For a UK finance or accounting dissertation the data question comes down to seven sources: LSEG Workspace and Datastream, Bloomberg, WRDS, FAME and Orbis, Companies House, the Bank of England database and Yahoo Finance. Which one you use depends on whether your library licenses it, whether your question is about listed or private firms, and whether you need prices, accounts or macro series. The table first, then the verdict.
| Source | What it holds | Access for a UK student | Best for | Where it falls short |
|---|---|---|---|---|
| LSEG Workspace / Datastream | Prices, fundamentals, estimates, macro time series (Datastream: 120 years of series, per LSEG) | Library licence, usually a fixed number of seats | Event studies, asset pricing, long time series | Steep interface; seats fill up in March |
| Bloomberg Terminal | Prices, fixed income, company financials, news | Library terminal room, one seat at a time | Bond and derivatives questions, live data | Cannot be used from home; export limits |
| WRDS (CRSP, Compustat, IBES) | US-centred research datasets, some global | Institutional subscription; many UK business schools have it | Replicating a published finance paper | UK coverage thinner than US |
| FAME / Orbis | Accounts of UK and Irish private and listed companies (FAME); global (Orbis) | Library licence, web access | SME, private-company, governance and audit questions | Smaller firms file abridged accounts |
| Companies House | Filed accounts, officers, PSC, bulk data | Free, public | Governance, director, insolvency, single-firm studies | Manual extraction; XBRL parsing for bulk |
| Bank of England database | Interest rates, exchange rates, money and credit series | Free, public | Monetary policy, banking, macro-finance | No firm-level data |
| Yahoo Finance and similar | Daily prices, basic fundamentals | Free | Small portfolio or volatility studies | Survivorship bias; no audit trail |
This guide is for UK undergraduate and MSc students in finance, accounting, banking, economics with a finance question and business studies with a quantitative project. The free public catalogue by subject is in our guide to UK data sources for a dissertation; this one is about the licensed financial databases behind the library login and the two public registers that finance students underuse.
The ranked shortlist
1. LSEG Workspace and Datastream: best for most finance dissertations
What was Refinitiv Eikon is now LSEG Workspace, and Datastream is its time-series engine. Together they cover equity prices, company fundamentals, analyst estimates, indices, bonds and macroeconomic series across markets, and they export to Excel through an add-in that most UK business school libraries teach in a one-hour session. For an event study, a capital-structure regression or a market-efficiency test on the FTSE, this is the source your supervisor expects. Who it suits: anyone with a question about listed companies and a library that holds the licence. Where it falls short: the interface has a learning curve measured in days, licences are usually limited to a handful of concurrent users, and the seats are all taken in the last month before every dissertation deadline. Book time in the autumn and download more than you think you need.
2. Bloomberg Terminal: best for fixed income and live markets
Bloomberg is the industry standard and many UK universities run a terminal room, sometimes with the Bloomberg Market Concepts certificate attached to a module. For bond yields, credit default swaps, derivatives pricing and anything that needs the news feed, it is the deepest source available to a student. Who it suits: students with a fixed-income or trading question and access to a terminal. Where it falls short: you cannot use it from home, the export functions have daily limits, and a dissertation that needs ten years of data for 300 firms will hit them. Use it for the specialised series and pull the bulk data from Workspace.
3. WRDS: best for replicating a published paper
Wharton Research Data Services has been the access layer for academic finance datasets for more than thirty years, and if your dissertation replicates or extends a paper from the Journal of Finance or the Journal of Financial Economics, the original authors almost certainly used CRSP for prices, Compustat for fundamentals and IBES for analyst forecasts through it. Who it suits: MSc students doing a replication with a UK twist, and anyone whose supervisor works in the US academic tradition. Where it falls short: the datasets are US-first, UK coverage through Compustat Global is thinner, and not every UK institution subscribes. Check the library database list before you plan around it.
4. FAME and Orbis: best for private companies and accounting questions
FAME holds the filed accounts of UK and Irish companies, private as well as listed, with up to ten years of financials, ownership and director information in a searchable, exportable form; Orbis is its global sibling. Both were Bureau van Dijk products and are now sold by Moody’s, so your library catalogue may list them under either name. Who it suits: accounting students working on audit quality, earnings management, SME financing, corporate governance or insolvency, where the population is private firms that never appear in a price database. Where it falls short: small companies file abridged accounts under the Companies Act thresholds, so profit and loss detail disappears exactly where you want it, and the search logic rewards students who have read the help pages.
5. Companies House: best free source, and better than students think
Every UK company’s filed accounts, confirmation statements, officers and persons with significant control are public on the Companies House register, free, and downloadable as PDFs one at a time. For bulk work, Companies House publishes a free monthly company snapshot in CSV and a daily accounts data product in XBRL, which a student comfortable with Python or R can parse into a panel. Who it suits: single-company or small-sample governance studies, director-turnover and insolvency questions, and anyone whose library lacks FAME. Where it falls short: extraction is manual at the single-firm level and technical at the bulk level, and there is no ratio calculation; you build the variables yourself.
6. Bank of England database: best for banking and monetary questions
The Bank’s statistical database publishes interest rates, exchange rates, yield curves, money and credit aggregates and lending data as downloadable series, updated on a published schedule. Who it suits: dissertations on monetary transmission, bank lending, mortgage pricing or sterling, usually paired with ONS series for output and prices. Where it falls short: there is no firm-level data, so a banking dissertation about specific banks needs FAME or Workspace alongside it.
