Lifang Peng, Ning Li, and Jingwen Pan (2019) test whether the monthly planetary Ap index, used as a proxy for disturbed geomagnetic days, is associated with S&P 500 returns over January 1998–December 2017.
High AP Index, low stock market performance—really?
The
working hypothesis is behavioral rather than mechanical: geomagnetic
disturbances have been linked to poorer sleep, mood, and health, which
can raise risk aversion and reduce demand for equities. The authors
therefore treat Ap as an environmental mood shock and ask whether that
shock shows up in US index returns after standard market controls. Data
come from NOAA for Ap and from a standard US equity series for the
S&P 500; the analysis is monthly, not daily, which reduces noise but
also blunts short-horizon storm effects.
The core result is a statistically significant inverse relationship: higher monthly Ap is associated with lower S&P 500 returns. Granger-style tests and regressions with year dummies support a directional link from geomagnetic activity to subsequent market performance rather than the reverse. The economic magnitude is small. The authors describe an adverse-signal probability on the order of 1.2 per thousand, so Ap is a detectable but not dominant driver. Liquidity matters: the negative Ap effect is stronger when market liquidity is higher, which they interpret as faster transmission of mood-driven order flow through more liquid books. This is the paper's most distinctive empirical claim relative to the earlier weekly-storm literature of Krivelyova and Robotti.
A second distinctive finding is seasonal. Long-term Ap has well-known spring and autumn peaks (the Russell–McPherron / equinoctial pattern). Prior work often attributed geomagnetic–return links to that semiannual cycle. Peng et al. argue that the Ap–S&P 500 association in their sample is not mainly an artifact of those equinoctial maxima. After allowing for seasonal structure, the inverse relationship still appears, which they take as evidence of a more general geomagnetic channel rather than a calendar-season dummy in disguise.
The contribution is therefore modest but clear: it brings a standard space-weather index (monthly Ap) into an econometric study of US equity returns, documents a negative and liquidity-amplified association, and rejects a purely semiannual explanation. It does not establish a tight causal mechanism, does not use daily or storm-threshold specifications, and does not claim usable short-horizon forecast skill.
The core result is a statistically significant inverse relationship: higher monthly Ap is associated with lower S&P 500 returns. Granger-style tests and regressions with year dummies support a directional link from geomagnetic activity to subsequent market performance rather than the reverse. The economic magnitude is small. The authors describe an adverse-signal probability on the order of 1.2 per thousand, so Ap is a detectable but not dominant driver. Liquidity matters: the negative Ap effect is stronger when market liquidity is higher, which they interpret as faster transmission of mood-driven order flow through more liquid books. This is the paper's most distinctive empirical claim relative to the earlier weekly-storm literature of Krivelyova and Robotti.
A second distinctive finding is seasonal. Long-term Ap has well-known spring and autumn peaks (the Russell–McPherron / equinoctial pattern). Prior work often attributed geomagnetic–return links to that semiannual cycle. Peng et al. argue that the Ap–S&P 500 association in their sample is not mainly an artifact of those equinoctial maxima. After allowing for seasonal structure, the inverse relationship still appears, which they take as evidence of a more general geomagnetic channel rather than a calendar-season dummy in disguise.
The contribution is therefore modest but clear: it brings a standard space-weather index (monthly Ap) into an econometric study of US equity returns, documents a negative and liquidity-amplified association, and rejects a purely semiannual explanation. It does not establish a tight causal mechanism, does not use daily or storm-threshold specifications, and does not claim usable short-horizon forecast skill.
The semiannual variation of geomagnetic activity is linked to the interaction between the solar wind and Earth's tilted magnetic field, which typically causes increased geomagnetic disturbances around the equinoxes and lower activity around the solstices.