06445 - Business Statistics

Course contents

Statistical Information for Business
National and international statistics, both official and non-official; web big data and web scraping techniques from static and dynamic websites (hints). Data from ERP, CRM, and PMS management systems; corporate surveys (organizational climate surveys and surveys of clients and suppliers).

Corporate Data Life Cycle (Statistical Information Quality)
Data cleaning: techniques for identifying and correcting duplicates, outliers, missing data, and for data standardization. Data validation: methods for assessing the consistency, continuity, and completeness of information. Data integration within data warehouses. Corporate statistical summaries: operational KPIs (Key performance indicators) for management control and financial KPIs for solvency analysis.

Statistical Methods for Decision-Making in Business and Finance (Cross-Section Data)
Key performance indicators (KPIs) for level and KPIs for risk (graphical summaries, means, and variability measures). Relationships among KPIs (cross-tabulations). Decision-making based on inferential comparisons. Identification of managerial drivers (conditional means, linear regression, logistic/multinomial regression). SWOT analysis applied to mergers and acquisitions.

KPIs for Measuring Latent Phenomena (Composite Indicators)
Process of constructing a composite indicator (theoretical framework, data selection, empirical structure, validation). Weighting schemes and statistical aggregation techniques: correlation, risk-based weighting, regression, factor analysis, and correspondence analysis (introductory notes).

Statistical Methods for Decision-Making in Business and Finance (Time-Series)
Financial market indices. KPIs for measuring an asset’s liquidity and other financial KPIs (technical charts, moving averages, and oscillators). Support and resistance levels interpreted as confidence intervals. Exponential smoothing. RiskMetrics (overview). Trend extraction: stochastic versus deterministic trends. ARIMA modeling. Black-box models (supervised machine learning techniques) for classification (e.g., credit risk) and forecasting.

Evaluation of Forecasting Performance
Expected loss and the decision maker’s risk aversion. Evaluation of predictive accuracy: predictors, forecast errors, loss/cost functions, and predictor optimality under a subjective loss function. Forecasting performance ranking: descriptive approaches and inferential testing.

Readings/Bibliography

Teacher's slides

Teaching methods

The course is delivered both in the classroom (where theoretical aspects are covered) and in the laboratory, where simulations and business/case studies are conducted.

For students planning to attend the laboratory sessions in person, it is mandatory to complete prior training on safety in study environments via e-learning https://elearning-sicurezza.unibo.it/ (modules 1 and 2).

OPTIONAL (i.e., for a limited number of students), it is possible to apply to discuss a case study with the teacher (in the classroom) (no expected loss or gain).

Assessment methods

The exam aims to verify the achievement of the following learning objectives:

• knowledge of the statistical tools presented during the lectures

• ability to apply these tools

• ability to use the results obtained to interpret the phenomenon under study and to support decision-making processes.

The assessment of learning consists of a mandatory written examination and an optional oral examination. The written test includes a number of questions on methods ("theory questions") and exercises through which students demonstrate their ability to apply the acquired tools. Each question is graded on a scale from 0 to 36 points. Six points are assigned for unanswered questions. The final score of the written examination is calculated as the average of the individual question scores

During the exam, the use of supporting materials such as textbooks, handouts, notes, or electronic devices is not allowed.

To take the examination, students are required to register through the online noticeboard system within the established deadlines. Those who are unable to complete their registration by the due date must promptly inform the Teaching Office (and in any case before the official closing of the registration lists). It will be at the instructor’s discretion to authorize their admission to the examination.

Teaching tools

Teacher's slides. Computer sessions with Excel, Gretl, R and Pyhon

Office hours

See the website of Andrea Guizzardi