Traders work on the floor of the New York Stock Exchange in New York City in 2015. (Spencer Platt/Getty Images)

Every week, it seems, we learn more about the consequences of air pollution — and in particular of the smallest airborne particles, known to scientists by the name PM2.5, which are capable of traveling deep into the lungs and entering the bloodstream, and from there, causing havoc.

Not only does air pollution kill millions of people each year — we’re also learning that we can’t necessarily escape it by staying indoors. The tiny particles travel indoors with us. Granted, their toll is clearly the worst on people outside, particularly those who work outdoors, but recent research suggests that tiny PM.25 particles also reduce the productivity of indoor office workers.

In a new working paper published by the National Bureau of Economic Research, three economists demonstrate that poor air quality may even be capable of dinging stock market returns, produced by traders who are very much indoors at the New York Stock Exchange.

Examining the daily returns of the S&P 500 index from January 2000 through November 2014, the researchers showed that on days with bad air quality in lower Manhattan within a mile of the exchange, stock returns tended to be lower.

“A one standard deviation increase in PM2.5 decreases the daily percentage returns by 11.9%, a substantial effect on daily NYSE returns,” the study reports.

Or to put it less technically: “In New York, if you take a day from the cleanest 20 percent of days, and make it one of the dirtiest 20 percent of days, other things being equal, the effect of that on the stock returns, it depresses them by 11 or 12 percent,” explained Anthony Heyes, a professor of economics at the University of Ottawa and the study’s lead author. “Which is a fairly significant number.”

Heyes conducted the study with his University of Ottawa colleague Soodeh Saberian and Matthew Neidell of Columbia University.

The study does not take a strong stand on how this happens. But the researchers think that being subtly irritated by air pollution, and not feeling well, contributes to being more risk averse in one’s behavior. This, in turn, would then lessen investors’ performance.

Exploring this idea, the study shows that air pollution levels in lower Manhattan also had a relationship to the Chicago Board Options Exchange (CBOE) Volatility Index, or VIX, which is a widely accepted indicator of investor fear or worry. The VIX tended to increase when pollution was worse, the study found, such that “a one unit increase in PM2.5 concentration increases the value of VIX by 1.9%.”

“We would say, the movements in VIX that we see are consistent with reduced risk appetite,” said Heyes, noting that the researchers are only suggesting this mechanism, rather than having definitively proved it.

The researchers also performed statistical techniques to control for other extraneous factors like weather, levels of traffic, and the presence of other pollutants than PM2.5 in deriving their findings.

The findings are not all that surprising, notes Tom Chang, a professor at the Marshall School of Business at the University of Southern California, in that prior studies have also shown psychological factors that affect investor behavior, such as whether it’s a sunny day or even who just won the World Cup. Chang has also recently published work on the subtle effects of air pollution on the productivity of indoor office workers — work he co-authored with Neidell — but was not involved in the current study.

“This paper makes a solid contribution to this literature by showing that pollution is an important factor in determining investor sentiment,” Chang said.

Chang also seemed to find the idea about an effect on risk aversion plausible, noting in an email that “when you are feeling sick, you have a lower tolerance for risky investments so you move your money from riskier to safer assets (e.g., stocks to bonds).” He added, though, that “whether increasing risk aversion is a good or bad thing depends on whether risk aversion was too low, too high, or just right before the change.  That is looking back we can probably say that we would be better off if people were a little more risk averse about the housing market a few years ago.”

One surprising thing about the new paper is that it finds that local air pollution in New York appears to have affected the entire S&P, whose component stocks are obviously traded by individuals far beyond the city. But Heyes suggests that top traders are so concentrated in Manhattan that they can still have an impact, and notes that the study performed tests to make sure that the effect was definitely local in nature.

“The effect that we find is very definitely a Manhattan effect, that’s just a statement of empirical reality,” he said.

In the end, it’s another telling finding for the field of behavioral economics, which tends to show that we are far from purely rational economic actors, notes Heyes. “The investors, the actors in the finance sector are actually human beings, they have emotions, they have moods, they have limitations in terms of their cognition and how they behave,” he said.