When Investing In Stocks, The Long Term Starts At Three Decades
February 29, 2016
The claim that the stock market is an investment for the long term is as common as empirical, robust thresholds for it are rare. Uncertainty of returns, possible losses and recovery from market crashes show that things will not start improving before fifteen to twenty years (with regular investments, it takes around four more years). But the full benefits of long-term investments are not reached before 30–35 years. This rather long time is what it takes to be fairly insulated from market cycles, so that when one happens to start investing does not matter too much.
When Investing In Stocks, The Long Term Starts At Three Decades – Introduction: The elusive long term
The short term is more often quantified than the long term
The claim that the stock market is an investment for the long term is everywhere, for instance in the titles of popular books such as Stocks for the long run (Siegel 2014). The idea is not recent, as show the books Common stocks as long term investments (Smith 1924) and Everyman and his common stocks: A study of long term investment policy (Sloan 1931).
But when exactly does this long term start? Banks and fund managers talk of things like minimum investment time horizon. Recommendations tend to look something like this: “A rule of thumb in investing is that you should not invest money in stocks or stock funds that you will need in less than five years” (Brennan 2002). William Bernstein (2002) also uses five years as threshold, while Jim Cramer (2009) prefers four and Suze Orman (2008) plays it safer with the range 5–10 years.
However, as Fama and French (2009) point out, “ten years is not a long period for stock returns, and ten-year periods with negative market premiums are common”. Indeed, these authors say that a stock investment shorter than some length would be risky, but they do not say that a longer investment would automatically be safe: four, five or ten years may be too short, but nobody says what would be long enough. And one cannot but wonder how these numbers were chosen. Investors, their advisors, fund managers, authors, etc. all love the long term, but they do not know (or ask) how long exactly is the long term.
The ambiguity of the concept of ‘long term’
In fact the very meaning of the phrase ‘long term’ is ambiguous. It is often limited to an unhelpful quasi-tautological contrast with the short term. Ang and Knut (2012) for instance “define a long-horizon investor as an investor having no specific short-term liabilities”.
In theoretical articles, the use of the phrase is generally asymptotic, as synonym for ‘eventually’. ‘Stocks are a long-term investment’ then means that if one waits long enough things will eventually go one’s way. But it says very little about what may be expected over a finite length of time, which is where people happen to live. This is just as unhelpful as saying that “in the long run we are all dead” (Keynes 1923). And Warren Buffett (1988) claiming that his “favorite holding period is forever” is not helping either. As Ed Easterling (2005) remarks, “A long-term perspective is quite relevant for scholarly studies and for the assessment of long-term relationships, but is not a good idea for developing assumptions that are to be used for investment planning.”
Numerical but fragile
The third type of use of ‘long term’ does aim at finding a specific length of time. For instance, Edgar Smith (1925) was “unable to find any twenty-year period within which diversification of common stocks has not, in the end, shown better results, both as to income return and safety of principal, than a similar investment in bonds.” Eighty years later, the mark had been pushed up: Pu Shen (2005) found that “when a holding period was at least 26 years, stocks never underperformed bonds.”
See full PDF below.