Bloomberg Data Retrieval 

1.  Introduction

Click on the Bloomberg Data Wizard  and you will be presented with the template for retrieving data from Bloomberg:

Let us first explain the tabs .  The default here is   for the retrieval of historical datasets.  The  tab allows you to request “Headers”, which is a snapshot of the latest trading information (one day only) of a security.  “Analytics”  brings down the last recorded value of whatever you have selected from the Bloomberg Analytics menu.

2.  Single Item Data Retrieval

In the  tab, look at the top left part of the template:

You will see the default Bloomberg identification type, Ticker-Exchange.  With this identification type you must enter the security symbol, the exchange on which it resides (or else it will use your default), and the Bloomberg market sector from a dropdown list.  Other Identification types will require different information, and the wizard will adjust accordingly.

If you  now enter a stock symbol (e.g. “ibm”) in the  box, accept the other defaults, and press    You will then be presented with a list of all available historical fields, arranged into related groups, each with a checkbox. 

Check any desired fields, on any number of the field groups, and click on the  button.  Then you will get as the default response: the last calendar year of the data fields you requested, displayed as daily data (if available). This dataset will be named “blp_ibm”, unless a dataset by that name already exits.  Should such a dataset already exist with that name, your downloaded dataset will be given a temporary name, such as “temp1”.  The dataset blp_ibm will be saved in the “data\Bloomberg\history” subfolder of the fdc folder:

A temporary dataset will not be saved, although you can SAVE AS and give it a name.  Downloaded datasets will automatically be in FDC data format, and will have a default plot that can be changed in a wide variety of ways.

When the operation is complete, the FDC “code” for your request will appear in a command line. 

The purpose for this is to show you how to make such a request directly from a command line.  You can edit the line and resubmit, or include the line in a set of lines to be automatically run, or in a list-based loop.  You may later wish to automate some of your data retrieval, in which case it will be helpful to have sample code.

In addition to the Ticker Exchange identification type, you may request data by CUSIP, ISIN, Sedol, Valoren, or by any identification type known to Bloomberg, and the wizard will adjust to request just the information it needs.  For example, Just click on the Identification Type drop-down menu, and choose CUSIP: 

Then simply enter the CUSIP number it the single field that appears to the right.  In this case, you will not be asked for the exchange or market sector, since the CUSIP identifier is unique.  You can then proceed as above to download the default dataset containing all fields you select for this security.

No matter what identification type you have used to specify data, there will be other fields to fill out if you don’t want to accept all defaults.

One is the the frequency of the data .   Of course, FDC gives you the ability to turn daily data into weekly, monthly, etc.  Additionally FDC allows you to specify the day of the week you choose as the end of the week.  Thus it is probably best to download daily data from Bloomberg, and then further manipulate it in FDC.  But that option is yours.  Daily is the default if you do nothing.

Next you can specify the date range:  .  If you do not enter a date range, FDC will get you the last year’s worth of data, as we have already mentioned.  If you enter just the beginning date, you will get all the available data from that date until the present.  Likewise, if you only enter the ending date, you will get all data up to that date.  A useful alternative to actually entering the current date is entering “today”, as the Enddate; FDC will always know what that means.   If you retrieve data while the markets are open, Bloomberg will provide the current day’s data, such that the high, low and last price for the current day will appear as though trading were concluded for today.  We think that’s an advantage, but if you do not want today’s prices, enter yesterday’s date or “today back 1” in the Enddate box.

Finally, you can make the selection as to whether your request is for price or yield, the default is Price:   As always, the last thing to do is to press .

3.  Retrieving Data Via a List

Retrieving a single dataset is useful for conducting a variety of analyses on one security; but more likely than not, you will need to retrieve larger amounts and varieties of data.  In that case, it is best to construct a list of your regular needs. Lists are just text documents in which each line contains data related to one dataset or other item of interest.  They can be created and edited with Wordpad or any text editor, or directly within FDC, and they are normally stored in the “Lists” subfolder of the main FDC folder.

