The feature list for any analytics system runs to many pages of detail. This list covers the must haves without drowning you in detail.
Product coverage is extensive, covering all the major markets; interest, foreign exchange, credit, commodities and equities. Product definitions are flexible, allowing different representations of a single product. Product field and cashflow documentation is linked to the validation of incoming trade data so it is always up to date.
With the Microsoft HPC architecture workload is distributed across multiple CPUs on multiple computers. Multi-threading within the analytics engine also allows a single calculation to be spread across multiple processor cores. However this is only part of the story. The most important technical feature is vectorisation. Please see the highlighted section to find out what this is because its importance cannot be understated.
ICE Data Services has struck an agreement with Vector Risk to resell their rates for use in our analytics service. A customer can request access to these rate sources for an add-on to their base subscription. Flexible rate source prioritisation allows a customer to mix their own rates and vendor rates in any fashion.
The simulation methodologies historic VaR and Monte Carlo are fully featured. For example you can specify attributions for VAR (breakdown by market segment), proxying of rates, confidence interval and any number of roll-ups up to global VaR. For credit you can see the effect of netting and collateral, track path dependent events such as option knock-outs, and apply correlations and stochastic processes appropriate to the evolution of different market rates.
CVA has brought about two important enhancements; risk neutral curve evolution and CVA sensitivities. The sensitivities are calculated in the Monte Carlo by running extra scenarios to shift each risk factor up and down. This is a very intensive calculation but Vector Risk's HPC architecture makes it possible. CVA, DVA and bilateral CVA are all supported.
Stress Testing is richly supported. Need to run historic scenarios on groups of curves, together or separately, parallel or non-parallel shifts, absolute or relative by industry sector or liquidity ranking? It's all there, and the same stresses can be reused for market, credit or liquidity calculations.
Daily Workflow is the second major component of the cloud solution after the analytics engine. Here the user can construct task dependencies (data loads, followed by valuation and reconciliation, followed by risk calculations, and finally the publishing of daily reports). Typically a daily process refreshes all the analytic results, and subsequent intra-day real-time calculations are routed directly to the analytics engine.
Parameter Estimation is part of the package. Apart from collecting raw rates you can automate the generation and update of curve evolution parameters and correlations used in the Monte Carlo.
Drilldown is fully supported. This means you can re-run a calculation and configure the system to capture and display any rate movements or trade mark to markets on any scenarios and time steps you wish.


Vectorisation gives an 80 times speed-up over normal code. How? Read on...

Simulations involve valuing the same trade across hundreds or thousands of potential future rate scenarios. Normal code would do this by performing the valuation one scenario at a time. For complex valuation formulas this means continuously swapping bits of the pricing code in and out of the CPU with little real processing going on. Vectorised code treats the rates across all the scenarios as an array so each piece of the valuation code stays in the CPU while a “pipeline” of rates gets passed through it. The upshot of this is that there is much less swapping of code in the CPU leading to much more useful throughput.

Vector Risk has developed a set of vector template functions that underpins its pricing library and analytics engine. The unique feature of these template functions is that they are non-invasive. The pricing code looks like normal code and can be supported or built on without the need to attain esoteric coding skills, however these templates allow the thousands of scenarios in a Monte Carlo or historic simulation to be processed in one pass of the pricing code, instead of looping through the scenarios one at a time.