The Pricepoint statistical asset valuation framework is based on recursive filtering algorithms widely used in many engineering applications where the issue is to extract a signal from a noisy background. Such situations occur e.g. in location tracking, mobile telephony, computer animation, voice recognition and navigation, to name but a few. Invector’s Pricepoint systems address the problem of isolating a price signal from noisy trading data. In the markets for heterogenous assets like residential housing or pre-owned personal vehicles, the asset attributes that drive prices have to be identified and assessed and the inherent price uncertainty and market variability arising from asymmetric information and macro-economic cyclicality must be efficiently addressed to minimize prediction error.
Invector’s estimation algorithms operate in real time. Whenever one or more transaction prices are transmitted to the system, they are compared to the current estimate, prediction errors are recorded and the predicted sales price of every asset in the relevant market universe is optimally updated to the next period’s expected value. Simultaneously, the updating process yields new estimate of the error co-variance matrix, and in particular the standard error of each value. The co-variance matrix evolves over time to reflect market variability and inherent price uncertainty at each point in time and the standard error of each valuation is a valuable measure of its accuracy and single-asset price risk. The availablility of these measures is a unique feature of Invector’s valuation technology, and can be systematically leveraged in many important management tasks. The mathematics behind Invector’s estimation technology is relatively simple and has been known since the early sixties, contributing to a number of engineering feats, notably the Appollo 11 moon landing in 1969. The development of a framework to efficiently apply these principles in low-frequency trading environments has required a number of years. With the Invector Pricepoint technology, it holds the promise of significant efficiency gains, as estimates of collateral asset value and volatility become an essential and productive component of modern credit risk management processes.