Forecasting Consumer Price Index of USA

Samantha Liu, Xiang, Han, Yongliang "Stanley", Review of Business Research

ABSTRACT

In this study, we find that the best fitting model among a collection of AR, MA, ARMA, ARIMA models for annual CPI data from 1913 to 2006 is ARIMA (2, 2, 1). Based on this model, we forecast the annual CPI value for 2007 to be 207.1909, with a 95% confidence interval between 205.1212 and 209.2607. We predict that inflation will increase in 2007 since the confidence interval of the forecast suggests a consistent increase in annual CPI during 2007. This finding will provide useful information for the Fed and financial and economic analysts who are concerned about the economy. We can infer from the increase of CPI that the Fed may take necessary actions to contract the economy in 2007.

Keywords: Consumer Price Index, Forecasting, United States Economy

1. INTRODUCTION

The purpose of this study is to forecast the Consumer Price Index (CPI) of the United States in 2007. Using historical annual CPI data from 1913 to 2006, we search for a time series model to predict the annual CPI of 2007. The CPI is the most widely used measure of inflation in financial analysis. It is one of the leading indicators of economic trends, which provides a preview of what is going to happen before the change actually occurs. The Federal Reserve Bank (the Fed) uses the CPI to formulate monetary policy to facilitate US economic growth (Fullwiler and Allen, 2007). In addition, the CPI is used to adjust annual changes to social security payments. This study is important because an accurate forecast of the CPI will provide information about inflation, help policy makers to set appropriate monetary policy, and assist people to predict the impact of inflation on financial markets.

In the next section we briefly introduce the role of the CPI in economic analysis. In the following section, we describe the four basic time series models and the model selection criteria. Finally, we forecast annual CPI in 2007 using the best fitting model and evaluate the accuracy of our forecast.

2. BACKGROUND INFORMATION: THE CONSUMER PRICE INDEX

The Bureau of Labor Statistics in the US publishes the CPI at 8:30 am EST around the 15th of each month. The CPI measures the "changes in the prices paid by urban consumers for a representative basket of goods and services" according to the Bureau of Labor Statistics. The basket includes about 200 types of goods and thousands of actual products including housing 42%, food 18%, transportation 17%, medical care 6%, apparel 6%, entertainment 4%, and others 7%. The prices are sampled at different stores. In each period, prices for a fixed-list of goods and services are compared to a base period. Currently, the base period is the 1982-1984 period, during which the CPI is set to equal 100. All other values of the CPI are a percentage relative to the value during the base period.

The CPI, proxying for inflation, has been widely used as a leading indicator of economic change. Financial markets continuously assess expectations on the CPI and react to the innovations contained in new published data (Espasa, Poncela and Senra, 2002). Inflation is most likely to affect interest rates, stock prices and exchange rates. Unexpected inflation causes bond prices to drop and yields to rise. An increasing interest rate will negatively affect stock prices. Unexpected inflation also decreases the value of a country's currency in the global market and impacts the exchange rate. The Fed will decrease money supply in view of high inflation in order to boost up interest rate. Thus, the Fed has been responding to the negative effect that inflation has on the economy by changing monetary policy. As a result, a key issue for many economists and policy makers is to predict and analyze inflation and take corresponding actions.

3. BACKGROUND INFORMATION: THE TIME SERIES MODELS

3.1. Time Series Models

There are four basic time series that may describe the behavior of the dataset (Enders, 2004), namely Autoregressive (AR), Moving Average (MA), Autoregressive Moving Average (ARMA) and …

Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Forecasting Consumer Price Index of USA. Contributors: Samantha Liu, Xiang - Author, Han, Yongliang "Stanley" - Author. Journal title: Review of Business Research. Volume: 7. Issue: 1 Publication date: January 2007. Page number: 123+. © 2008 International Academy of Business and Economics. COPYRIGHT 2007 Gale Group.

This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.