Numerous studies have been carried out to study periods where markets exhibit certain characteristics that imply a good entry or exit point. Some of the more well-known studies include the 'January effect', the 'October effect', and even the 'Santa Claus effect'! In addition, research has been done on the effect of the 'unlucky 7s': this refers to the years that end in a '7', such as 1987 and 1997, where equity markets met with strong downturns. The more recent one in 1997 is probably still fresh in the mind of some investors, as the Asian Financial Crisis saw a broad-based crash in Asian equity markets.

But is there any truth in these studies which seem to suggest that there is a best time to enter and exit markets? We shed light on this question in the second quarter issue of the FSM magazine. In this first instalment on 'Is There A Best Time To Invest', we will focus specifically on whether there is a best month to invest in. In the second instalment, as we will mention later, we will focus on whether there is a best quarter to invest in.

Best time to invest: The 'January effect'

We start by focusing specifically on the months of the year where some studies have observed certain common characteristics. One of the studies touches on the 'January effect'. This refers to equity markets rising markedly during the period starting on the last day of December and ending on the fifth trading day of January.

What could account for the 'January effect'? One reason could be due to the corporations and individuals closing their tax books at the end of December. Thus, for investors who are sitting on paper losses, they would be more willing to sell their investments to create a situation of a tax loss. Unlike in Singapore, investors in the US need to pay taxes on their 'net realized capital gains and losses' in their investments and by reporting their losses, the overall taxes paid on capital gains tend to be reduced. The other reason for selling one's investments in December is to raise cash for the holidays by selling investments that are already profitable.

The sell-out in December may temporarily depress the equity markets without any real fundamental change. As a result, bargain-hunters may start buying at the beginning of the following year, causing the markets to rise in early January. Therefore, the 'January effect' suggests that investors should wait till January to invest as it is the month which posts the best returns in a year!

Worst time to invest: the 'October effect'

In the same vein, is there a worst time to invest in equities? The answer would be yes if we examine the 'October effect'. It suggests that equities tend to decline in October as investors get jittery as historically, several market crashes have occurred in that month. Examples include October 1929, following which the Great Depression started. Another example of a downfall in October is the 'Great Crash of 1987' which happened on 19 October, when the Dow Jones Industrial Average (DJIA) Index in the US plummeted 22.6% within a single day (in US dollar terms). Investors now remember it as 'Black Monday'. This event marked the start of a ripple effect as other major regional equity markets suffered from the after-effects. In the same month, the DJ Stoxx 50 Index (a proxy for the European market) fell 23%, and Japan's Nikkei 225 Index fell 12.5% (returns are in their respective local currency terms).

While there are many theories that attempt to explain why the crash happened, there was no obvious fundamental reason behind the crash. However, one common finding was that mass panic made things worse and many stock exchange regulators have instituted measures such as trading curbs and circuit breakers since then to prevent panic-selling.

Other than the 'January effect' and the 'October effect', there are also some lesser-known 'anomalies' recorded. For example, in the 'Turn of the month effect', Frank Russell Company studied the returns of the S&P 500 Index in the US over a period of 65 years, and found that US large-cap stocks consistently posted higher returns at the turn of the month. This was attributed to the cash flows at the end of each month when the salaries and interest payments were made to investors, and it was assumed that they would then use these monies to purchase stocks. The authors, Chris R. Hensel and William T. Ziemba, found that the returns at the turn of the month were significantly above the average over the period from 1928 to 1993, and 'that the total return from the S&P 500 Index over this sixty-five year period was received mostly during the turn of the month'.

So how much truth is there in these hypotheses? Do markets perform particularly well in January? And likewise, should a fall in the markets be expected in October? We examine these various anomalies by going through the historical returns of some of the main regional indices and our own local market, the Straits Times Index (STI). Let us start with the 'October effect'.

Analysing the 'October effect'

The month of October is supposed to be the worst month of the year to invest into, while January is supposed to be the best. In determining whether a month has performed badly, we look at its probability of giving a negative return, which would be high if the effect was true, and its average return over the past years, which would be low if the effect was true.

Chart 1 shows the probability of having negative returns in a particular month in the past 18 years (1989 to 2006). In the first instance, we test out the probability based on the STI only. In the second instance, we test out the probability on a basket of five indices, including the S&P 500 Index, DJ Stoxx 50 Index, Nikkei 225 Index, MSCI Asia ex-Japan Index, and the STI.

We observe that the probability of having negative returns in the month of August is 55.6% among the five indices we cover. This means that out of the 90 months of Augusts surveyed, 50 of them showed negative returns. When we look at the results of the STI, the probability of it having a negative return in August is highest at 61.1%. Thus, from Chart 1, we can actually see that the month with the most frequent number of negative returns in the past 18 years would be the month of August, not October.

