No Cover Image

Conference Paper/Proceeding/Abstract 1020 views

Forecasting Branded and Generic Pharmaceutical Life Cycles when the the Branded drug is outperforming the Generic

Sam Buxton Orcid Logo, Kostas Nikolopoulos, Marwan Khammash, Philip Stern

Swansea University Author: Sam Buxton Orcid Logo

Abstract

Branded and generic pharmaceuticals have until recently competed in virtually distinct worlds (Bernard, 2011). This world is separated by the patent protection of branded products. Companies were focused on a single product type. There was also a wide discrepancy in the prices of branded pharmaceuti...

Full description

Published: 2016
URI: https://cronfa.swan.ac.uk/Record/cronfa43654
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract: Branded and generic pharmaceuticals have until recently competed in virtually distinct worlds (Bernard, 2011). This world is separated by the patent protection of branded products. Companies were focused on a single product type. There was also a wide discrepancy in the prices of branded pharmaceuticals as compared with their generic counterparts. Over the past decade these worlds have collided creating a new environment. Bernard (2011) terms this “braneric competition”. There are three factors that have activated this union competitive duration, corporate convergence and commercial hybridisation. The need to plan for and forecast this market is a key opportunity that needs to investigated and invested in. Comparing the life cycles of branded and generic drugs that are no longer patent protected, it can be seen that a new pharmaceutical life cycle has started to emerge and is inextricably associated with the new environment. This research shows now that branded drugs even many years after patent expiration are still actively competing with and often outperforming the generic equivalents. This research uses prescription level data from GP’s in the UK to speculate why this may be the case and investigate whether these life cycles can be forecasted. The research is being undertaken in two stages. Stage 1 saw the Bass diffusion model, repeat purchase diffusion model (RPDM), moving average, exponential smoothing, Naïve and Naïve with trend models applied to the data. There is competing literature suggesting that complex models are better able to forecast pharmaceutical life cycles. While other research supports the conclusion that simple models are often better. When stage 1 had been completed it was noted that none of the complex models were able to outperform the simple models. The second stage will see the Holt winters exponential smoothing model, auto-regressive integrated moving average (ARIMA) model, robust regression, and regression over t, regression over t-1 all applied to the data. The empirical evidence linked to stage 1 shows that for the branded life cycle the Naïve with 50% trend would perform the best. For the generic equivalents the empirical evidence suggests that the Naïve model with the addition of a 10% trend would provide the most accurate and robust modelling and forecasting method.
Keywords: : Forecasting; Diffusion Models; Pharmaceutical Life cycles; Branded drugs; Generic drugs.
College: Faculty of Humanities and Social Sciences