Investors are constantly advised not to try to time the market and to remain invested throughout the inevitable ups and downs.
The evidence in favour of this approach is compelling.
Between 1986 and 2015, missing the FTSE All Share index’s 10 best days would have cut your overall return from investing in the index almost in half – and no investor, private or professional, seems to have found a reliable way to time their sales and purchases of shares to get the best from the good days while avoiding the bad.
Yet there are a number of market phenomena that seem to suggest successful market timing is possible.
Telegraph Money has taken a look at some of the adages and anomalies that are at odds with conventional investing wisdom, and the data behind them.
Much of the data used has been sourced from the Stock Market Almanac 2016 and 2017 editions, written by Stephen Eckett.
1. Sell in May
The best known investment adage is “sell in May and go away”. The idea is to avoid shares entirely between the beginning of May and the end of October, when markets are supposedly more volatile and more prone to steep falls.
The performance of the FTSE All Share index, which is a broad representation of the UK market, appears to support this belief.
In the 34 years since 1982, the winter period – November 1 to April 30 – has outperformed the summer period 29 times. The average annual outperformance of the winter period is 8.8 percentage points.
Money has still been made in the summer months, although the returns have been much lower and have relied on dividends.
Both the FTSE 100 and FTSE All Share returned less than 1pc on average in the summer periods between 1986 and 2015, with dividends reinvested. Without dividends reinvested, the average falls to minus 1pc.
The cumulative return of the FTSE 100 over those periods is 21pc with dividends reinvested, and minus 30pc without.
But adopting the “sell in May” approach is not a certain route to stock market success: unexpected events can easily intervene.
Take last year as an example. Thanks to the fall in the pound after the surprise vote to leave the EU, which boosted the income of large international firms in sterling terms, the FTSE 100 climbed from around 6,100 points to nearly 7,000 between May 1 and October 30, delivering a total return of 14pc.
Any “sell in May” investor would have sold low and bought high.
Adrian Lowcock of Architas, the investment firm, said: “It is impossible to predict short-term trends with enough accuracy to bet your hard-earned savings. Last year is a reminder that markets are not guaranteed to fall in the summer months.”
He suggested that investors instead add defensive holdings to their portfolio as protection against weaker summer months, rather than selling out of markets altogether.
2. Solar eclipses
One of the more surprising events to exert an apparent influence on markets is the solar eclipse.
Since 1900, 15 total solar eclipses have been visible from American soil.
On the day of each eclipse, and on the day before, the average return of the Dow Jones index has been minus 0.3pc.
The average for the day after the eclipse has been 0.2pc.
Just 13pc of the days before an eclipse experienced a positive return, compared with 47pc for the day itself and 80pc on the day afterwards.
As the 15 examples go back to 1900, the results could be skewed by superstition and fear not present in the modern world.
Additionally, as the average profits on offer are so small, the benefit could be wiped out by the cost of trading.
3. ‘Turn of the month’ investing
This phenomenon is perhaps more soundly based.
The argument goes that most market gains come in the days surrounding the end of each month, making it possible to capture profits while being invested for only a week of each month.
The reasoning is that the end of the month is when large institutional investors such as pension funds routinely reinvest dividends, giving markets a boost.
Technical factors to do with pension fund administration also play a part.
The evidence is striking. Between 1969 and last year, an investment in the FTSE All Share index during the six days around the turn of each month would have returned nearly 3,000pc.
An investment excluding those days would be down by 30pc over the 47 years. Remaining invested for the whole time would have led to a 2,000pc return.
The costs of the extra trading involved in the “turn of the month” strategy would reduce the excess returns to some degree.
Darius McDermott of Chelsea Financial Services said: “Data can generally be produced to support any argument – for and against. Most investors don’t have the time, resources or knowledge to be this active.
“Trying to time the market is a huge risk, today more than ever when markets and stocks are not necessarily reacting to events in the way you might previously have expected.”
4. Santa rally
The flip-side of the “sell in May” effect is that markets are often strong over the winter months, but the “Santa Rally” involves a steeper rise in the last days of the year.
There are numerous theories about the cause of the rally, including fund managers making portfolio changes before the year end, people investing Christmas bonuses and even Christmas cheer improving investor sentiment.
While the putative reasons may vary, the data in favour of the rally’s existence is convincing.
If we look at nine trading days over the festive season – the three before Christmas, the three between Christmas and new year and the first three of the new year – the average return for the FTSE 100 over those days between 1984 and 2015 significantly exceeded the average daily return for the rest of the year.
The FTSE 100 rose by an average of 0.03pc a day between 1984 and 2015.
By comparison, the average of each of the nine days is substantially higher, ranging from 0.08pc to 0.47pc – three to 15 times greater.
It is the fourth trading day of the nine that has been the strongest, with a positive return 83pc of the time and a 0.47pc average rise.
- Follow us for more investing ideas: @TelegraphInvest
- For investment tips and ideas five days a week, read Questor