Think at London Business School
The year is 1816. Investors are excited by a boom in a new technology. Cash is pouring in and the returns are appetising. Wind forward a few years and people who put their money into this innovative industry are doing well: between 1816 and 1824, stock prices more than doubled, up 140%.
The industry in question was canals. Emerging as a disruptive technology, the canal systems made the carrying of goods far cheaper than by horse and cart.
Then, in 1825, the Stockton and Darlington Railway opened for business. The disruptor became the disrupted.
Once rail freight became established, it was 60 times more efficient than canals in ton miles per day. Over the following quarter century, stock prices in canal companies fell by 70%.
It was the turn of railways to boom: in 1835, the stock prices of railroad companies doubled within a single year – although they subsequently fell back to their previous level. They doubled again by 1845, then fell by two-thirds by 1849.
There are clear lessons to be learned, even now. We vividly remember the boom in technology stocks in the second half of the 1990s and the subsequent bursting of that bubble: the index of US technology stocks rose nine-fold to a peak in March 2000, then fell by 82% in the following two-and-a-half years.
There is nothing new about booms and busts in the valuations of innovative industries. In his book Engines that Move Markets, Alasdair Nairn looks at the history of investments in new technologies. In many cases, these innovations are greeted with scepticism and derision. Take one example. In 1825, the year of the Stockton and Darlington railway, the Quarterly Review – an influential literary and political periodical of the time – opined: “What could be more palpably absurd than the prospect of locomotives travelling twice as fast as stage-coaches?”
In many cases though, this scepticism is followed by over-enthusiasm, with new technologies leading to stock market bubbles: investors suspend rational valuation. In time, that euphoria wanes, and valuations come back to earth.
So does this mean that investors should steer clear of new-fangled technologies? Are they simply being lured into a boom that is going to go bust? Not necessarily.
Look again at the example of the dotcom boom. It is certainly true that the huge increase in valuations from 1995 to 2000 was not sustained. But after that boom went bust, there was a marked recovery. In fact, since 1995, technology has turned out to be a strikingly good investment for those with the nerve to hold on throughout: annualised returns have grown 10.3% a year, beating the 6.9% for US stocks as a whole over the past 24 years.
Perhaps, then, the lesson is to avoid industries that appear to be in terminal decline? Again, that’s not necessarily true. Take the example of railways. At the start of the 20th century, in the US they accounted for 63% of the value of US stock market; in Britain, they represented almost half the market.
No longer. Railroads now account for a vanishingly small proportion of stock market investments. Over much of the 20th century, railway stocks performed badly as their industry faced new competition. The 1950s and 1960s were particularly difficult as trucking lured away freight traffic. Americans took to their cars. Airlines creamed off long-haul passenger traffic. Many railroad companies went bust. In 1970, the collapse of Penn Central marked the largest bankruptcy the US had ever seen.
But look at what happened subsequently. The railroad industry in the US was nationalised and productivity increased. That meant railroad stocks picked up. In fact, they did so well that they more than compensated for their earlier weakness. Indeed, since 1900, they actually outperformed the US stock market as a whole.
‘Certainly, there can be times where stock prices in new industries reflect an unrealistic enthusiasm for their future. And there can be times when investors are too pessimistic about the prospects for declining industries’
So it appears that investors should shun neither new nor old industries. Certainly, there can be times where stock prices in new industries reflect an unrealistic enthusiasm for their future. And there can be times when investors are too pessimistic about the prospects for declining industries. But these errors will not always be in the same direction: there will also be times when investors underestimate the potential of new industries and overestimate the survival prospects of industries in decline.
That is when the tools of classic investment analysis need to be deployed to sort the wheat from the chaff – be it in new, innovative enterprises or old, long-established ones.
Investors’ over-enthusiasm for young businesses in pioneering industries is well illustrated by the performance of such investments after they make their stock market debut through initial public offerings (IPOs). From the canal-building boom more than 200 years ago to the internet revolution of 1990s, the birth of new industries has been accompanied by successive waves of companies joining the stock market through IPOs.
Evidence from IPOs in both the US and the UK over recent decades shows a consistent pattern. On average, investors who buy IPOs at their issue price see a profit on day one: stocks go up. But that profit is likely to be more than offset by under-performance over the following few years.
