The Presidential Election Cycle And The Stock Market: A Classic Calendar Anomaly
Calendar anomalies are regular, predictable and exploitable patterns in the stock market that seem not to follow the purportedly “true drivers” of stock prices, such as earnings, or news events – but are linked to some external clock or season or recurring cycle that should seem to have nothing to do with financial or economic fundamentals. They comprise some of the strangest examples of inefficient markets, persistent mispricings, and inexplicably reliable opportunities to beat the index averages. From the classic old chestnut “Sell in May and Go Away” (which refers to the tendency of stocks to go nowhere in the summer months, all the gains clustering between Oct/Nov and May 1) to the “September effect” (September is on average the worst month for the market, by far), these violations of efficient market theory show amazing longevity in time series going back to 1957 or 1926 or 1896. (All those dates pertain to the September effect, by way of example.)
The grand-daddy of calendar anomalies – based on time series going back to 1833 (or even, stretching the point, to 1789)– is the one that holds special interest right now: the bizarre, but oddly sense-making cycle of down-and-up returns based on the 4-year American Presidential Election calendar.
The Presidential Election Calendar Anomaly
Start with the well-established fact that the returns in the second two years of the U.S. Presidential Term far surpass the returns in the first two years.
Year 3 returns are spectacular. Over 13 election cycles (1954-2006), Year 3 averaged gains of 23.5%, without a single down year. (Year 4 of the cycle only had 1 down year.)
For investors, the anomaly offers a rich vein of “alpha” (i.e., above-market returns). Since 1947 (through 2011), Years 1, 2 and 4 of the Presidential cycle returned – cumulatively – about a 300% gain. Years 3 returned 2100% gain.
Election Cycles and the Turn-of-the-Month Anomaly
The pattern is pronounced, and there are many nuances to explore. One study combined the Presidential Cycle anomaly with an even stronger calendar anomaly: the so-called “turn-of-the-month” effect.
Many payment schedules – monthly paychecks, rental receipts, pension plan distributions and contributions, social security payments, income fund distributions – are oriented to the last day or first day of each month (the “turn” of the month). Consumers/investors/retirement-fund managers receive these injections of cash around the beginning of the month. Consumers pay monthly bills, and will often promptly invest the surplus. Fund managers will do the same with the incoming 401-K contributions.
- “The turn-of-the-month construct simply postulates that additional individual or household liquidity will positively impact stock returns at the turn of the month.”
Turn-of-the-month studies typically examine a 4-day period, from the day before the 1st day of the month to three days after. The monthly surge in funds flowing into the market in this window drives prices and returns higher. Combining these two variables — the election cycle and the turn-of-the-month effect – amplifies the difference in market performance between the first two years and the last two years of the Presidential term.
The NASDAQ composite index shows an annualized return of almost 90% for turn-of-the-month days in the last two years of the election cycle (compared to 25% for the same window in the first two years of the cycle).
Why would the last two years of the presidential cycle show such a marked improvement in stock market returns? Several explanations have been proposed.
Vote-buying… um, excuse me, Stimulus measures…
This “explanation” suggests that the party in power attempts to juice up household liquidity in the year or two leading up to the election, to put people in the mood for “more of the same.” It is a fairly common speculation in the academic studies.
In short, the idea here is that government intervention in the economy (stimulus) ramps up in the last two years of the election cycle, causing the stock market to rise.
I find this a rather facile view, based on folk models of the workings of the political system. The evidence is against it. A study of the effect of US elections on foreign markets explicitly examined the possible conflation of variables related to US fiscal or monetary policy, and found that the Election Effect was strong even when controlling for these factors.
A more recent and quite careful study confirmed this result specifically for the U.S. As they phrased the question:
- “If the [Presidential Election Cycle] in stock returns is driven by economic policy, then a cycle should be detected in the government’s tools of fiscal policy. Governmental fiscal policy is carried out through changes in tax law (revenue to the government) and spending policies (expenses of the government)…”
They found that while government economic policies clearly do influence the stock market, it is not sync’ed to the Election Cycle.
- “The presidential election cycle in stock returns and the government’s economic policy influence on stock returns are two separate phenomena.”
In fact, a direct examination of the patterns in fiscal policy finds that
In other words, there is no pattern of systematic “juicing” of the economy or the stock markets by the government leading up to a Presidential election.
Government Activism Depresses the Stock Market
This is the precise inverse of the previous explanation. It proposes that during the first two years of a President’s term, the markets are “depressed.” Why? Because a newly elected (or re-elected) administration assumes (or resumes) power with a mandate for change. Markets hate change. Major policy moves normally take place in the first two years of the presidential term. For example, between 1972 and 2007, major tax legislation – an important expression of policy activism – was much more prevalent in Years 1 and 2 of the Presidential term.
In 8 out of these 9 Presidential terms, there were major tax law changes during the first two years of the cycle. Tax legislation is one of the most uncertainty-inducing forms of government activism. It unsettles the stock market. The new measures make take a while to achieve their full effect – leading to prolonged uncertainty about economic impacts, and depressed market sentiment.
- “Presumably, a president’s new agenda can take a couple of years to work its way through the economy—and might even produce some financial indigestion if said agenda is not found ‘market friendly.’”
Other lines of reasoning and evidence support this view. The findings by some studies (though not all) of apparently positive effects of government gridlock on market performance would support the idea that the opposite of gridlock – a united, activist administration with a fresh political mandate – is not favorable for investors. Similarly, the Congressional Effect, which suggests that markets outperform when Congress is not in session (and can’t pass new laws), points to the same conclusion. (This will be the topic of a subsequent column.)
In short, the better explanation for the Presidential Election cycle in market returns is the diminished policy activism by the government in the 2nd half of the term, as all parties begin to focus more on the next election.
Admittedly it is hard for an investor to know quite how to exploit this anomaly. Getting into the market for Year 3, and maybe Year 4, and staying out for Years 1 and 2 – on paper this generates a spectacular return, but it doesn’t conform with human psychology. We are all entranced by the challenges of investing, the game – as John Maynard Keynes observed long ago:
- “The game of professional investment is intolerably boring and over-exacting to anyone who is entirely exempt from the gambling instinct; whilst he who has it must pay to this propensity the appropriate toll…”
It is not reasonable to ask someone to sit out the action for years at a time just to be able to enjoy “beating the benchmark.” The secret about investing (which everybody knows) is that it is often has as much to do with entertainment as with making money.
Perhaps the lesson is more relevant at the level of asset allocation. If the pattern holds, 2021 and 2022 will not be stellar years for U.S. equities. A “tactical tilt” away from equities in the first two years — especially Year 2 – of the next cycle might be an actionable investment response to the likely repeat of the Presidential Cycle.