-performing industry was high. Averaged across the 115 individual years the gap in returns was 108%. I don’t see any reason why this performance gap should narrow on average. People are constantly guessing about what is going on, and the news is available faster than it was. Small changes in information about an industry can still give rise to large changes in value.