7. Yahoo Finance and other free price feeds: acceptable, with a caveat
Free daily price histories are fine for a small portfolio study or a volatility comparison, and markers accept them for undergraduate work when the licensed sources are unavailable. The caveat, which must appear in your limitations, is survivorship bias: delisted firms vanish from free feeds, which flatters any strategy tested on them. Who it suits: students without library access to the sources above. Where it falls short: no audit trail, occasional data errors, and no fundamentals worth the name.
Which source for which question: five worked matches
The shortlist is easier to apply when the question comes first. Five dissertation questions UK finance students actually propose, and the source each one needs.
- Does the announcement of a dividend cut move the share price of FTSE 250 firms? An event study. Daily prices and the announcement dates come from Workspace; the market index from Datastream. Sample: every dividend cut announced over five years, screened for confounding announcements in the same window.
- Do family-owned UK private companies carry less debt than non-family firms? A private-company question, so FAME: filter by ownership, export ten years of balance sheets, and build leverage yourself. Expect abridged accounts to remove the smallest firms from the sample and say so.
- Has the pass-through from Bank Rate to mortgage rates changed since 2022? Bank of England database for Bank Rate and quoted mortgage rates, monthly, with ONS inflation alongside. No firm-level data needed; the whole dataset is free and downloads in an afternoon.
- Did audit-firm rotation change reported earnings quality at UK listed companies? An accounting question that needs auditor names and accounts together: FAME or Orbis for the auditor field and financials, Companies House to check filings where the database field is blank.
- Does the momentum anomaly still hold in UK equities? A replication of a US paper on UK data. WRDS if your library has it, for the methodology the original used; otherwise Datastream for a constituent list with delisted firms included, which is the point at which a free price feed fails.
Notice that none of the five needs Bloomberg, and only one needs WRDS. The two sources with the most prestige are the two most undergraduate finance dissertations can do without.
The recommendation
Use LSEG Workspace and Datastream if your question is about listed companies and your library has it; it covers the widest range of finance dissertations with a single export. Use FAME if your question is about private companies or accounting quality. Use Companies House when the sample is small or the budget is zero, and treat Bloomberg and WRDS as specialist supplements rather than the primary source. Whatever you choose, record the exact database, the date of extraction and the screens applied, because the marker will ask and because you will not remember.
Three things the methodology chapter must say about the data
- Population and sample. Which index or register, which years, which exclusions (financials, utilities, firms with missing observations), and the final N. State the number of firm-years, not just the number of firms.
- Variable construction. The database field or formula behind every variable, so that leverage means the same thing in your table as in the paper you are following.
- Known biases. Survivorship bias in free feeds, abridged accounts in FAME, US-centred coverage in WRDS. Naming the bias is worth more than pretending the data are clean.
If part of the data is survey-based, for example a corporate finance dissertation that asks treasurers about hedging, the realistic response rates are in our analysis of business dissertation survey response rates, and they are lower than most proposals assume. For social survey microdata such as the Wealth and Assets Survey, the registration route is set out in our guide to getting UK Data Service data.
Which software, and how do you keep the data safe?
Most licensed sources export to Excel, and most UK finance dissertations are then analysed in Stata, EViews or R; SPSS handles regression but not panel data well. The trade-offs, and what happens to each package when your university licence lapses, are in our comparison of SPSS, R and jamovi. Keep the raw export untouched and version every cleaned file; a corrupted spreadsheet in April is the finance student’s specific nightmare, and the storage options are compared in our guide to dissertation backup and version history. Licensed data also carry terms: you may use it for your dissertation and you may not redistribute the raw series, so appendices should carry summary statistics, not the download.
Where finance students lose marks on data
Not on the source. On the paragraph that never gets written: the one that says which database, which screens, which years and which biases. If the data are downloaded and the methodology chapter is still a blank page, Tesify can draft the data and methodology sections from your own extraction notes, your own variable definitions and your own sample counts, so the chapter states what you did in the order a marker checks it. Every sentence stays yours and is 100% written by you; what you get is the chapter existing before the deadline rather than after.
Frequently asked questions
Which database do most UK finance dissertations use?
LSEG Workspace with Datastream, where the library holds it, because it covers prices, fundamentals and macro series for listed companies in one export. FAME is the usual choice for private-company and accounting questions.
Is Bloomberg better than LSEG Workspace for a dissertation?
Not usually. Bloomberg is deeper for fixed income and live markets but must be used at a library terminal with export limits, so most students pull the bulk data from Workspace and use Bloomberg for specialised series.
Can I write a finance dissertation with only free data?
Yes. Companies House, the Bank of England database, ONS series and a free price feed support a defensible undergraduate project, provided the limitations section names the survivorship and coverage biases those sources carry.
What is the difference between FAME and Orbis?
FAME covers UK and Irish companies; Orbis is the global version. Both originated with Bureau van Dijk and are now Moody’s products, so libraries list them under either name.
Does my university have WRDS?
Many UK business schools subscribe, but not all. Check the library’s database A to Z before planning a replication study, because the CRSP and Compustat datasets are only available through it.
How many firm-years do I need?
Enough for the regression you plan: a panel of 100 firms over 10 years gives 1,000 firm-years, which supports fixed-effects models comfortably. State the count after exclusions, not before.
Can I put the downloaded data in my appendix?
Summary statistics, yes; the raw licensed series, no. The database terms prohibit redistribution, and a marker does not need the download to assess the work.
What should I record when I extract the data?
The database and product name, the date, the exact screens and filters, the fields exported and the number of observations at each step. That note becomes the sample section of the methodology chapter.