No matter the security or market, investors and analysts have attempted to find a seasonality factor to exploit. One of the key principals most traders live by is that history repeats itself; and in that repetition there is an edge that can provide a profit. So, what is seasonality? Broadly, the term seasonal effect can be applied to any market pattern in price action that occurs during a specific period of a calendar year. This could be a specific direction in price action during a certain season or a consistent jump in volatility over the same month every year.

Over the years, traders and analysts have picked up on a number of well-known and relatively obscure seasonal effects across the spectrum of securities. Without a doubt, the equities market is the most popular arena to find these influences. Investors booking losses at year end, quarterly earnings, election cycles and ‘sell in May and go away’ are all relatively well-known axioms. However, stocks aren’t the only instruments that have been assigned a tradable pattern. The commodity markets are exposed to consumption cycles that lead to an imbalance of supply and demand, and even the fixed income market has its own response to government reshuffling. In this article we will offer a fundamental explanation behind each of these supposed seasonal effects, support or dispel them with data, and suggest how those confirmed patterns can present high probability trades.

Stocks are highly susceptible to seasonal patterns (or perceived patterns) due to the strict tax and governmental regulations that exist in most of the world’s developed economies. While the same effects can be seen in many countries - with some adjustments - this section will focus on the US benchmark Dow Jones Industrial Average for ease’s sake.

Earnings Session – Corporate governance and legislation designed to protect investors were the initial impetus behind quarterly earnings reports. However, these good intentions were quickly overshadowed by the consequential speculating on profit statements and periods of spectacular volatility. For years, these episodes have drawn greater numbers of traders out of the woodworks with expectations for quick trades. However, from the chart below, it is hard to draw the conclusions that there is any consistent direction during the bulk of the reportings (typically in the middle of January, April, July and October). On the other hand, there is a consistent – albeit modest – increase in volatility specifically during these periods. This is what would be expected as the earnings season will surprise on both the upside and downside over time. Considering that there is no bias on long-term direction during this period, strategies that focus on an increase in volatility alone (pair trading, potential breakouts, etc) would work best.

Year End Taxes – Another frequently followed seasonal effect is the theory that investors will book losses before the end of the year for tax purposes. Capital losses can be used to offset gains for that same year and may be rolled forward (with some limitations) to balance positive returns in later years. Furthermore, after a portfolio is liquidated and the new year begins, traders and fund managers supposedly reestablish their portfolios and long-term positions. If this assumption held water, there would be a notable dip in average returns through December with a substantial rebound in January. At the same time, volatility would rise through the year end (as selling is more frequently associated with active markets) and taper off after the tax benefits had been procured. However, this supposition runs counter to what the data reflects. Considering that price action does not follow theory in this situation, there is no discernable edge with the age-old adage.

Elections Years – As long as there have been democratic-style politics there have been promises from candidates to fix whatever may be ailing an economy. Whether the economy is facing a recession (whereby nominees will vow to jump start growth) or expansion is already a staple (and candidates will pledge to cut taxes), there seems always to be an upside for the economy when everything is said and done. In the US, the grandstanding happens every four years; but the premise that investors are actually wooed by such ideals is certainly absent from hard data. Curbing enthusiasm for such a reliable truism, it follows that between 1921 and 2007 the average annual return for the Dow Jones Industrial Average was 8.08 percent. When only election years are considered, the average over the same time frame stands at 8.45 percent. However, there may still be some truth in this seasonal effect. Some analysts and traders have suggested that there is a pattern of greater returns in pre-election years. Consequently, returns from 1921 to 2007 for pre-election years average 13.5 percent. Therefore, a long-term buy-and-hold strategy in major equities in the year prior to elections may offer a profitable filter over the time.

Cashing In For The Spring – One of the most commonly held proverbs in the equities markets is ‘sell in May and go away.’ In this, market veterans believe stocks produced the bulk of their positive gains from November through April; so long trades should be liquidated by May Day and reestablished at the beginning of November. Fundamentally, there are few suggestions as to why this phenomenon may exist. Nonetheless, statistics do support just such a historical effect. In fact, not only do the returns through the invested six months account for 80 percent (8.19 percent) of the year’s return on average, there is reduced volatility through the same period in comparison to activity through the late summer. Putting this relatively high probability pattern to work, a macro trader can establish a long-bias in equities from November through April and then switch to a different market through the Spring and Summer months.

