Bachelor’s and Master’s Theses

A descriptive analysis of the data is conducted by dividing the sample of wheat and maize producers into two subgroups: (a) farmers who adhere to organic farming standards, and (b) farmers who do not use pesticides but do not meet organic standards (possibly enrolled in pesticide-free production schemes).

Based on survey data of 1540 Swiss wheat and maize producers, the thesis focuses on the crop protection strategies adopted by organic wheat and maize producers in Switzerland, comparing these practices to those of non-organic, pesticide-free producers. The latter group refers to farmers who avoid using chemical synthetic pesticides specifically for wheat and maize but may apply them to other crops within their rotation. Differences between wheat and maize cultivation methods are analyzed to understand how protection strategies may vary depending on the crop. The analysis furthermore looks at the types of Integrated Pest Management (IPM) strategies employed by both groups, such as long-term strategies like wider crop rotation or more resource-intensive investments, such as specialized machinery. Moreover, differences in farmers’ ratings of the effectiveness of IPM measures are investigated. This comparison sheds light on the diversity of crop protection strategies within different farming systems and provides insights into the perceived efficacy of those approaches. Main research questions: 1. How does the portfolio of IPM measures differ between organic and non-organic, pesticide free wheat and maize production? 2. What kinds of IPM strategies (i.e., long-term strategies such as wider crop rotation or strategies with a higher investment e.g. for machinery) are used by organic as compared to non-organic but pesticide free producing farmers? 3. Do organic farmers rely more on preventive measures? 4. How do both groups of farmers evaluate the effectiveness of IPM measures?

Robert Finger (rofinger@ethz.ch) Sharmin Akter (sharmin.akter@mtec.ethz.ch)

This project is set to limited visibility by its publisher. To see the project description you need to log in at SiROP. Please follow these instructions: - Click link "Open this project..." below. - Log in to SiROP using your university login or create an account to see the details. If your affiliation is not created automatically, please follow these instructions: http://bit.ly/sirop-affiliate

Methods and Data The AECP group has developed a model to assess benefits of variable rate application in space for nitrogen in Swiss wheat production (Späti et al., 2021) and investigated farmers’ acceptance of such technologies with different resolutions (Späti et al., 2022). With respect to temporal aspects of decision-making, the group has also investigated why farmers might deviate from recommendations when to apply pesticides (Möhring et al., 2020). Overview of opportunities and policy aspects of digital innovations for sustainable agriculture have been provided (Finger, 2023; Finger et al., 2019). This literature serves as a basis for the development of a new, extended bio-economic model. This thesis is based on a collaboration with the ETH Spin-off Digit Soil. For example, data input for the parameterization of the model will be provided by Digit Soil.

Development and application of a bio-economic model to assess the costs and benefits of spatial and temporal information for improving nitrogen efficiency in agriculture.

Supervision Robert Finger Agricultural Economics and Policy ETH rofinger@ethz.ch Hélène Iven, CEO Digit Soil

Ideas for Analysis and Methods: Step 1: Define and identify over-performers – and the degree to which they overperform. A categorisation will need to be made, which could be binary, multi-category and/or continuous. Potentially trying all three as robustness checks. Step 2: Model which management/structural etc. factors affect the observed performance state at the intensive and extensive margin – i.e. underperformer, performer, overperformer or similar versus a continuum of performance where minimum standards for scheme = 0, lower/no adoption a negative value and higher/greater adoption a positive value. Step 3 (more consideration needed): Correct for the fact that farmers can still do the practices even if they do not enrol in the scheme. Exact methods TBD but a Heckman or double-hurdle model could be appropriate. For example, farms that do not enrol in the scheme and/or perform the practices just below the 60% required and/or do not perform the practices should also be considered and corrected for. Switching regressions could be something to explore. More work needed to define this (can be part of the Thesis process).

This analysis aims to examine the management characteristics and farm structural constraints that are correlated with exceeding minimum standards – and/or the degree to which minimum standards are exceeded – that must be met to receive payments for voluntary agri-environmental schemes. The case study would be based on the reduced tillage scheme, where to qualify, farmers must perform ploughless cultivation on at least 60% of their arable land using either mulch, strip tillage, direct drilling, or a combination of these methods. However, many farmers exceed the bare minimum levels needed to qualify and, as such, “over-perform” relative to the basic incentive. The analysis would utilise farmer characteristics, farm usage of complementary practices, and structural data to understand the determinants of “over-performance” adoption among farmers. This is an important policy question, as it may help determine which factors can be harnessed to improve the uptake and impact of soil conservation schemes. Additionally, this analysis could shed some light on aspects such as the additionality of current soil conservation schemes relative to the uptake behaviour of farmers, irrespective of the current scheme requirements. Potential inspiration / starting points from where we can add to with our data/analysis: - Farmer behavior toward herbicide-free agriculture and conservation tillage - Garcia - American Journal of Agricultural Economics - Wiley Online Library - Estimation and comparison of the performance of low-input and conventional agricultural production systems | Q Open | Oxford Academic - Pesticide-free agriculture: insights into farmer adoption across crops | Q Open | Oxford Academic (Documents attached)

