The Consumer Price Index (CPI) is released on a monthly basis by the Central Bureau of Statistics, at a lag of 15 days after the end of the indexed month. Toward the end of each month, the Bank of Israel publishes a forecast of the CPI for analysis purposes. Although its weight in the CPI is low, the fruit and vegetables component is extremely volatile and therefore potentially makes a significant contribution to errors in the CPI forecast. In recent years the Bank of Israel has begun collecting data on the prices of products sold in the outlets of Israel’s major food retailers. This information is reported by law1 , is transmitted from the major retailers’ special websites, and is entered into a unique database created by the Bank known as “the Retail Price Database.” This study presents a forecast of the fruit and vegetables component of the CPI based on the prices that were entered into the Retail Price Database. A combination of this forecast and other forecasts that currently exist at the Bank of Israel improved predictive quality by 23 percent in the tested period. This project, which uses Big Data to improve existing forecasting methods, is the first of its kind at the Bank and constitutes a pilot for more extensive use of Big Data to support policy making.

​To the full PDF file​
​