Hanken School of Economics Finance Master's Thesis Guide: 1720, 30 ECTS, 17170 Seminar and Empirical Methods

Quick answer: what is the Hanken Finance thesis route?

The current Hanken School of Economics Master’s programme in Finance is an English-language, two-year 120 ECTS Master of Science (Economics and Business Administration) programme in Helsinki. For the current 2025-2027 curriculum, the Finance thesis itself is course 1720, 30 ECTS. The current English/non-Swedish research seminar is 17170 Research Seminar in Finance, 5 ECTS. Hanken also requires 17011 Empirical Methods in Finance, 10 ECTS and 17012 Advanced Topics in Financial Economics, 10 ECTS in the current Finance MSc plan. The seminar is separate from the thesis credits.

1. The current Finance code structure needs careful reading

Finance has a language-code detail that is easy to misread from older Hanken material. The controlling 2025-2027 public study plan lists the thesis as 1720, 30 ECTS. Older English-track plans used 1720-E, so students may still find the older code in historical documents or completed-course records. For the research seminar, Hanken’s current live English course page and 2026 Moodle clearly identify 17170 Research Seminar in Finance for non-Swedish speakers, while 17160 is the Swedish parallel. This guide therefore uses 1720 as the current thesis code and 17170 as the current English seminar code.

2. 1720 is the current 30 ECTS Master’s thesis object

The exact current Finance thesis object in the 2025-2027 plan is 1720, 30 ECTS. Hanken’s degree regulations independently require a 30 ECTS Master’s thesis inside the advanced studies of the two-year Master’s degree. This confirms the credit value from two different official levels. Students should not carry over a thesis code from Accounting, Economics or an older Finance plan simply because the ECTS value is also 30.

3. 17170 Research Seminar in Finance is a separate 5 ECTS course

The current English seminar is 17170 Research Seminar in Finance, 5 ECTS. Hanken’s current Moodle pages explicitly state that 17170 is for non-Swedish speakers and 17160 is the Swedish-speaking parallel. The seminar is not embedded inside the 30 ECTS thesis. This matters for study planning because students need to account for the 5 ECTS seminar and the 30 ECTS thesis as separate parts of the advanced studies.

4. Check Sisu before registering for 17170

The public thesis instructions state that students must meet the prerequisites for the research seminar, but the formal current prerequisites are controlled by Sisu. The public Finance programme page, Moodle overview and this guide are not substitutes for the current Sisu implementation. Before choosing a thesis start semester, check the latest 17170 version, prerequisites, registration dates, teaching periods and any instructions concerning attendance, presentation or topic approval.

5. The Finance programme has a strong empirical and quantitative methods environment

The current Finance MSc makes 17011 Empirical Methods in Finance, 10 ECTS mandatory. Hanken describes the programme as developing advanced analytical and quantitative skills, and the methods course is specifically designed to prepare Finance students for quantitative research used in the Master’s thesis. This makes empirical work especially common, but it does not create one compulsory estimator or software package for every thesis. The method still has to fit the research problem and the available data.

6. 17012 provides the advanced financial-economic theory environment

The programme also requires 17012 Advanced Topics in Financial Economics, 10 ECTS. Current Hanken course information covers theoretical foundations and modern developments in asset pricing and financial decision-making under uncertainty, including portfolio choice, pricing theory, information problems, agency and capital structure. A Finance thesis should use theory as an analytical framework where relevant, rather than treating the literature review and empirical analysis as unrelated sections.

7. Optional quantitative courses do not become universal thesis requirements

The 2025-2027 Finance plan includes electives such as Mathematical and Quantitative Finance, FINTECH and Blockchain, Factor Investing, Value Investing, Strategic Growth Investing and Machine Learning for Finance. These options broaden the methods environment, but they do not prove that every Finance thesis must use machine learning, advanced mathematics, an event study or a specific asset-pricing model. The thesis design should be selected from the research question, theory, data structure and evidence needed to support the intended claim.