In fact, there is only a 38.9% chance of the returns being negative in the month of October among the five markets and the STI. Given that this is the case, it is difficult to conclude that October is a bad month to invest.

Is January really the best month to invest in?

The 'January effect' suggests that equity markets tend to rise markedly in the month of January, and average returns tend to be greater. In Chart 2, we examine the probability of the various markets returning positive returns in the past 18 years from 1989 to 2006. We find that the probability for returns in January to be positive stood at 61.1% for the basket of five markets, which is reasonably high. However, the probability of positive returns in the months of November and December is even higher at 64.4% and 73.3% respectively. And although the probability of January delivering positive returns for the STI is relatively high at 72.2%, the probability is also high in the months of July and December too. It is, therefore, not conclusive that January is the month with the highest probability of delivering a positive return.

Some may argue that although the probability of a positive return in the month of January is not the highest, the actual returns in the month of January may actually be higher compared to the other months. In this case, it could be argued that the 'January effect' still applies. Similarly, although the probability of having a negative return in October is not the highest relative to the other months, the average return in October could be the lowest. In order to find out, we also crunched some numbers to calculate the average returns in each month over the past 18 years.

Average monthly returns in the past 18 years

From Chart 3, we can see that January is not the month which posted the strongest returns among both the basket of indices and the STI. For the STI, the average return in January for the past 18 years was 1.2%, but the average return in October, November and December was much higher at 2.4%, 3.0% and 3.8% respectively. The findings are similar for the basket of five indices, as the monthly returns for October, November and December turned out to be the highest: the average return in January was only 0.9%, compared to 1.8%, 1.8% and 2.5% in October, November and December respectively. Thus, we find that it is difficult to argue that the returns in January are higher compared to the other months.

We also note from Chart 3 that August and September turned out to be the worst months in terms of the average returns as they were the only two months that delivered negative returns. Based on the data, if you were to invest in the basket of five indices in the months of August or September, you would have lost 1.5% of your investments on average. On the other hand, you would have made 1.8% on average if you invested in October.

In fact, the study on the monthly returns reveals that December was the best month to stay invested, given that the average monthly return in December is the highest, whether it is for the basket of five indices, or the STI.

How variable are the returns?

While the returns in December have turned out to be the highest in our study, we also want to find out whether the variability of these returns is high too.

Based on our study, the variation of returns in December is quite substantial. Investing in December can give investors a return as high as 25%, which happened in 1993 for the MSCI Asia ex-Japan Index or it could lead to a decline of 10.1% if you had invested in the Nikkei 225 Index in 2000. Other months are also quite volatile: the difference between the best and worst returns can on average be as high as 33.5 percentage points!

The volatility, coupled with our finding that out of 10 Decembers, investors could lose money in 3 of them, support the fact that timing the market based on calendar months is a risky business. On the other hand, October displays the strongest negative return in our study: a decline of 20.1%. But the return in October (October 1998 more specifically) can be as high as 28.2% (for the STI). Thus, it is also difficult to categorically state that October is a bad month to invest in.

Conclusion

In our study, we have managed to debunk the 'October effect' and have shown that contrary to popular belief, October is not the worst month to invest in. We have seen that based on our studies, there was a 55.6% probability that instead of October, August would turn out to be the worst month to invest in. In fact, the probability of making money in October was 61.1%! We have also seen that there is evidence that the 'January effect' does not hold true either as the probability of positive returns in the last three months of the year is higher (based on the basket of five indices). One possible reason has been attributed to the 'January effect' being so widely documented, that it has become self-fulfilling and has caused investors to pile their money in earlier and earlier in the hopes of reaping the benefits from the 'January effect'.

However, we do not encourage investors to invest solely based on these results. While our analysis shows strong upside in certain months, the studies are based on the average returns in several markets. When market fundamentals are not sound, and a market goes through a sharp downturn, it does not matter which month or quarter an investor invests in - the investor would still be hit.

Another point to note is that to go in and out of markets during certain months or quarters actually entails high transaction costs, which would eat into the returns in the long run. In addition, there is no single month that has a 100% probability of having positive returns. It is better to look at the fundamentals in an equity market, including the attractiveness of a market (based on its PE ratio), its earnings growth, and its economic and political situation.

So far, our results show that there are some months that seem to deliver much better returns, namely October, November and December, and they happen to fall in the fourth quarter. This has also prompted us to study the differences in the probability of having positive returns in the four quarters of a year, and the magnitude of the returns in the different quarters. Thus, in the second instalment on 'Is There A Best Time To Invest', we will focus on whether there is a best quarter to invest in.

» Is there a best quarter to invest?

Is this article useful to you?