In the commodities market, there are no set, market-wide time frames for accounting like there are with stocks. However, there are far more consistent periods where supply and demand are unbalanced leading prices to rise or fall. This makes sense as commodities, unlike other financial securities, are based on physical and often limited resources. When weather seasons change or holidays approach, production and consumption shifts can easily push the market out of equilibrium – and it often does.

The Need For Fuel – Since the industrial age, growth for the world’s largest economies has been dependent on the availability of combustible fuels. Over the past quarter century, light sweet crude has become the benchmark for energy demand and inventories. And, many analysts and traders expect this market proxy to be highly sensitive to the changing weather – particularly a build up in demand for refined fuel ahead of the cold winter season and before summer vacation trips.

Conventional wisdom has it that there is typically a rise in demand for crude oil from March through May, so refiners can process the raw material into gasoline and stockpile it before the official start of the driving season (Memorial day). There are similar expectations for a pick up in the demand ahead of the blustery winter season, which sees greater use of refined heating oil to warm homes. Cross checking these beliefs against the data, however, offers mixed results. The traditional held bullish run up in price before the summer driving season does seen a sharp jump in crude prices in March – with moderate gains through April and May. On the other hand, the supposed increase in demand before the frigid winter months begin in December is not as apparent. Instead, the expected lull in August and September actually sees significant increases to price, while the more pertinent October and November months actually see losses on average. Considering the lack of consistency with this effect, it is best not to force an edge.

Inventory Tax – Another, lesser known seasonal influence that market participants expect on a yearly basis is a broad selling by holders of the physical to avoid inventory taxes. And, while there has been a rise in volatility over the past 20 years on average, the theoretical drop in prices certainly does not follow.

When looking for seasonality influences in the financial markets, the most effective and consistent patterns are borne out of outside influences or inconsistencies in normal market operation. Where does that leave government-issued fixed income? Debt issued by governments is usually deeply liquid and considered essentially risk-free (depending on the country). Nevertheless, there are still commonly held recurrences in trading circles. We will focus on price action from short-term, US Treasury paper to test these theories.

Year-End Portfolio Padding – Similar to the connection between tax season and equities discussed above, there is a common notion that investors will liquidate their more risky assets and reallocate their capital into Treasuries at the year end. This belief originates from the idea that market participants will look to offset yearly profits by booking losses on short-term stock, currency and futures trades – which are taxed at a higher rate. Looking to reduce their tax burden - yet not wanting to fully forfeit returns through the year end - portfolio managers and other large traders will buy government debt to provide a riskless return. Looking at the data below, there is evidence to suggest that this influence may exist from year to year. On average, yields have fallen 0.9 percent through December (prices and yields trade inversely to each other), suggesting buying pressure. However, the level of demand for this specific period is weaker than that seen in other monthly changes. As such there is little opportunity to make a consistent profit in this seasonal effect.

Election Years – Sharing another common bond that the fixed income market shares with equities is suspected, consistent relationship to the election cycle. There are a number of theoretical changes to government spending through the election cycle (shoring up spending before an election, issuing more debt after an election to fund spending projects, etc); but there is little support for any response from the market in response to these shifts. Since 1963, the average change in five-year T-Note yields was 2.0 percent (with significant volatility around the mean). In election years, this average merely rose to 2.98 percent and in the year following the vote it grows to 4.47 percent. This may seem like a meaningful change, but the significant deviation from year to year certainly erodes its statistical relevance.

Seasonal effects are one of the most frequently referred to strategies in the media and educational materials. However, history has shown that where these trading opportunities truly exist, they are usually fraught with volatility and/or the rewards are generally small. This may be due to their never having really existed at all or perhaps it is the natural response to increased interest. Regardless of the reason, it is clear that these trading maxims should not be used alone in strategies, nor should they replace reliable fundamental and technical analysis.