Market Anomalies and Seasonal Cycles

7.1What Is a Market Anomaly?

An anomaly is an exception or irregularity. In stock markets the word refers to statistical price biases that theories such as the efficient market hypothesis struggle to explain. Seasonal tendencies — prices tending to rise or fall in particular months, on particular weekdays or around particular events — are the typical examples.

The explanations offered come from participants' behaviour and market rules: tax-driven trading, the timing of bonus and pension inflows, institutions' fiscal year-ends and mandates, thinner participation during holidays and so on. In most cases, however, the reasons have not been firmly proven.

The value of studying anomalies is said to lie less in using them to decide trades than in understanding the backdrop — that this period tends to be volatile, or that money flows tend to shift at this time of year.

• Anomaly: a price bias that theory struggles to explain but that has been observed in the past
• Proposed reasons include tax rules, the timing of inflows, market rules and participant behaviour
• They are usually used as context for understanding the market

7.2Monthly Anomalies

The January effect is the tendency of small-cap stocks to outperform large caps and the market in January, reported in US research in the 1970s. The best-known explanation is the tax-loss selling hypothesis — losing positions are sold in December for tax purposes and bought back in the new year — but some argue the effect has weakened since it became widely known.

“Sell in May and go away” is a US market saying. The tendency for stock returns from November to April to exceed those from May to October is also called the Halloween effect, and a 2002 study reported similar patterns in many countries' stock markets. Thinner participation during the summer holiday season is among the reasons given.

In US stocks, September is known as a weak month on average. The diagram is a schematic of these monthly tendencies, not actual averages. Every month's tendency varies widely from year to year, and years that go against the average are not unusual.

Japan has its own saying, “a top at Setsubun, a bottom at Higan,” an old rule of thumb that prices tend to peak in early February and bottom around the spring equinox in March.

• January effect: small caps are said to be strong in January (some say it has weakened)
• Halloween effect: November–April returns are said to tend to exceed May–October
• In US stocks, September is known as a weak month on average

7.3Year-End and Turn-of-the-Month Anomalies

The Santa Claus rally is the tendency for stocks to rise over the last five trading days of December and the first two of January, popularised by a US investors' almanac. Lighter selling during the holidays and year-end bonus money are among the reasons given. Japan also has a phrase, tobi no isshin, for a final upswing at the end of the year.

The turn-of-the-month effect is the tendency for stocks to rise from the last trading day of the month through the first few days of the next. It is attributed to lump-sum money entering the market around the start of the month — salaries, pension payments and regular contributions to investment funds.

Japanese stocks also drop on the day after the last day to hold rights at fiscal period-ends (often March and September), by roughly the value of the dividend or shareholder benefit that has gone ex. This is a mechanical move rather than an anomaly, but it is sometimes described as a supply-demand bias, with buying gathering ahead of the last rights day and selling afterwards.

• Santa Claus rally: the last five trading days of December and the first two of January
• Turn-of-the-month effect: money flowing in around month-end and month-start
• Around record and ex-dividend dates, separate mechanical moves from supply-demand bias

7.4The Four-Year US Presidential Cycle

A US presidential term lasts four years: the first year in office, the midterm-election year (year 2), the pre-election year (year 3) and the presidential-election year (year 4). Historical data show stock-market tendencies that follow this cycle, known as the presidential election cycle.

The commonly cited pattern is that stocks struggle in the midterm year amid policy uncertainty and tend to bottom in the second half of the year, then tend to rise from after the midterms into the pre-election year, which has historically been the strongest of the four. One explanation offered is that administrations tend to favour growth-friendly policies ahead of the next election.

The diagram is a schematic of the four-year pattern, not actual averages. The pattern is based on decades of data, but each phase is observed only once every four years, so the number of observations is small. Wars, financial crises or large interest-rate moves can produce behaviour completely unlike the cycle.

• The midterm year (year 2) is said to tend to struggle
• The pre-election year (year 3) is known as historically strong on average
• Each phase is observed only once every four years, so data are few

7.5The FOMC and the Blackout Period

The FOMC (the US Federal Open Market Committee) sets US policy rates and meets eight times a year. Because interest rates weigh heavily on stock valuations, markets are said to move more around FOMC meetings.

A study published in 2015 reported that from 1994 to 2011 US stocks rose sharply on average in the 24 hours before FOMC statements, a pattern dubbed the pre-FOMC drift. Later research, however, has reported that the pattern weakened after the study became widely known.

Before each meeting comes a blackout period: from the second Saturday before the meeting until the day after it, Fed officials refrain from public comment on monetary policy. With no officials' remarks to move markets, prices are said to be driven more by other news such as economic data and corporate results.

• The FOMC meets eight times a year; rate decisions are major market events
• The tendency to rise before FOMC announcements (the pre-FOMC drift) has reportedly weakened since its publication
• During the blackout period, Fed officials refrain from public comment

7.6Futures and Options Expirations

Index futures and options expire, and a settlement price is calculated on expiration day. In Japan, index options settle on the second Friday of every month, and in March, June, September and December the futures expiry coincides — the so-called major SQ. In the US, index futures, index options and single-stock options expire together on the third Friday of March, June, September and December, a day known as triple witching.

Around expirations, rollovers to the next contract month, position unwinding and option hedge adjustments concentrate trading, so volume tends to surge and prices tend to be more volatile. The diagram is a schematic of volume swelling on expiration day.

In the US, constituent changes and weight rebalancing for major indices such as the S&P 500 often take effect on the same day, concentrating index-tracking money into the close. Much of the movement on such days reflects supply and demand that has nothing to do with company results or the economic outlook.

• Japan's major SQ: the second Friday of March, June, September and December
• US triple witching: the third Friday of March, June, September and December
• Rollovers and hedge adjustments tend to swell volume around expirations

7.7Cautions When Reading Anomalies

The biggest weakness of anomalies is data mining — finding chance biases in past data. Test enough combinations of months, weekdays and events and some bias will turn up somewhere. Biases with no plausible explanation, or that appear only in particular periods, are likely to be chance.

Anomalies are also said to fade as they become well known, because more participants try to front-run them. Studies have found that many anomalies reported in academic research shrank after publication. The January effect and the pre-FOMC drift are not the only examples said to have weakened over time.

Looking only at averages is risky too. A month with a positive average still includes years of steep declines. Use anomalies as background on what tends to happen at a given time, and weigh them together with the market's direction and specific news.