Contagion Risk

Financial contagion differs from fundamental comovement: it is the "excess" correlation observed during crises above what economic linkages would predict. A country with no direct trade or financial ties to Greece should not, in theory, see its bond spreads rise when Greece's spreads widen — but during 2011-2012, exactly this happened across Southern Europe. This excess correlation is contagion: the transmission of fear, the repricing of sovereign risk as a category, and the flight-to-quality dynamics that treat all peripheral assets as homogenous.

Three contagion channels are empirically well-documented: (1) Trade channel — a crisis in an export market reduces revenues and external accounts; (2) Financial channel — bank exposures, cross-holdings, and funding dependencies mean that losses at one node force deleveraging at connected nodes; (3) Confidence/wake-up call channel — investors reassess all similar-profile countries when one fails, causing correlated sell-offs even without direct links. The latter channel is the hardest to model mathematically and produces the most surprising contagion events.

The Noosphere contagion model uses network theory to represent country-country financial linkages and the Hawkes process to model the self-exciting nature of contagion events. The R₀_financial coefficient — analogous to the epidemiological basic reproduction number — measures how many secondary country crises are expected for each primary crisis, given the current network topology. When R₀ > 1, a crisis anywhere in the network becomes a global systemic event.