Daily Supervisor: Charlie Rees (FiBL and ETH) charles.rees@fibl.org, Christian Grovermann (christian.grovermann@fibl.org), ETH supervisor: Robert Finger (rofinger@ethz.ch )

Gaëlle Leduc (gleduc@ethz.ch)

**Hintergrund:** Die Wertschöpfungskette von Nahrungsmitteln weist in der Schweiz eine X-Struktur auf. Die X-Struktur beschreibt eine Wertschöpfungskette in welcher viele Produzentinnen und Produzenten wenigen Verarbeitern beziehungswiese wenigen Händlern gegenüberstehen, diese aber wiederum viele Konsumentinnen und Konsumenten beliefern. In der Schweiz werden die Nahrungsmittel von vielen Landwirtschaftsbetrieben hergestellt. Sie entsprechen einer atomistischen Struktur. Das hat zur Folge, dass der einzelne Landwirt durch seine Produktionsmenge den Preis für das Produkt nicht bestimmen kann. Auch die Konsumentinnen und Konsumenten weisen eine atomistische Struktur auf. Niemand kann durch seine Nachfrage den Preis für die Nahrungsmittel beeinflussen. Demgegenüber stehen die Abnehmer und Händler. Diese umfassen nur wenige Akteure und ihre Entscheidungen über Mengen können die Preise beeinflussen. Dadurch kann Marktmacht entstehen. Die Politik greift in die Agrarmärkte ein, um ein besseres Gleichgewicht zwischen den Akteuren der Wertschöpfungskette zu schaffen. Aus der Kombination der Wettbewerbssituation (Anzahl Akteure und Organisation) und den agrarpolitischen Eingriffen entsteht eine sogenannte Marktordnung. Für unterschiedliche Agrarmärkte wie beispielsweise Kalb- oder Geflügelfleisch, Brotgetreide oder Zucker aber auch Äpfel oder Trauben bestehen unterschiedliche Marktordnungen. **Methodisches Vorgehen:** Die Bachelorarbeit besteht aus einer ausführlichen Literatur- und Datenrecherche zu einem ausgewählten Agrarmarkt in der Schweiz. Die Studentin oder der Student kann selber entscheiden, welchen Markt sie oder er untersuchen möchte.

Darstellung und ökonomische Interpretation der Marktordnung eines ausgewählten Schweizer Agrarmarktes.

