Clean datasets

This page shows the data used in the study after it's been off corrupts records. The detail of the corrections carried out is available in the 'corrupts' section. The first 50 records of the resulting clean datasets are shown below with the full datasets individually available for download. A number of seemingly corrupt records are still pending correction with the detail also available in the 'corrupts' section. The corresponding firm-date combinations are excluded from the book dataset while the full time series of prices, returns and events for the corresponding firms are excluded from the prices, returns and events datasets below.

Book

Source file: 'H and peck data_Final.xlsx'; tab: 'Dataset'

Prices

Returns

Events

Corrupt records

This page shows the corrupts records found in the raw data extracted from the main source files. Most corrupts records have been cleaned as resulting from an obvious typo. The detail of these corrections is shown in the 'sorted' section below. A number of other incorrect entries for which the error is not obvious are still to be addessed. The detail of these is shown in the 'pending' section below.

Sorted

Book descriptive statistics

This page shows descriptive statistics for the book dataset. The results are shown by reporting frequency starting with the statistics for the firms reporting annually, followed by those for the firms reporting semi-annually, and finally the statistics for all firms combined. For each year the statistics are shown for each reporting period. Results for annually reporting firms naturally show only 1 reporting period, while those for semi-annually reporting firms show 2 reporting periods. The 'n' statistic documents the number of available observations for the correpsonding variable-year-reporting period combination.

Returns descriptive statistics

This page shows descriptive statistics for the returns dataset. The results are shown by reporting frequency starting with the statistics for the firms reporting annually, followed by those for the firms reporting semi-annually, and finally the statistics for all firms combined. For each year the statistics are shown for each reporting period. Results for annually reporting firms naturally show only 1 reporting period, while those for semi-annually reporting firms show 2 reporting periods.
The '# days available' family of statistics refers to the number of return days and non-zero return days available around corporate events. In this context, availability is defined as no more that 5 successive missing records.

Firms

This table shows descriptive statistics for individual firms.

Aggregate

This table shows descriptive statistics for goups of firms, namely those reporting annually, semi-annually, and all firms combined. The statistics are shown for each year and reporting period. The 'n' statistic documents the number of available observations for the correpsonding variable-year-reporting period combination.

High data availability firms' descritive statistics

This page shows the same descriptive statistics as those shown in the 'book' and 'returns' pages re-computed for firms with 'high data availability'. In this context, high data availability is defined in four steps. The first involves selecting years of interest; these are the ones for which data is available in both the book and return datasets, namely, 1912 and 1913. The second step involves selecting high data availability book firms: first the variables with are least 150 observation for the two years of interest are selected; second firms with data available for these variables for both years are selected. The third involves selecting high data availability returns firms where firms are selected if for both years of interest they show at least 15 returns days available before as well as after at least one of their recorded corporate event. The fourth step involves taking the intersection of the high data availability book and high data availability returns firms.

Names

Book descriptive statistics

Returns descriptive statistics