Credit Suisse’s Global Investment Returns Yearbook 2015 is a must read for investors. This year, the yearbook contains some extremely interesting long-term (115 years) data on industries.

Changing industries

Today, in the US and UK markets only the banks and mining industries have weightings close to their 1900 levels. Industries have risen and fallen as technology has advanced, but industries remain one of the original and most important factors in portfolio organization. When fund managers build, alter, or report on portfolios, they refer to industry weightings. So, getting these weightings right — or wrong — has consequences. Industry membership is the most common method for grouping stocks for portfolio risk management, relative valuation and peer-group valuation.

Credit Suisse Group AG (ADR) (NYSE:CS)’s research team covers more than 100 years of data in its research report. Starting at 1900 markets were dominated by railroads. In the UK, railway companies accounted for almost half the value of the stock market while in the USA they had a 63% weighting.

It’s clear that the market today is much more diversified than it was 100 years ago, and rail stocks have all but disappeared.

Long-term performance

Using Ken French’s industry data (Fama and French, 1997), Credit Suisse then goes on to look at the performance of US industries. Using two data sets, that of Fama and French, 1997 (1926 to 2014) and 57 Cowles industries (1900 to 1925) Credit Suisse’s data shows that a dollar invested in the US market at start-1900 would have grown, with dividends reinvested, to USD 38,255 by end- 2014, representing an annualized return of 9.6%.

The worst performing industry was shipbuilding. A dollar invested in shipbuilding and shipping in 1900 would have grown to just USD 1,225 today, representing an annualized return of 6.4%. Tobacco came out top with an annualized return of 14.6%.

It should be noted that this data does suffer from hindsight bias. The sample only contains those industries that existed in 1900 and which survived.

New industries

100 years ago the market was dominated with industrial companies. Today, in the US at least, there’s a heavier weighting towards technology.

Credit Suisse’s data shows that over the 20 year period from 1995, the beginning of the dot-com bubble, to 2015 tech stocks actually beat the market, with an annualized return of 10.5% versus 9.9% for US stocks as a whole. Despite the bubble, the technology sector has, for most investors, generated good returns — unless you were unlucky enough to buy at the top of the market in 2000.

However, Credit Suisse’s data also shows that old; declining industries can also provide good returns. A prime example is the railroad industry.

According to long-term data from 1900, to date, railroads outperformed the market. The industry struggled during the 1950s and 1960s, as trucking took much of their freight traffic as Americans began to drive or fly rather than taking the train. But since the 80s railroads have outperformed airlines, road transport and the wider US market in general.

Conclusion

Industries are a key investment factor. To exploit diversification opportunities to the full, investors need to diversify across a wide spread of industries. It is interesting to see which have done best and worst, although this tells us little about the future.

The industrial landscape will change during the 21st century perhaps even more radically than in the past. Should investors focus on new industries and shun the old? Or should they be contrarian? It’s not possible to say but as the chart above shows, even the market’s oldest industries (rail) have the ability to outperform over the long-term.