For those of us who have been in the Financial Services industry for a while, Bloomberg always meant the terminal, but over the last 10 years Bloomberg has been increasing the set of data feeds they offer, until we are left with a (sometimes bewilidering) set of feeds each with its own unique characteristics. During this time, I’ve worked with all of them. Written feed handlers for them and helped clients write applications that consume data from the feed handlers. This is a quick tour.
Normally one might expect with a market data feed, that you request a security and you get all the data. Bloomberg is rather different in this respect because of the huge quantity of reference data available, which must be requested separately. I make a deliberate distinction between static data (price data whose values may change during the day but not broadcast, e.g. FWDRT1MO) and reference data (e.g. IDISIN, INDUSTRYSECTOR even though they are requested and delivered in the same way. Also Bloomberg provide some calculation facilities through their 'override fields'.
I'm just getting into this myself. There are two options for requesting data: SFTP and Web Services.
To my understanding, the SFTP option requires a Bloomberg application ('Request Builder') in order to retrieve data. The second option (Web Services) doesn't seem well-documented, at least for those working with R (like myself). So, I doubt a library exists for Web Services at this point.
Bloomberg provides an authentication certificate in order to gain access to their network, as well as their web services host and port information. Now, in terms of using this information to connect and download data, that is still beyond me.
If you or anyone else has been able to successfully connect and extract data using Bloomberg Web Services and R, please post the detailed code to this Blog! Joe, Bloomberg will shut down Web Services access, only SFTP will be available. It's not that bad, as you don't need to write all this boilerplate code to keep track of connections, requests, etc.
Even though there are some tricks in getting data via SFTP (say, you request 'every day at 23:00, but a file can appear at 23:05, then an update at 23:10) Regarding Request builder - it is possible to create your own 'request files', this is basically about serializing/deserializing from/to.csv-like. Same-ish boiler-plate code. – Dec 30 '15 at 8:19.
The Bloomberg Data Licence provides access to a huge range of reference data delivered as downloaded files via an ftp based mechanism. You simply parse.