The CSSC Group, L.L.C. has programmed a mathematical concept called the Bounded Influence Estimator (BIE) to assist in the reporting of statistical data results. All data was returned with the appropriately-calculated BIE, mean and median.
The BIE was offered to the user in three forms:
-
-
- As on on-line calculator for the calculation of one set of variables,
- As a batch calculator for the calculation of many sets of variables and included in a custom software application,
- As a Web Service which accepts data in almost any form over the Web, calculates and returns results to the client.
-
What is a BIE?
The BIE limits the influence of a small subset of the data. Its robustness against outliers and model misspecification is examined and supported. We further use BIE to develop a method for detection of additive outliers.
The BIE can be applied to:
-
-
- Currency exchange rates *,
- Data collection results that are sensitive to outliers that present a heavy tail due to outliers,
- “Just-in-Time” inventory planning,
- Other evaluations of data where anomalous outliers could skew statistical results,
- Survey/Questionnaire results.
-
* Knowledge of the distribution of exchange rates has important implications for theories of international finance and their applications. It is also of importance to varying issues related to foreign exchange. For instance, options pricing on foreign currencies relies on the right specification of the stochastic processes of exchange rates. In testing exchange market efficiency, the information of the statistical properties of exchange rate distribution is essential. Moreover, the volatility of exchange rates itself is a major risk component in international investing. Hence, clear understanding of the behavior and variance of exchange rates is important both for portfolio selection and for the evaluation of the performance of international asset portfolios. Empirical evidence on the distribution of exchange rates, however, has been far from conclusive. While most previous studies have recognized that the rate of change in a foreign currency is not normally distributed, there is a lack of consensus on what type of distribution is most appropriate for describing the behavior of exchange rates.
Click the blue button to learn more about the BIE.