We compute monthly correlation matrices of 25 global futures markets in four asset classes: fixed income, commodities, equities, fx. Comparing and grouping those correlation matrices leads to distinct «regimes» in time. We can characterize these regimes by futures market returns, finding patterns between risk-on and risk-off assets. One of those regimes is especially «risk-off». We can also characterize these regimes by CS hedge fund index returns. In the «risk-off» regime, they also underperform. The Eurekahedge EHF funds show a similar performance behaviour according to strategies across regimes as the CS hedge fund indices. The dispersion across the Eurekahedge EHF funds for each month is largest in the «risk-off» regime.