8. Start with the claim you need to support

A strong Finance thesis begins with a clear research problem and purpose, not with a favourite model. The current Hanken thesis rubric separately evaluates the problem statement, purpose, theoretical framework, method choice, method use, results and interpretation. This structure rewards alignment. Before coding anything, define what the study is trying to explain, estimate, compare or predict and what type of evidence would be sufficient for that claim.

9. Theory should guide variable and model choices

Finance models are not only technical tools. They represent assumptions about investors, firms, information, risk, incentives or markets. If a thesis uses variables such as leverage, liquidity, abnormal returns, ownership concentration, volatility or ESG measures, the theoretical section should explain why those variables are connected to the research question. The empirical specification should then reflect that logic rather than collecting variables simply because they are available in a database.

10. Association is not automatically causation

A major evidence boundary in Finance is the difference between association and causal effect. A panel coefficient linking leverage with performance, or ESG scores with valuation, can describe a conditional relationship without proving that one variable caused the other. Causal language requires a defensible identification strategy and assumptions appropriate to the design. Fixed effects, instrumental variables, difference-in-differences, matching or natural experiments can support causal inference in some settings, but none becomes causal merely because the method name is used.

11. Event studies require a defensible event and benchmark

Event studies are common in Finance, but their logic should be explained rather than treated as automatic. Define the event, event window, estimation window and benchmark model. Explain why the event timing is sufficiently precise and why the expected-return model is suitable for the sample. If other information arrives at the same time, interpretation becomes more difficult. Abnormal returns can show market reaction relative to a benchmark, but they do not automatically prove the deeper mechanism behind the reaction.

12. Panel-data models need attention to unobserved differences and timing

Firm-level Finance studies often use panel data. A panel structure creates opportunities to control for persistent firm or time effects, but it also raises decisions about clustering, lag structure, dynamic relationships and reverse causality. Students should explain why fixed effects, random effects, pooled models or other approaches fit the research question. A statistically strong coefficient does not solve an identification problem if the explanatory variable is itself affected by the outcome.

13. Financial time-series analysis must respect temporal ordering

Time-series work should preserve the order in which information becomes available. Future information must not leak into model training, portfolio construction or forecasting evaluation. Stationarity, persistence, structural breaks and changing volatility can also matter. If the project forecasts returns, risk or volatility, separate model estimation from genuine out-of-sample evaluation and explain how the forecast origin and test period are defined.

14. Backtests can look strong because of overfitting

A strategy can perform extremely well in historical data because the researcher has tried many signals, windows, filters or portfolio rules. This creates data-mining and multiple-testing risk. A credible thesis should distinguish exploratory choices from confirmatory analysis, avoid using the test period to tune the model and report robustness when many specifications have been examined. In-sample fit is not the same as evidence that the strategy will generalise.

15. Machine learning needs leakage and validation controls

Machine learning can be useful for prediction in Finance, especially when data are high-dimensional or relationships are nonlinear. However, model complexity does not remove the need for research design. Prevent target leakage, keep training and test information separate, tune hyperparameters without contaminating the final evaluation and compare performance with meaningful baselines. Explain whether the objective is prediction, explanation or causal inference because these are not interchangeable research goals.

16. Statistical significance is not the same as economic importance

A coefficient can be statistically significant but economically small. A Finance thesis should therefore report magnitudes, units and uncertainty in a way that allows the reader to interpret practical meaning. For portfolio or investment results, transaction costs, turnover, liquidity and implementability may matter. For corporate-finance results, relate the estimated magnitude to economically meaningful changes in the underlying variable rather than discussing only p-values.

17. Asset-pricing tests are model-dependent

When a thesis evaluates asset-pricing models or factors, distinguish observed returns from the model used to describe expected returns. Alpha, factor loading, pricing error or model fit are conditional on the selected model and sample. Rejecting one model does not automatically establish a unique alternative explanation. If comparing factors or models, explain the economic motivation, sample period, portfolio construction and how multiple-model testing is handled.

