TAOS is essentially a “matrix.” Similarly to the movie Matrix (and with improving AI), one may live in a simulated path of actions that resemble the real world so closely that noticing a significant difference between the simulated and real world events can be quite a challenge. The actions taken by the agents, and their interactions, have “real” effects on each other, but only in the simulated world. That is, in the matrix.

The main reason why this approach is useful in risk management is clear: by running an increasing number of realisations, all of them simultaneously, the average statistics across them will get more and more accurate by the laws of classical statistics. The simulated paths are independent from each other, currently, but they could also be tied together to present a portfolio of assets, say, with realistic correlations.

In mathematical finance, which is used by most quants in the profession of trading and investing, Monte Carlo techniques form the basis of, say, numerical option pricing. By simulating complicated stochastic processes several times, it is possible to find out the underlying risks (and therefore the true value of an option) especially in cases of extreme complexity for which closed-form mathematical formulas do not yet exist.

Since TAOS produces multiple layers of L3 MBO data, and not only the L1 top-level or trade price, the depth of the simulation is greater and therefore also potentially more useful in more cases, especially in modelling of liquidity, market impact, and crashes. The simulations are gradually made more realistic, with endogenous and exogenous shocks. The paths can always be monitored in live at MONITOR link below.