Title:
A Vetting of the Acuity of the Bloomberg One-period-ahead Forecasts

Authors:
Moncef Belhadjali, David Leon, Edward Lusk, Germany

Abstract:

Context One of the progeny of the Big Data Analytics [BDA] circa 2017 is the recent emergence of a commercial sector offering: SQL: Relational Database [SQL:RDb] Data-Analytics & Artificial Intelligence modular driven platforms that can be used by their clients as “Dedicated Out-source Links”. Most of these SQL:RDb firms offer Forecasting Platforms where clients can pass their datasets through 25-30 Forecasting Models [FM]s.

Focus As there are often Quantity v. Quality issues with this forecasting-barrage tack, we offer a simple and effective vetting-check. We suggest using a protocol where the Bloomberg One-Period Ahead forecasts are used to create a simple but effective benchmark for deciding on forecasting acuity.

Results For the measure of forecasting acuity, we selected the Absolute Percentage Error [APE] for the One Period Ahead forecasts.  For the Vetting-phase, as expected, usually the [Random Walk FM] outperformed the OLSR-Forecasts. When this was the case, we moved to the Benchmarking-phase, where the Bloomberg forecasts were, in the main, either inferentially more effective or equally effective compared to those of the RW FM for the [APE%]-measure. Conclusion We suggest that the SQL:RDb firms offer Forecasting Platforms that are also linked to the BBTs-platforms. Using the BBTs as an initial forecast profiler may simplify the forecasting process without compromising its acuity.

DOI: https://doi.org/10.51505/IJEBMR.2026.10705