Construction of an FDC list for Bloomberg retrieval is easy but detailed, particularly with all of the possible Bloomberg options.  However, don’t fret, as once done, it will serve you forever.  So as not to disturb the flow of an illustration on downloading Bloomberg data, we have put a section on list construction at the end of this illustration. For the remainder of this illustration, we will assume that any lists mentioned are of the correct type.

Once again we choose the History tab, and concentrate on the list retrieval section shown here.

When you click on , you will be presented with the names of those lists available in the “Lists” folder:

Just select the one you want and click . 

Then . as explained in the previous section.  Note that the fields selection is universal to the list members.  That is, whatever you request for one, you request for all the items.  As with single data requests, enter a date range: .

You may leave the Enddate blank, in which case FDC will insert today’s date.

The chosen list could reference a considerable amount of data, and if you store it all in one location, it may be harder to find if you need it.  For this reason, we give you the option of saving the data to different locations:  Click the button, and then choose the location (if you leave this choice blank, the data will go into the “data\Bloomberg\history” folder as usual.)

Once you have done this, here’s an example of what you might  see:

As in the previous section, you can then override defaults for Data Period, and Prices vs Yields.  Then click on , and all requested data will be downloaded (with translation and default plot) to the desired location  This will happen in the background with no further notification.

At the same time, FDC inserts the equivalent command line code on the current line of your FDC command window.  This enables you to save this language in a command set, so that it can be run automatically every day, or reused with possible changes made.

Note that when you download historical data via a list, the data downloaded will replace any data with the same name.  Should you wish the current list download not to replace data, choose a different path. 

Should you wish to view what occurred, click the button  .  This will show any error that occurred, or any of the datasets that could not be downloaded.  For example, you could have entered a wrong symbol, which would have prevented you from downloading a dataset.  But you are under the assumption that everything was entered correctly, and still you don’t have that needed dataset.  If you go to the download log, you can find the specific request, and there will be an error message describing its failure to download and the reason.

4.  Retrieving Data Automatically

Above we have mentioned running the same commands for data retrieval every day.  We have also shown you that FDC provides the command line text for each request.  Ideally you should now build (and save) a command set with all of your data retrieval requests. 

After making your typical requests as above, just save the command set.

Then enter an appropriate name and save:

On a subsequent day, all you have to do is , select  and .  The command set will appear exactly as you saved it, and all you need do is click on the autoplay button .

Suppose you need to delay downloading for a few hours until all of the data is available, but you cannot wait around.  In that case, you need to tell FDC to start the autoplay commands at a specific time. The line:

initiates a “hold” or “wait” until 16 August 2005 at 230 hours.  You need not specify the date, if it is the same date as today:

With the above text line added to the beginning of your command set, click on autoplay , and FDC will begin the automatic execution.  The first line it encounters will tell it to wait until later, which it will do, before running the other command lines.  Startat can either be set to play at a particular date and time, or should you only specify the time, it will run at that time.  If you need it to run tomorrow (i.e. after midnight), you need to specify the date.  If this is something you do every day, insert and save the Startat command at the beginning of your command set as shown above.

When you initiate an autoplay sequence of commands, FDC will place a “Stop” pop-up button at the upper left of the command palate.  This will enable you to cancel out of autoplay should you need to.

5.  Retrieving Headers

Instead of retrieving historical data, you may click on the  tab shown here.

If you look at the top section, which retrieves a single header, you will see that you only need to specify the identification type and related information.  As before, you may use any Bloomberg identification type.  There is no need to specify fields, data period, dates, or any other information, as that is handled automatically in headers.  Then press the download button as before.

It works similarly to the procedure for retrieving historical data, with a few exceptions.  Headers datasets will be named with the prefix “blph_”, and saved in the subdirectory “data\Bloomberg\headers”.  Also, each time you download a particular Header, it will replace and save over an existing header dataset of the same name.  Once again, the equivalent command line is automatically created for you:

.

The bottom section has two parts.  The first part downloads and saves a list of headers to any designated location.  Again it is only necessary to specify a list of security identifiers.  One reason for downloading such a list of headers is to include them in command sets that update and process historical data.

The second part can be particularly useful: Downloading a list to a Spreadsheet.

