Basic information on financial databases: cook books, tips and tricks & economic news

This blog contains schematic easy to grasp - hands on - help in performing searches in economic databases, making work sets and making them inter-exchangeable between the databases.

* Disclaimer. I am not a finance professional. Most posts are the result of personal findings.

All presented images are scaled and can be enlarged to original size (click the picture).

Search This Blog


Return, expected return and abnormal return

In finance, an abnormal return is the difference between the expected return of a security and the actual return. Abnormal returns are sometimes triggered by "events." Events can include mergers, dividend announcements, company earning announcements, interest rate increases, lawsuits, etc. all which can contribute to an abnormal return. Events in finance can typically be classified as occurrences or information that has not already been priced into the market.

In stock market trading, abnormal returns are the differences between a single stock or portfolio's performance in regard to the average market performance over a set period of time.[1] Usually a broad index, such as the S&P 500 or a national index like the Nikkei 225, is used as a reference for the average market performance.
For example if a stock increased by 5% because of some news which affected the stock price, but the average market only increased by 3%, then the abnormal return was 2% (5% - 3% = 2%). If the market average performs better than the individual stock then the abnormal return will be negative.

Abnormal return = Actual return - normal return

Definition: Used in the context of stock returns; abnormal returns means the return to a portfolio in excess of the return to a market portfolio. Contrast excess returns which means something else. Note that abnormal returns can be negative.
Example: Suppose average market return to a stock was 10% for some calendar year, meaning stocks overall were 10% higher at the end of the year than at the beginning, and suppose that stock S had risen 12% in that period. Then stock S's abnormal return was 2%.

In contrast, excess returns are returns above the risk-free rate. What is used in the CAPM is the expected excess return instead of excess return itself. Expected excess return is the so called risk premium. The measure of these excess returns is expressed by the measure alpha.

Where can we find it and how do we find it?

In Datastream.
You'll have to calculate the abnormal return yourself.

Ri = Total Return (datatype RI)
The return in Datastream is expressed as an index. The base date of the company is 100 and from there on the return evolves per day. The datatype for Return Index is RI. Usually the numbers for RI are very high. This differs from an actual return, obviously, but trends should be similar.

Actual return
The actual return is a percentual calculation of the differences per day of the RI.
The trend, can of course be calculated from stock prices also.
See my blog post here. < click

The actual return is a percentual calculation of the differences per day of the RI. You can ask for this percentual return in Datastream as follows:
  • Go to Time series request
  • Enter a company or a list of companies
  • Enter RI in the datatype field
  • Click on the Fx button
  • Search for the Percentage change, Period
  • Leave X as is
  • enter 1D
  • Now you see this in the datatype field: RI,PCH#(X,1D)
The result now will be the index numbers of the return + the percentual change (= stock return)
You can also do this with a based return (= 100 on your start date). To do that you enter this in the datatype field
Source blog:
 This formula also works for index Return Index: PCH#(mnemonic(RI),1M) e.g. S&P 500: PCH#(S&PCOMP(RI),1M) (put this in the item rule) When working with the constituent LS&PCOMP it'd be  PCH#(X(RI), 1M)

Expected return
How do you calculate the average of a probability distribution? As denoted by the above formula, simply take the probability of each possible return outcome and multiply it by the return outcome itself. For example, if you knew a given investment had a 50% chance of earning a 10% return, a 25% chance of earning 20% and a 25% chance of earning -10%, the expected return would be equal to 7.5%:
The forecast return on a stock, Return on Equity (ROE)
Abundant in Datastream. Noteworthy imho are Return on Equity (Datastream Worldscope DWRE andthe IBES section forecast for 12 month period:
filter on return on equity
12 Month Forward Return on Equity ROEF1FD12 (frequency yearly)

Cumulative abnormal return (CAR)
Cumulative abnormal return, or CAR, is the sum of all abnormal returns up to time . If no event occurs then CAR equals zero.

Market value:
International Valuation Standards defines market value as "the estimated amount for which a property should exchange on the date of valuation between a willing buyer and a willing seller in an arm’s-length transaction after proper marketing wherein the parties had each acted knowledgeably, prudently, and without compulsion.

The following table shows a Datastream output on a yearly search on H:AH (AHOLD) with datatypes RI and formulas PCH#(X,1D) = % change; Return of Equity ROE (DWRE) and PCH#(X(LI), 1D) = % change compared to LI (Local Market Index = AEX); plus the % change with 7 decimals and lastly the abnormal return. The latter was provided by the Thomson Financial helpdesk. REGR = regression, PCH = % change
As far as I understand you can calculate abnormal return G column yourself by using columns E and F (or column C, for that matter)

Mark's Blog provides more information and a formula
For those  that don't want to read the explanation, here is the formula:

Disclaimer: I am still working out how this exactly works, and I cannot claim this to be correct. I am open for all suggestions/ corrections:


No comments:

Post a Comment