18. Corporate-finance research should separate mechanisms from correlations

Corporate-finance topics can involve capital structure, governance, investment, payout, financing constraints or executive incentives. Observational firm data often contain strong selection and endogeneity problems. A relationship between governance and performance, for example, may reflect omitted firm characteristics, reverse causality or common shocks. The conclusion should be no stronger than the design permits, and alternative mechanisms should be discussed when they remain plausible.

19. Risk and volatility are measured through models

Risk is not one directly observable quantity. Historical volatility, implied volatility, value at risk, downside risk and factor exposures capture different aspects of uncertainty. Students should define the measure used, justify why it fits the research question and avoid writing as if one model captures all possible forms of financial risk. If results are sensitive to the chosen risk measure, that is part of the interpretation rather than something to hide.

20. Data access should be tested before the topic is fixed

The current Finance seminar tells students to check data availability early. This is practical because many Finance topics depend on proprietary company, market, transaction, analyst, fund or ESG datasets. Before confirming a topic, check whether the required variables exist, whether the time period is sufficient, whether identifiers allow merging, what licence or access restrictions apply and whether the data can be used in a public Master’s thesis.

21. Keep a reproducible data trail

Empirical Finance work often involves several data sources and many transformations. Record the source, access date, raw variable names, sample filters, merges, currency conversions, winsorisation, lag construction, portfolio rules and model inputs. Where possible, separate raw data from processed data and keep analysis scripts in a logical order. This makes it easier to explain changes in sample size or estimates during supervision and strengthens auditability.

22. Missing data and outliers need predefined logic

Financial datasets frequently contain missing observations, extreme values, delistings or reporting changes. Do not remove observations only because they weaken the preferred result. Explain inclusion and exclusion rules, whether outliers are winsorised or trimmed, and why. If the final regression sample differs from the descriptive-statistics sample, make the reason transparent. A clear sample-construction table can prevent many later questions from the supervisor or examiner.

23. Hanken requires a data processing description

Hanken’s thesis process explicitly requires a data processing description. Treat this as part of research design. Identify what data are used, where they come from, where they are stored, who can access them, whether they contain personal or confidential information and what happens after the project. Hanken also provides research data management support specifically relevant to bachelor’s and Master’s students.

24. Personal-data obligations depend on the dataset

Many Finance theses use market or company data without directly identifying natural persons, but some projects use executive, investor, employee, client, survey or interview data. Personal-data obligations therefore depend on the actual dataset, not the programme title. If individual-level records are used, check the legal basis, access controls, minimisation, retention and disclosure risk. Pseudonymised data may still be personal data if re-identification remains possible.

25. Ethical review is design-dependent

A standard secondary market-data study does not automatically require human-participant ethical review. Surveys, interviews, experiments or sensitive individual-level research can create a different route. Hanken lists specific advance-review triggers, including departures from informed consent, certain research involving minors, unusually strong stimuli, above-normal mental-harm risk and safety threats. Check the actual design early enough that ethics requirements do not become a late-stage obstacle.

26. A public thesis cannot contain confidential company material

Hanken states that Master’s theses are public after approval and cannot be classified as confidential. Company-sponsored Finance projects must therefore keep confidential background material outside the public manuscript. If reviewers need protected information for grading, Hanken allows it to be supplied separately. Agree the boundary before analysis begins so the final thesis can explain the method and evidence without exposing restricted material.

27. The normal thesis length is 60-70 factual pages

Hanken’s general guidance recommends 60-70 pages of factual content for the Master’s thesis. The normal maximum is 100 pages including cover, contents, references and appendices, unless a deviation is agreed in advance. Finance theses can become long because of tables, robustness tests and appendices. Use appendices for supporting material where appropriate, but keep the main argument focused enough that the research question, method, key results and interpretation remain visible.

28. Referencing should follow the current Hanken or subject instruction

Hanken’s general reference guide follows APA 7 and notes that some departments prefer Oxford. The current Finance sources reviewed for this guide do not establish a unique public Finance-only citation style that overrides the general boundary. Follow the current seminar, supervisor or subject instruction. Whatever style is used, keep references consistent and make data, literature and methodological sources traceable.

