1st International Congress on Environmental Modelling and Software - Lugano, Switzerland - June 2002
Keywords
patents, trends, volatility, garch, gjr, egarch, asymmetry, regularity conditions, asymptotic theory, international rankings, non-nested tests
Start Date
1-7-2002 12:00 AM
Abstract
This paper analyses trends and volatilities in registered patents for the top 12 foreign patenting countries inthe USA. In terms of total US patents by foreign countries, Japan is ranked first, followed distantly by Germany andthen France. Patent registrations from each of the 12 countries have increased steadily over time, but at different rates.Using monthly time series data for 1975-1998, the time-varying volatility of the patents ratio, namely US patentsregistered by each of the top 12 foreign countries relative to total US patents, is examined in detail. Internationalrankings based on both the number of patents and patent intensity (or patents per capita) are provided. The asymmetricAR(1)-EGARCH(1,1) model is found to be suitable for most countries, while the AR(1)-GARCH(1,1) and AR(1)-GJR(1,1) models also provide useful insights. Non-nested testing procedures are used to discriminate betweenGARCH(1,1) and EGARCH(1,1), and between GJR(1,1) and EGARCH(1,1).
An International Comparison of Foreign Patents Registered in the USA
This paper analyses trends and volatilities in registered patents for the top 12 foreign patenting countries inthe USA. In terms of total US patents by foreign countries, Japan is ranked first, followed distantly by Germany andthen France. Patent registrations from each of the 12 countries have increased steadily over time, but at different rates.Using monthly time series data for 1975-1998, the time-varying volatility of the patents ratio, namely US patentsregistered by each of the top 12 foreign countries relative to total US patents, is examined in detail. Internationalrankings based on both the number of patents and patent intensity (or patents per capita) are provided. The asymmetricAR(1)-EGARCH(1,1) model is found to be suitable for most countries, while the AR(1)-GARCH(1,1) and AR(1)-GJR(1,1) models also provide useful insights. Non-nested testing procedures are used to discriminate betweenGARCH(1,1) and EGARCH(1,1), and between GJR(1,1) and EGARCH(1,1).
