Financial crises have a peculiar way of exposing connections that investors never knew existed. One day a pandemic in Wuhan or a war in Eastern Europe seems like a distant problem, and the next, trading desks in New York, Frankfurt, Hong Kong and Tokyo are all moving in lockstep as if the world’s stock exchanges were wired together by a single electrical circuit. Economists have long acknowledged that equity markets are interconnected, but the deeper question — what actually drives those connections, and whether they strengthen or weaken as monetary policy, uncertainty and global crises evolve — has remained stubbornly difficult to pin down. A new study published in the Atlantic Economic Journal by Wan-Shin Mo of National Chiayi University in Taiwan and Shun-Chuan Chuang of Chung Yuan Christian University offers one of the most detailed statistical portraits to date of how return shocks travel across nine developed stock markets between 2017 and 2023, a window that happens to capture three of the most turbulent episodes in modern financial history: the pandemic-era crash and recovery, the tightening cycle of the United States Federal Reserve, and the outbreak of the Russia–Ukraine war.

At the heart of the research lies a now-standard but powerful econometric toolkit: the spillover framework developed by Francis Diebold and Kamil Yilmaz, applied within a rolling-window generalized vector autoregression, or VAR. In essence, the method treats each stock index as a node in a network and asks a deceptively simple question: when a shock hits one market’s returns, how much of that market’s future forecast-error variance can be explained by shocks originating in other markets? The generalized VAR structure matters because it does not require the analyst to make arbitrary assumptions about which market came “first” — it computes the variance decompositions in a way that is invariant to the ordering of the variables, a crucial feature when markets in Asia, Europe and North America close and open at different hours of the day. By sliding a rolling window across the sample, Mo and Chuang were able to track, day by day, how the total connectedness of the system rose and fell, which markets were net exporters of shocks and which were net importers, and how these roles shifted when conditions changed.

The nine markets in the sample — the United States’ S&P 500, Canada’s S&P/TSX Composite, Hong Kong’s Hang Seng, South Korea’s KOSPI, Singapore’s Straits Times, France’s CAC 40, Germany’s DAX, the United Kingdom’s FTSE 100 and Japan’s Nikkei 225 — were chosen as representative benchmarks of developed economies. Using daily benchmark index returns, the researchers decomposed the forecast-error variance of each market into contributions from itself and from the other eight, then aggregated these contributions into a “total spillover” index measuring overall interconnectedness, and “net directional” measures identifying whether each individual market was, on balance, a transmitter or a receiver of shocks. The results paint a clear and geographically structured picture. North American and European markets consistently acted as net transmitters of return shocks, radiating influence outward to the rest of the system, while the Asian markets in the sample — Hong Kong, South Korea, Japan and Singapore — behaved as net receivers, absorbing more influence from abroad than they exported.

The time dimension of the analysis is where the story becomes genuinely dramatic. Total spillovers across the nine markets surged to extraordinary heights during the COVID-19 pandemic, when lockdowns, collapsing earnings expectations and a chaotic “dash for cash” in March 2020 synchronized markets with brutal efficiency. The study then documents a second wave of elevated connectedness around the onset of the Russia–Ukraine war in early 2022, when energy price shocks, sanctions and renewed geopolitical risk again caused markets to move together. These episodes confirm what many practitioners have suspected anecdotally: crises do not merely damage markets individually, they weld them together, transforming a set of semi-independent national exchanges into a single highly coupled system precisely at the moment when diversification is needed most.

But Mo and Chuang went beyond simply describing when spillovers spiked — they asked what caused them. In a second stage of analysis, the researchers ran determinant regressions linking the rolling total spillover index to a set of macroeconomic and policy variables, using lagged regressors to mitigate potential endogeneity problems, with contemporaneous specifications as a robustness check yielding qualitatively similar results. The findings are strikingly consistent: total spillovers increased when the United States monetary policy stance was tighter, when United States policy uncertainty was higher, and during the COVID-19 pandemic itself. This pattern aligns with a growing body of literature on the “global financial cycle,” which argues that US monetary conditions propagate through global risk appetite, capital flows and exchange rates to influence asset prices far beyond America’s borders. When the Fed tightens, funding becomes scarcer, investors de-risk, and the resulting portfolio adjustments ripple outward — showing up in the study’s data as a measurable rise in cross-market connectedness.

