creates a substantial impact in financial institutionsIt brings unparalleled benefits to several business units
MoCaX already delivers a calculation speed for vanilla trades 5 to 1,000 times faster. The speed of response for an organisation for an accurate price will increase accordingly.
Also, scenario and stress testing, that need to repeat the the same calculation several times, will be possible in a new way.
NEW MARKET FOR EXOTICS
For exotics, the calculation speed increase between 1,000 to 100,000+. Consequently we can now calculate CVA, FVA, or Capital, ultra-fast, for exotic products.
This can give a user an critical edge in the currently stagnant market of complex OTC derivatives.
STABLE PRICE AND CAPITAL
Monte Carlo simulation with 1,000,000 scenarios are now doable with MoCaX. CVA, FVA and capital will not fluctuate when we re-run the calculation.
Regulators require that any risk metric needs to be conservative. Consequently, any pricer in an IMM capital calculation engine needs to be either exact or give a controlled error.
If we do not have a simple analytical pricer for a derivative and we want to put it through the IMM calculation, we need to use fast approximations inside the engine. This creates a massive problem: how can we make sure that the approximated price is conservative, always?
MoCaX Intelligence removes that problem all-in-all. With MoCaX we can achieve pricing machine-precision (0.000000000000001) if needed at ultra-fast speeds.
SPEED & PRECISION
Until now, we lived in a world in which the numerical noise of Monte Carlo simulations was something "we had to live with". We could only decrease it to certain extent.
MoCaX changes the game in that space. Now we can control it. That is because simulations with 1,000,000+ scenarios are now doable, hence the calculation speed and precision becomes now a variable that can be adapted to our requirements.
HIGH CONTROL ON UNCERTAINTY
Given a pricer, there is a clear relationship between the precision that MoCaX can deliver and the number of anchor points (N) we use.
Consequently we can control very well any error we may be inducing in the calculation. Lattice methods lack this control completely.