WRDS

Introduction to CRSP Stock Price Data

This document is intended to provide background information, descriptions, definitions, and sources useful to those who want to use CRSP stock price data for financial research.

CRSP refers to the Center for Research in Security Prices, located at the University of Chicago. The CRSP database includes monthly and daily price quotations for common stocks, excluding preferred stock and including ADRs, traded on the New York and American stock exchanges and NASDAQ.

The terms of the CRSP database license maintain that the data may be used only for academic research. Commercial use and republication of the data are prohibited without permission. The following citation should be given when CRSP data are used in academic research:

Source: CRSP, Center for Research in Security Prices. Graduate School of Business, The University of Chicago (2004). Used with permission. All rights reserved. www.crsp.uchicago.edu

CRSP data are perhaps most conveniently accessed through the WRDS system at the Wharton School of University of Pennsylvania. This brief document can give only a cursory overview of the contents of the CRSP database. Comprehensive documentation is available through WRDS, or directly from CRSP.

This site also provides presentations and tutorials prepared by CRSP. Additional help with the use of CRSP data is also available through Baker Library Research Services.

Data availability

CRSP monthly stock price data cover the period from December 1925 through the present.* Daily data begin in July 1962. However, while NYSE data are available throughout the entire period, AMEX data start in 1962, and NASDAQ data begin in 1972. The following table summarizes the availability of data by date and exchange:

* CRSP stock price data as licensed by Harvard Business School are updated on a quarterly basis. Updates are generally within six weeks of the end of the latest calendar quarter. When accessing the data in WRDS, take care to select "Quarterly Updates..." to access the most recent data.

Data items

The principal data items available in CRSP include individual stock prices (high, low, and close; raw and adjusted); trading volume; dividends and other distributions; holding period returns; and market indices, including both CRSP proprietary indices and the S&P 500. The following table summarizes the principal data items in CRSP. Brief definitions are also provided later in this document.

Identifiers

CRSP also includes identifying data for each company and stock covered by the database. These identifiers include the proprietary CRSP data items PERMCO and PERMNO, as well as other, more commonly encountered concepts:

Definitions

The following definitions of the most critical CRSP variables may aid in the effective use of CRSP data.

• PERMNO
  Every stock issue covered within CRSP is assigned a unique number, its PERMNO. While a company may change its name, ticker, exchange, or CUSIP from time to time, the PERMNO of a stock within CRSP remains unchanged. Consequently, the PERMNO is the principal identifier of a stock in the CRSP database, and provides a reliable way of tracking a stock over time. As discussed immediately below, a PERMNO identifies an individual stock. For the most part, companies will have only one issue of common stock on the market at any time, so there is usually a unique link between PERMNO and company. However, a company with multiple common stock issues on the market will be associated with multiple PERMNOs.


• PERMCO
  If a company has issued multiple tranches of publicly traded common stock, CRSP will track all of them individually with different PERMNOs. The CRSP PERMCO is a number that uniquely identifies each company within the CRSP database. This makes it possible to associate different stock issues from the same company. Most companies have only a single common stock issue outstanding at any time, so there is usually a one-to-one correspondence between PERMNO and PERMCO. As of December 2003, for example, CRSP was covering three different series of common stock (A, E, and H) related to General Motors. These three stocks have different PERMNOs within CRSP, but the common PERMCO indicates that the stocks belong to the same company.


• Split factor
  When a company splits its stock, prices and numbers of shares are not directly comparable over time. For example, on May 27, 1999, IBM split 2-for-1. The number of shares outstanding doubled from the previous day, from 920 million to 1.84 billion, while the price per share was cut roughly in half, from $236.25 to $116.00. Looking only at the raw, unadjusted, prices would give us a very inaccurate measure of the single-day change in price. We have to apply an adjustment factor, dividing the earlier price by two or multiplying the later price by two, to have comparable prices. By the same token, any share-based measurements - trading volume, for example - would have to be adjusted if we want to compare before- and after-split values


• Price
  Usually, the CRSP price is the closing price, the price of the last reported trade on any given day. However, if no trade is recorded on a given trading day, the reported price will be the negative of the average of the bid and asked prices for that day. The price reported in CRSP monthly files is the price of the last trading day of the month.
  
  When comparing prices at different times, care must be taken to apply the split factor appropriately, as described above, to assure that the prices are truly comparable.


• Price - Ask or High, Bid or Low
  Daily: The high price reported by CRSP is the highest trading price each day or, if no trade is reported, the closing ask price. Similarly, the low price is the lowest trading price of the day or, if no trade is reported, the closing bid price. CRSP also reports the high and low prices for each trading period.
  
  Monthly: The reported monthly high and low prices are the highest and lowest daily closing prices, rather than the highest and lowest daily high and low prices. Again, if no trade is recorded, the negative of the average of the daily bid and ask prices are reported instead.


• Return
  The return measures the percentage change in value over a period of time, and must take into account the effects of splits and other capital actions. Further, depending on the research objective, one might include or exclude dividends and other distributions in the calculation of stock returns.
  
  In the following examples rt represents the holding period return from period t-1 through period t, pt represents the closing price at period t, ft represents the split factor, and dt represents the per share value of dividends or other distributions.


• Excluding dividends
  
  
• Including dividends
  
  
• Calculating returns
  Splits and dividends are relatively infrequent events, so the return most days is simply the relative or percentage change in price from the end of one day to the end of the next:
  
  
  
  For example, the price of IBM on the NYSE at the end of May 12, 1999, was $225.50, while the price at the end of May 13, 1999 was $246. So the return between May 12, 1999 and May 13, 1999 was:
  
  
  
  On the other hand, IBM paid a dividend of $0.24 per share the previous week, so the return from May 5, 1999, through May 6, 1999, was:
  
  
  
  Finally, IBM split 2-for-1 later in the month, so the return from May 26, 1999, through May 27, 1999, was, incorporating a split factor:
  
  
  
  
• Compound returns
  A compound return accumulates returns over a number of successive periods. It is calculated as:
  
  
  
  That is, add one to each individual period return; multiply all these elements; and subtract one from the final product.
  
  When cumulating returns, it is important to remember to add one to the individual period returns before multiplying the successive returns. Remember also to subtract one at the end of the process to convert the total cumulative value (1+r) to a total cumulative return, or total change in value.
  
  The following table illustrates the calculation of the compound return for IBM over the first week of May 1999. The stock accumulated nearly 4 percent in value over that five-day trading period.
  
  

  PERMNO	Date	Return	Compound Return
  12490	19990503	0.014640	0.014640
  12490	19990504	-0.001178	0.013445
  12490	19990505	0.000590	0.014042
  12490	19990506	-0.012422	0.001446
  12490	19990507	0.038232	0.039733



• Market index returns
  CRSP calculates returns for equal-weighted and value-weighted portfolios of all available stocks each trading period, based on individual stock returns. Each of these is calculated in two ways, including dividends and excluding dividends.
  
  
  • Equal-weighted: The CRSP equal-weighted market return is a simple average of the returns of all traded stocks.
    
    
  • Value-weighted: The value-weighted market return is a weighted average of all stock returns, with the weights given by the market value of the stock issue (price times shares outstanding) at the end of the previous trading period.