What's behind next generation portfolio analytics
Using High Frequency (HF) data dramatically improves precision of statistical estimates due to the so-called bias-variance tradeoff. HF data provides many more recent/fresh data points, thus decreasing the variance of estimates, without using stale data points that would increase the estimation bias.
PortfolioEffect is a cloud service that runs on specialized HPC grade hardware. All portfolio computations stay in the cloud to offload client systems when processing large volumes of intraday data. We use FAST protocol, adopted by exchanges such as NYSE, Nasdaq OMX & BATS for high throughput low-latency messaging.
Market microstructure model pipeline
Portfolioeffect features a next generation model pipeline for high frequency data. Price feeds are processed with a series of auto-calibrating models for market microstructure effect, price jumps/outliers, fat distribution tails, long memory (fractality) and intraday risk factors.