A Markov-chain market model describes a market as moving between a small number of named states, for example calm, normal and stressed, where the chance of moving to each state next period depends only on the current state. It is a simple stochastic process used in risk work and token-economy simulations to capture long spells of quiet or turbulent conditions.
The switching probabilities are set out in a transition matrix, with one row per current state whose entries sum to one. If the calm row reads 0.90, 0.08 and 0.02, a calm month is followed by another calm month 90% of the time, a normal month 8% of the time and a stressed month 2% of the time. Each state can be given its own volatility, demand or liquidity. Running the chain many times in a Monte Carlo simulation produces paths for stress testing. The main weakness is that the probabilities are estimated from past data and may not hold.
Ref 1100001414