29. AI can support tasks but cannot replace the research contribution

Hanken’s current AI guidance permits some forms of support, including brainstorming, planning, coding, data processing, visualisation and language assistance, subject to the applicable course or thesis rules. The student remains responsible for the research question, model choices, results, interpretation and conclusions. AI must not fabricate data, references or results, and it must not produce the final academic analysis as if it were the student’s own work.

30. Proprietary financial data need extra caution with AI tools

Finance projects often use licensed or confidential datasets. Permission to use AI for study support does not automatically permit uploading proprietary market data, company material, personal data or restricted research files to an external AI system. Check the data classification, licence and Hanken’s approved tools before uploading anything. Even a dataset without names can contain sensitive information or contractual restrictions.

31. The maturity test is compulsory

Hanken requires a maturity test in connection with the Master’s thesis. It is a supervised digital essay related to the thesis and is assessed pass/fail. The current guidance recommends 400-800 words. The language route depends partly on the student’s previous school education and whether Finnish or Swedish proficiency has already been demonstrated in an earlier degree. The supervisor evaluates subject knowledge and a language teacher may also evaluate language where required.

32. Submission is final and requires PDF/A

When the thesis is ready, Hanken uses an authenticated electronic form and requires the final thesis as PDF/A. The submitted version is final: no corrections or additions can be made after submission, and a thesis submitted for grading and failed cannot simply be submitted again through the normal route. Before uploading, complete the supervisor-agreed corrections, verify the PDF/A file, remove confidential material and arrange the maturity test with the supervisor and department Administrative Coordinator.

33. English theses go through Turnitin plagiarism control

Every Hanken Master’s thesis must pass plagiarism control, and English-language theses are checked in Turnitin after electronic submission. The similarity report is a tool for the supervisor and is not an automatic plagiarism decision. Correct quotation, paraphrasing, attribution and source management remain essential. Data fabrication, misleading manipulation and unauthorised AI use can also fall within Hanken’s academic-misconduct framework.

34. Two reviewers propose the grade and the Education Council grades the thesis

The current formal rules require two examiners from different subjects, with at least one holding a doctoral degree. Hanken’s submission instructions state that one reviewer is the supervisor. The reviewers provide an evaluation and proposed grade to the Education Council, which formally approves and grades the thesis. Hanken uses the Master’s thesis scale 1-5, with 0 as fail.

35. Use the current 2026 AoL rubric during the project

From 1 August 2026 Hanken uses a common Master’s thesis AoL rubric. It evaluates the problem and purpose, theoretical framework and use of theory, choice and use of methods, results, analysis and interpretation, fulfilment of purpose, contribution and information sources. It also checks independence and process, maturity test, plagiarism control, research ethics and AI compliance. Using the rubric during drafting can reveal structural problems before final submission.

36. A practical Hanken Finance thesis sequence

A safe workflow is to confirm the current 17170 seminar requirements in Sisu, enter the seminar, define a feasible research question, verify data access, agree a research plan with the supervisor and complete the data-processing requirements. Build the theoretical framework and empirical design together, document sample construction and code, then keep conclusions within the identification and model limits. Use seminar feedback to revise the manuscript, complete formatting and reference checks, arrange the maturity test, prepare the final PDF/A and submit only when the final supervisor-agreed version is ready.

Final note

The most important programme-specific facts are: current thesis code 1720, 30 ECTS; current English/non-Swedish research seminar 17170, 5 ECTS; mandatory 17011 Empirical Methods in Finance, 10 ECTS; mandatory 17012 Advanced Topics in Financial Economics, 10 ECTS. Older Finance material may show 1720-E, and Swedish materials use 17160, so the code context should be checked rather than guessed. For operational actions such as Sisu registration, current seminar instructions, deadlines, coordinator details, data access and AI use, always recheck the current Hanken source before acting.