The curious aspect of this is that the apparent over-pricing of IPO stocks is a well-known phenomenon. By now, investors should have learned their lesson; they should require that stocks are more modestly priced when they first come to market. Yet the phenomenon persists. It seems that investor hopes triumph over their experiences. They are prepared to bet on a long shot in the hope that they have identified the next big winner: they accept that there may be some laggards and failures among these innovative stock market newcomers, but believe that losses on these losers will be more than offset by the gains by backing the next decade’s new Microsoft. Experience suggests that belief is misplaced.
There is an investment strategy that avoids over-paying for newly-listed stocks and exploits their growth as companies mature or ‘season’. Consider the impact on UK stock returns of seasoning: the time that has elapsed since a company’s IPO. We tracked the returns over the years 1980 to 2015 from a strategy of investing in stocks which at the start of each year had gone through differing periods of seasoning: three years or less; four to seven years; eight to 20 years; and more than 20 years. The four portfolios were rebalanced annually to makes sure they always captured the desired range of seasoning.
The results of this experiment showed a pretty clear picture: the greater the seasoning, the higher returns. There was one exception to this pattern – around the dotcom boom and bust. But overall, by the end of the period, an investor’s wealth from investing in the most-seasoned stocks would have been almost three times higher than from investing in the least-seasoned. Put simply: old beats new.
If industries can experience periods of over- or under-valuation, perhaps it is possible to exploit this in making choices about which industries to invest in. If certain industries have become over-valued, then we might expect that the pattern will be reversed, with past losers beating past winners: concentrating on stocks in industries that have lagged in the year just ended should reap rewards as their prices recover in the year ahead.
In fact, the reverse is true. History suggests a momentum effect: industries where investment returns have been strong in year one are likely to have offered superior returns in year two; meanwhile, the losers – those that have lagged in one year will probably have continued to do so in in the following 12 months.
This momentum effect is substantial. Taking data for 1900 to 2014, we ranked industries by their returns over each past year, then divided them into five groups from best-to-worst performers. Equal amounts were invested in each quintile. Industries were re-ranked annually, bringing in new ones that emerged and dropping those for which indices were no longer produced. The strategy was reproduced annually over the whole period.
‘For the patient, long-run investor who can weather such episodes, the past success of these strategies may provide food for thought’
The results: over the full period, in both the UK and US, winners tended to continue to win. In the US, investing in the previous year’s winners produced an annualised return of more than 13%, compared to losses of 8% from investing in the previous year’s losers.
Dealing costs notwithstanding, over the full 1900 to 2014 period, investors in the US would have grown 870 times richer by buying winning industries rather than losers.
We also looked at a value-based rotation strategy. At the start of each year, we ranked industries by their yield or book-to-market ratios, divided them up into five groups, and then looked at their subsequent returns. Both low yield and low book-to-market are associated with growth industries. Companies from new industries and technologies tend initially to pay low dividends, retaining cash for growth and investment. Mature and declining industries with fewer prospects pay out more. In growth industries, a large part of market value will comprise capitalised future opportunities not yet reflected in book value or assets in place. Thus lower yield and book-to-market industries tend to be newer growth industries, while those with higher valuation metrics tend to be older, or ‘value’ industries.
Our analysis revealed an industry value premium in both the USA and UK. In other words, ‘value’ industries gave higher returns than ‘growth’ industries. The existence of this premium is consistent with there being periods of overvaluation for growth industries that the rotation strategy helps to avoid, and periods of undervaluation for value industries that the rotation strategy helps to exploit.
Of course, we are describing past patterns here, which cannot definitively tell us what will happen in the future: a future of as yet unimagined innovations and inventions. Even looking at the past, one needs to bear in mind that rotation strategies have failed in around one year in three.
Sticking to such strategies can be acutely painful, especially at market turning-points. But for the patient, long-run investor who can weather such episodes, the past success of these strategies may provide food for thought.
Elroy Dimson is Emeritus Professor of Finance at London Business School and chairs the Centre for Endowment Asset Management at Cambridge Judge Business School.
Paul Marsh is Emeritus Professor of Finance at London Business School and has been Chair of the Finance subject area, Deputy Principal, Faculty Dean, an elected Governor and Dean of the Finance Programmes (including the Masters in Finance).
Dr Mike Staunton is the Director of the London Share Price Database, a research resource of London Business School.
Every year, Professors Elroy Dimson and Paul Marsh and Dr Mike Staunton collaborate with the Credit Suisse Research Institute to produce the Credit Suisse Global Investment Returns Yearbook, the authority on long-term financial asset returns.