Enter the name of a list in the edit box.  FDC will then download the headers for your list, put them in an Excel® spreadsheet, and pop open that spreadsheet. 

With this feature you can get an immediate table of the current values you need.  Again, the equivalent command is constructed on a command line, should you wish to save the instructions for future use.

.

The spreadsheet is automatically saved in the “excel” sub-folder of the “fdc” folder. 

If you were to save that command line and use it again, the spreadsheet created will overwrite the previous one of the same name.  You will not be asked if it is okay to overwrite.  If you want to save spreadsheets that are slightly different (perhaps because of dates or times), just give them different names. 

6.  Retrieving Analytics

Retrieving Analytics is similar to retrieving Headers, except you also have the ability to choose from a huge assortment of fields.  You select these fields exactly as illustrated in the Get History section above (i.e. by ticking off the boxes).  Analytics downloaded will be saved to the Bloomberg Analytics subdirectory with the prefix “blpa_” added to the name.  As with Headers, a new download will overwrite any existing analytics file of the same name.  You will not be asked if it is okay to overwrite.

You may also save a list of analytics to storage, or to an Excel spreadsheet, as with Headers.

7.  List Construction

Constructing a list within FDC is easy.  Click on the List Wizard icon , and .  If you already have the names of the datasets somewhere, say in a Word® or Notepad® file, just copy them in.  If you already have datasets with those names, then click on the  button, select the datasets and .  Their names will appear in the edit box, and you can then , and give that list a name and .  If you need to change items in the list, , and .  You can also construct and/or edit a list in any text editor, and FDC contains commands for combining and modifying lists in practically any logical manner.

However if you are constructing a list to download Bloomberg data, that list must contain the information required by Bloomberg to obtain the data.  It is likely that the additional information required in the “download” list may render that list usable only for downloading.  Thus you may want to give your "download” list a name appropriate for its use.

The chief requirement is that Bloomberg download requests identify the data being downloaded.  There are several alternate ways to construct each line in the list:

1.     If you are using the Ticker-Exchange Method of identification (the most common type), then the entry should be as follows:

 NAME   [Exchange]  MarketSector

where

a)      Name is the ticker symbol. 

b)      Exchange need only be specified if it is different from your default exchange.  For example, if you are in New York and want the price of British Telecom in London, you must specify LN.

c)      MarketSector corresponds to the Bloomberg Yellow Key.

A slight variation allows you additional flexibility.  You may specify an alternate name for the dataset to be saved under.  In this case, the line will look as follows (braces { } are needed):

NAME  [Exchange]  MarketSector  {alternate name}

 

2. For any other method of security identification, e.g. ISINs, CUSIPs, etc. you must specify the identifier Type.  This is done by listing “CUSIP, ISIN, etc. after a semi-colon “ ; ”,

For example

US6311001043 ; ISIN

You may also use the alternate name syntax in these cases, and this would probably be advisable.  For example, the command shown above will produce a dataset whose default name will be blp_US6311001043.  On the other hand the command

US6311001043 {abc} ; ISIN

will store the data under the name abc.

Below are all examples of legitimate List constructions to retrieve Bloomberg data.  Case sensitivity is not required within the FDC List Wizard.  A Ticker-Exchange type may optionally be added to downloads of that type, to clarify the entry more for the user.

         

 

Should you copy a list of security names in from another source, and face the seemingly annoying task of adding additional text to each line in your list, fear not.  FDC makes it easy for you to add text:

Lists are democratic; whatever you do to each item in a list, you do to all in the list.  That is, if you are going to request certain fields of data (say Open, High, Low and Close), your request will provide that data for all items in the list.  If you want volume data for only 3 items out of the Dow30, you will have to ask for that via a different request.

Also, if you have specified a path for saving data in a subfolder, every item in that list will be saved in the same subfolder.  Thus if you are going to keep your various data in different subfolders (which we recommend) you should segment the data into different lists accordingly.  That is, you will have a list of currencies to be downloaded and stored in the “currency” subfolder.  Likewise, futures prices may go into the “futures” subfolder.