This website uses cookies in order to improve user experience. If you close this box or continue browsing, we will assume that you are happy with this. For more information about the cookies we use or to find out how you can disable cookies, see our Cookies Notice.

Seasonal investing

Sell in May, reinvest in September – this rule of thumb has raked in extra returns in the past, but how reliable is it?

This bull market has certainly clocked up plenty of mileage. For eight long years, global equity markets have been rising. Is now the time to consider selling? Not yet, we would argue. For one thing, share-price rallies in the past have occasionally lasted even longer, sometimes well over ten years. In addition, there are fundamental reasons in favor of holding out. The global economy is gaining momentum, and interest rates in industrialized countries are likely to remain fairly low due to only moderate rises in inflation. Furthermore, growing but still subdued inflationary pressures afford central banks an opportunity to proceed slowly and cautiously in returning to somewhat more normal monetary-policy conditions.

Another reason for staying active is seasonality. "Sell in May and go away" is one old stock-market rule, often followed up with "but remember to come back in September." A look at the Dax share-price history shows that there is a kernel of truth to this rule. This can be seen by doing a monthly performance analysis. Such an analysis reveals that since its launch in late 1987, Germany's leading index has posted price gains in more than two-thirds of the cases in the months of October, November and December.

In the majority of instances and thus on average, the Dax also performed well in the first four months of the year. In the past 30 years, the weakest months have tended to be August and September. These months have been marked by large price reversals and negative average returns. Usually, it was well worth waiting until late September, before considering to re-enter the markets again.

Profitable timing

The above stock-market rule has some validity beyond German shores. Given that the advice about selling in May originally emerged on Wall Street, this should come as no surprise. In an interesting piece of financial research, Professors Sandro C. Andrade, Vidhi Chhaochharia and Michael E. Fuerst from the University of Miami examined seasonal effects in 37 countries from 1998 to 2012. Based on local MSCI indices, they determined that the gains on the various equity markets from November to April were around 10% higher on average than those from May to October. #

Ben Jacobsen, a professor at TIAS Business School, and Sven Bouman, CEO of Saemor Capital, reached a similar conclusion for the 37 countries for the period from May 1970 to October 1998. # Since October 31st is Halloween, they named this anomaly the Halloween effect. # The statistical evidence should have led to investors capitalizing on this market anomaly since 2002, when the article by Bouman and Jacobsen was published. Instead, the anomaly continues to occur. For whatever reason, savvy investors have not yet been able to neutralize the Halloween effect.

The Halloween effect's persistence in the stock markets has long been a source of fascination among market participants and academics alike. Bouman and Jacobsen have suggested that vacation periods account for its timing and length. Their theory is that investors reduce their risk exposure before going on vacation by selling securities. During the vacation period, trading volumes are low. Once the vacation season ends, investors begin expanding the share of equities in their portfolios again. A correlation with vacation periods is apparent, but that does not prove causality.

Unpredictable October

The risk of a crash is regarded as quite substantial in October, and for good reasons. On average, volatility is relatively high, compared to other months. However, in a majority of cases during the month of October, the Dax also saw gains. As a result, this has contributed to much debate as to the timing of Dax equity purchases − as is customary in international analyses − in late October or in late September.

Though past performance may not necessarily be indicative of future results, a calculation of the data from the end of 1987 to the end of September 2017 remains compelling: Investors who purchased Dax equities at the end of September and sold them again in late April generated a profit of 3,222% over this period. In contrast, investors who did not join the action until late October saw a profit of just 1,581%. Investors who purchased Dax equities at the end of 1987 and held on to them until today have done even worse, achieving a performance of 1,183%.

Reduced risk

Timing also paid off from a risk perspective. Volatility − which measures the intensity of share-price fluctuations – has historically amounted to 22% for "buy and hold" since the Dax was launched. For investors who bought equities each year in late September and sold in late April, volatility decreased to 17.1%. This figure even dropped to 14.7% for those who bought in late October and sold in late April. However, investors who waited until late October to buy forfeited significant profits. Despite the higher risk in October, for some, buying late in September may have been worthwhile, depending on the level of risk they were willing to take.

There are two flies in the proverbial ointment: Investors who followed the stock-market rule this year missed out on hefty share-price gains in September. Moreover, it still seems unlikely that the Halloween effect will persist forever. In theory, investors will eventually capitalize on this market anomaly. If enough do so, the excess returns from late September to late April will disappear, thereby making the market more efficient. As yet, however, there are few indications of this happening.

Dax boasts strong final quarter

Since its launch in 1988, the Dax has posted the highest average returns in the fourth quarter, while usually seeing a lull in August and September.

Source: Thomson Reuters Datastream; as of 10/2/17

* In the period from Jan. 1988 to Sep. 2017

Erratic October

In the past, the Dax has seen greater fluctuations in October. The index has traditionally been calmer in spring and summer.

Source: Thomson Reuters Datastream; as of 10/2/17

* In the period from Jan. 1988 to Sep. 2017


Sandro C. Andrade, Vidhi Chhaochharia, Michael E. Fuerst: "Sell in May and Go Away" Just Won't Go Away. In Financial Analysts Journal, Volume 69, Number 4, 2013.


Sven Bouman, Ben Jacobsen: The Halloween Indicator, "Sell in May and Go Away": Another Puzzle, American Economic Review 92, 2002.


In the research papers by Andrade, Chhaochharia and Fuerst, as well as by Bouman and Jacobsen, April 30th was specified as the time to sell and October 31st was specified as the time to buy.

Related Articles

Jul 16, 2018 New Americas CIO View

Americas CIO View

How much Value is to be found abroad?

Jul 13, 2018 New Chart of the week

Chart of the week

China's monetary base looks set to grow faster again soon

Jul 06, 2018 Chart of the week

Chart of the week

On trade, the Trump administration might have some powerful allies.

Jul 03, 2018 Investment Traffic Lights

Investment Traffic Lights

Our tactical and strategic view

Jul 03, 2018 CIO Special

Dollar pros and cons

The dollar has already reached our target. Currently, the arguments are balanced.

Jun 29, 2018 Chart of the week

Chart of the week

Do weak currencies hurt emerging markets?

Jun 22, 2018 Chart of the week

Chart of the week

In the line of fire

Jun 21, 2018 Macro Outlook

Ten years after

The long reach of the financial crisis

Jun 15, 2018 Americas CIO View

Americas CIO View

Inflation: Sometimes it skips a generation

Jun 15, 2018 Chart of the week

Chart of the Week

Looking at real federal funds rates, not a lot has happened after 7 hikes

Jun 08, 2018 Americas CIO View

Americas CIO View

Is it time for U.S. Small Caps to shine? If you pick them right

Jun 08, 2018 Chart of the week

Chart of the week

In some areas, Italy is actually doing quite well

Jun 04, 2018 Investment Traffic Lights

Investment Traffic Lights

Our tactical and strategic view

Jun 01, 2018 Chart of the week

Chart of the Week

Why emerging markets may be less vulnerable than they used to be

May 29, 2018 CIO Flash

Euro crisis 2.0?

Italy's political woes are dragging down markets while boosting our dollar call.

May 29, 2018 Americas CIO View

Americas CIO View

What do investors want from active managers?

May 25, 2018 Chart of the week

Chart of the week

A tale of two economies in the Eurozone periphery

May 24, 2018 CIO Flash

Italy's new coalition

Italian political turmoil might prove less worrisome than many think.


Please let us know what you think about this article/page.