Does Weather Affect the Stock Market?

Despite the best efforts of many highly trained economists and market specialists, there is no widespread consensus about how, or even if, weather affects the performance of the stock market.

It seems commonsensical that it must have some impact, since weather is a ubiquitous phenomenon from which traders are never fully isolated. On the other hand, there isn't a clear-cut, logical reason to expect that rain on Wall Street or a hurricane in Mexico should systematically change valuations or trader optimism. Ultimately it's an interesting question, but one that financial economics isn't really equipped to answer.

As a practical matter, it isn't difficult to test the correlation between stock market performance and weather pattern data. Meteorologists and climatologists chart everything from average sunshine to ocean currents, and stock market performance is a matter of public record.

The trick is trying to pick the right data to compare. Peer-reviewed studies have disparate and conflicting results. One recent and famous example was "Weather-Induced Mood, Institutional Investors, and Stock Returns," which came out of Case Western Reserve University in Cleveland in 2014. It found that relatively cloudier days increased perceived overpricing in individual stocks and, subsequently, led to more selling by institutions.

"Stock Returns and The Weather Effect" was published in the Journal of Financial Economics in 1980. It seemed to find a very large impact factor, 3.72, under what was referred to as a "calendar time hypothesis." However, further review found that weather was a much smaller predictive variable than whether or not the trading day was a Monday.

Another study, "Stocks and the Weather: An Exercise in Data Mining or Yet Another Capital Market Anomaly?" appeared in Empirical Economics in 1997. This study attempted to replicate a 1993 study that showed that stock prices were "systematically affected by the weather." The 1997 study could not reject the null hypothesis, ultimately conceding "that no systematic relationship seemed to exist."

The scientific method works wonderfully in physics or chemistry, where independent tests are controlled and variables are isolated, but nobody can run controlled tests on the ecosystem or the global economy. The systems are too large to replicate and too monstrously complex to fully understand. Data has its limits, and the best a market analyst can hope for is to show correlation, not causation.

Most causal models in economics or environmental science are regression-based. Modelers have to identify which factors seem relevant or irrelevant, and they need to have reliable and comparable data on all of the relevant factors. They also need to weight the relevant variables and add controls for possible corruption or biases. Many of these models are sophisticated and mathematically beautiful, but they can never accurately account for every potentiality.

One reasonable theory about weather and Wall Street suggests that severe weather interrupts business processes, supply chains and consumer movements, among other factors. In fact, the financial media often blames a sluggish quarter of gross domestic product (GDP) growth or stock market performance on weather problems. Though a popular idea, not everyone agrees.

One skeptic is Gemma Godfrey, head of investment strategy at Brooks Macdonald, who said that "the markets are insulated" from weather problems. "Markets have priced this in so there has been little downside reaction in the markets ... and less upside room when the weather warms." Many agree with her, arguing that meteorologists are good enough now that markets can anticipate fluctuations well in advance.

One alternative theory, an offshoot of behavioral finance, states that weather clearly affects mood, and mood clearly affects investor behavior. This link appears like a good argument for weather-influenced stock returns, but it's probably not as strong as its proponents make it sound.

For instance, it's not enough to demonstrate that weather affects mood; it must be demonstrated that weather affects mood in ways that alter decision-making about securities transactions (or, alternatively, alters saving and spending habits enough where securities volume is substantially different). Despite several studies in this area, economists don't really have the answers.

One such study, conducted between 2009 and 2011 on the Borsa Istanbul Stock Market in Turkey, found that investor behavior wasn't impacted by sunny days, overcast days or sunshine duration, but that it was probably affected by "the level of cloudiness and temperature."

A different U.C. Berkeley study, published in the Undergraduate Economic Review in 2011, concluded that "sunshine affects mood and mood can shape behavior" and found a "significant relationship" between sunshine and stock prices over the preceding half-century.

One study finds no effect from sunny days in Turkey, but a competing study argues that sunshine affects Wall Street performance. It's theoretically possible that sunshine affects Turkish traders differently than New Yorkers, but the far more reasonable conclusion is that model-based regression economics isn't really prepared to handle such an intricate causal relationship.