Carlo Altavilla and Domenico Giannone, "The Effectiveness of
Non-Standard Monetary Policy Measures: Evidence from Survey Data",
Journal of Applied Econometrics, Vol. 32, No. 5, 2017, pp. 952-964.

There are two sets of data files, which contain the same information.
The spreadsheets (.xls and .xlsx) are zipped in ag-xls.zip, and the
ASCII text files (in CSV format) are zipped in ag-data.zip. Unix/Linux
users should use "unzip -a" when unzipping ag-data.zip.

Treasury_Corporate_yield:

  Contains the following variables:
	
-- Date: date of reference for yields
-- AAA Corporate Bond Yields
-- BAA Corporate Bond Yields
-- 10-year Treasury Bond Rate

Files that begin with Individual_ :

These files give the breakdown of individual forecasts of the
forecasters involved in the Survey of Professional Forecasters.

Sheets and variable of reference (X): 
	BAABOND - BAA corporate bond yields
	BOND - AAA corporate bond yields
	CPI - Consumer Price Index
	PGDP - Price Index for Gross National Product
	RGDP - Real Gross Domestic Product
	TBILL - 3-Month Treasury Bill Rate
	TBOND - 10-Year Treasury Bond Rate
	UNEMP - Unemployment rate

How to read sheet "Individual_X"

Variables:

-- YEAR: Year of reference for the forecast

-- QUARTER: Quarter of reference for the forecast

-- ID: Identifies different surveyed forecasters

-- INDUSTRY: Industry the forecaster belongs to. Here "1" indicates
the forecaster works in a firm classified as "financial services", "2"
indicates the worker works in a firm classified as "non-financial
services" and "3" indicates that the forecaster works in an "unknown"
industry.

-- X# (where X is as the latter part of the name of the sheet):
This is the #-quarter ahead forecast for variable X (e.g. RGDP3
indicates the 3-quarter ahead forecast for Real GDP)

-- XA (where X is as the name of the sheet): This is the annual
average of variable X in the current year. Similarly for X followed by
other letters, these indicate the annual avarage of the year indicated
by the horizon (mapping from the alphabet to the integers, i.e. A-0
(present), B-1 (next year), C-2 (two years from now), etc.) Note:
there are many #N/A indicating that that entry is missing (i.e. not
available).

SPF_deadline:

Sheet identifies the SPF forecasters' deadline to submit their forecasts. 

Variables:

-- Year of Reference: Year of reference for the forecast

-- Quarter of Reference: Quarter of reference for the forecast

-- Deadline: Deadline for forecasters' submissions

-- Deadline (excel date): Deadline for forecasters' submission using
excel dating convention

Google:

The sheet shows the count of google trends (normalized) for certain
words.

-- First Column: this column indicates the dates (with a weekly
frequency) of when the Google Trends count refers to

-- "quantitative easing", "operation twist", "forward guidance",
"tapering". These columns indicate the Google Trends count of Internet
search queries for these variables from the United States from January
2009 to March 2015. Data are scaled to the maximum search traffic for
the specific array of terms (represented as 100) during the period
from January 2009 to March 2015 in the United States.

-- the other variables are "QE1", "QE2", "QE3" as well as "OT"
(=operation twist), "FG" (=forward guidance), "TAP" (=tapering) .These
are dummy variables indicating when these events occurred.

-- Note: there are many 0's because for many of these weeks there were
few or no such searches (in the case of the first three variable) or
the events in question were not occurring in that period (for the
dummy variables)