The researchers also examined how the direction of spillovers — who gives and who receives — depends on the environment. Net directional spillovers varied with pandemic severity, measured through new COVID-19 infection cases, and with the prevailing interest-rate environment. To handle zeros in the infection data caused by reporting irregularities, the authors applied a transformation recommended by Chen and Roth, and for net spillover series that could be non-positive, they used an inverse hyperbolic sine transformation. The finding that net transmission roles shift with these variables is consistent with a mechanism of cross-border portfolio rebalancing: as interest-rate differentials and risk conditions change, international investors reallocate capital across markets, altering which exchanges serve as sources of shocks and which act as sinks. In a further analysis sorted by central bank policy rates, the US, Canadian, Hong Kong, South Korean and Singaporean markets belonged to a “high-yield” group, while the French, German, British and Japanese markets belonged to a “low-yield” group — a distinction that helps explain how rate environments condition the geography of transmission.

The paper’s emphasis on interest-rate environments speaks to one of the most consequential policy debates of the past decade. If tight US monetary policy systematically raises spillover intensity across global equity markets, then the tightening cycle that began in 2022 did not simply raise discount rates everywhere — it also made the entire global system more fragile, more correlated, and more vulnerable to the next shock. For portfolio managers, the implication is that hedging strategies calibrated during calm periods may fail precisely when they are needed, because diversification benefits evaporate as connectedness rises. For regulators and central banks, the study adds empirical weight to the argument that macroprudential monitoring should track cross-border connectedness measures in real time, treating spikes in spillover indices as an early-warning signal of systemic stress rather than waiting for outright market crashes to reveal the wiring.

Methodologically, the study’s rolling-window generalized VAR approach offers a middle path between static analyses, which average over entire decades and obscure regime changes, and event-study designs that focus narrowly on a single crisis. By measuring connectedness continuously through time, Mo and Chuang capture the fluid character of market integration: it is not a fixed parameter but a state variable, responsive to pandemic severity, policy uncertainty and the stance of the world’s most influential central bank. The careful handling of econometric pitfalls — the generalized decomposition that sidesteps ordering assumptions, the lagged regressors that guard against reverse causality, and the nonlinear transformations that accommodate zeros and negative values — makes the results unusually robust for a field where measurement choices often drive conclusions.

For a general audience, the takeaway is both sobering and clarifying. The world’s developed stock markets are not merely traded side by side; they form a living network whose density is set by decisions made in Washington, by the trajectory of a virus, and by the outbreak of wars. When the system tightens, Asia’s major exchanges tend to inherit turbulence that originates in the West, while North American and European markets remain the dominant exporters of shocks. When the system relaxes, those flows attenuate, though they never disappear entirely. The study’s period ends in 2023, but its findings carry obvious relevance to the present, as markets continue to digest the aftereffects of rapid rate hikes, geopolitical fragmentation and lingering pandemic-era distortions. Investors who want to understand tomorrow’s crashes, the research suggests, should spend less time watching individual national exchanges and more time monitoring the invisible wiring between them — because that wiring, as this study demonstrates, is never static, and it is almost always tightening precisely when it matters most.

The work of Mo and Chuang ultimately transforms a vague intuition — that markets are connected — into a precise, time-stamped, testable account of when, where and why those connections intensify. By tying the ebb and flow of spillovers to measurable macroeconomic forces such as monetary policy stances and policy uncertainty, the study moves the field beyond crisis narratives and toward a genuine predictive framework for systemic risk. In an era when a single Federal Reserve meeting or geopolitical tremor can cascade across continents within hours, that kind of understanding is not academic luxury. It is a map of the fault lines beneath the global financial system — and, increasingly, it is the map that investors, regulators and policymakers will need to consult as the next shock inevitably arrives.