Crisis simulation
A live demo of my open-source regime-mc library: a Markov regime-switching Monte-Carlo over a two-asset book — a broad equity index and long-duration bonds — that flips between a calm regime and a crisis regime. The point it makes is one a single historical covariance can't: when you most need the hedge, correlations can break. These are generic illustrative asset classes and parameters — the technique, not specific tickers or a calibrated scenario set.
Assets: a broad equity index + long-duration bonds (two illustrative asset classes, not specific tickers)
The headline knob is crisis correlation: negative = bonds rally while stocks crash (the hedge works); positive = both fall together (the hedge fails). Run it at −0.3, then crank it toward +0.8 and watch the tail.
Equity weight %
rest in long bonds
Crisis correlation
−0.9 hedge … +0.9 fail
Crisis frequency %/mo
chance a crisis starts
Crisis severity ×
depth of the equity drop
Horizon (months)
e.g. 60 = 5 years
Paths
simulations (200–10k)
Seed
Set the dials and hit “Run simulation”.
Educational tool — not investment advice. A generic two-regime model on illustrative parameters; results depend entirely on those assumptions and do not predict real-world outcomes.