Robert Huber rhuber@ethz.ch

Farm households increasingly combine on-farm production with non-agricultural activities, forming diverse income portfolios in response to market volatility, policy shifts, and spatial opportunity structures (Grilli et al., 2024; Finger & El Benni 2021; Meraner et al., 2015). Understanding the drivers of these diversification decisions is essential for designing effective agricultural and rural development policies. A growing body of evidence highlights the importance of the spatial context for farm diversification decisions. For example, Vroege et al. (2020) identify three spatial factors as particularly relevant. First, opportunity costs of diversification are location-specific. Second, farms situated near attractive landscapes tend to diversify more, especially into agri-tourism. Third, proximity to urban areas and the non-farm population strongly shapes the uptake of activities such as on-farm sales and care farming. Vroege et al. (2020) also demonstrate that spill-over effects between neighbouring farms are significant: positive spill-overs arise from cooperation and information-sharing within local farm networks, while negative spill-overs at the regional level reflect competition. Together, these findings underscore that diversification cannot be understood at the level of the individual farm alone. Yet one spatial dimension has received comparatively little attention in this literature, i.e., the role of proximity to processing infrastructure outside of farms. Processing facilities, such as slaughterhouses, milk processors, and feed and meat processors, shape the local value chain environment in which farms operate. Their presence may encourage specialisation within agricultural production (horizontal diversification) by lowering transaction costs and offering reliable output markets, or alternatively constrain farm-level diversification by locking farms into commodity supply chains. Whether and how the spatial configuration of processing infrastructure influences farm-household diversification strategies remains an open empirical question, particularly in a small, fragmented agricultural economy such as Switzerland. To ground the empirical analysis and identify the most relevant types of processing infrastructure and spatial clusters to examine, the thesis will begin with a non-systematic scoping review of the literature on spatial drivers of farm specialisation and diversification. Beyond proximity to generic processing facilities, Switzerland offers a particularly relevant institutional context, i.e., a dense network of products carrying protected designations of origin (AOP/GUB) and protected geographical indications (IGP/GGA), such as Gruyère, Emmentaler, or L'Etivaz. These AOP/IGP networks may constitute distinct spatial clusters of influence that systematically shape whether farms in a given region specialise to meet protected-designation requirements or, conversely, diversify away from commodity supply chains. The scoping review will identify such Switzerland-specific factors alongside evidence from other European contexts, informing both the product selection for the empirical analysis and the interpretation of results. You will also connect to policy instrumrtns in this field, such as regional development projects (Projekte zur regionalen Entwicklung (PRE)). Research questions The thesis is guided by three interrelated research questions: • How is proximity to processing facilities, including e.g., slaughterhouses, meat processors, milk processors, and animal feed producers, distributed across the Swiss agricultural landscape, and how has this distribution evolved between 2015 and 2023? • Does proximity to processing facilities increase the likelihood that farm households adopt specialised (i.e., horizontally concentrated) production strategies, (and does this relationship differ by type of processing facility)? • Are there spatial spill-over effects in strategy choice associated with processing proximity, that is, do farms in processing clusters show correlated diversification patterns beyond what individual farm characteristics would predict? Methods and Data Proximity to processing facilities is expected to influence diversification through two channels. First, access to nearby processors reduces transport and transaction costs for farm output, which may make specialised on-farm production more profitable, thereby reducing the incentive to diversify horizontally. Second, and analoguous to the cooperation effects documented for agri-tourism by Vroege et al. (2020), the presence of a local processing cluster may create agglomeration economies and knowledge spill-overs that reinforce specialisation patterns across farms in a region. At the same time, geographic isolation from processors may increase vertical diversification incentives, as farms seek to capture more value by processing and marketing their own products. This suggests that proximity effects may operate differently along the vertical diversification dimension, generating testable hypotheses about specialisation vs. self-processing trade-offs. To investigate these hypotheses, a comprehensive data set will be provided that covers the years from 2015 to 2023 and is derived from AGIS census data and FADN microdata including structural and socioeconomic variables that can be combined with georeferenced processing facility data from the Swiss Business and Enterprise Register under NOGA 2025 C10 (FSO STATENT, 2025). Given the breadth of product categories available in this register and the heterogeneity of Swiss agricultural production, the empirical analysis will focus on a selected set of processing types to be defined in consultation with the supervisors based on the scoping review findings. Likely candidates may include dairy processing (with a distinction between generic processors and AOP/IGP dairies), meat and slaughter facilities, and potentially feed producers. Spatio-temporal econometric analyses will be applied to assess the relationship between processing proximity and farm-household strategies, detecting regional clustering and potential spillover effects.

To emprically examine how proximity to processing facilities such as slaughterhouses, feed and meat processors or milk processors influences the farm household decision to diversify or specialize in on-farm production. You will combine various datasets and employ cutting edge econometric methods.

Supervision: The thesis is jointly supervised by the AECP group (Robert Finger, rofinger@ethz.ch) and Agroscope (Sabine Liebenehm, sabine.liebenehm@agroscope.admin.ch) References Finger, R., & El Benni, N. (2021). Farm income in European agriculture: new perspectives on measurement and implications for policy evaluation. European Review of Agricultural Economics, 48(2), 253-265. Grilli, G., Pagliacci, F., & Gatto, P. (2024). Determinants of agricultural diversification: What really matters? A review. Journal of Rural Studies, 110, 103365. Meraner, M., Heijman, W., Kuhlman, T., & Finger, R. (2015). Determinants of farm diversification in the Netherlands. Land Use Policy, 42, 767-780. Vroege, W., Meraner, M., Polman, N., Storm, H., Heijman, W., & Finger, R. (2020). Beyond the single farm–A spatial econometric analysis of spill-overs in farm diversification in the Netherlands. Land Use Policy, 99, 105019.

Society and policymakers increasingly focus on reducing pesticide use through more sustainable pest management strategies. Since 2009, the EU has promoted Integrated Pest Management (IPM) as a central approach under the Sustainable Use Directive, alongside organic farming (Helepciuc & Todor, 2021). IPM is based on eight guiding principles (Barzman et al., 2015) that encourage a flexible and integrated approach to pest control. However, farmers rarely implement all principles jointly, which may limit the effectiveness of this holistic pest management strategy (Lefebvre et al., 2015). For example, early preventive actions can strongly influence later pesticide needs (Steiro et al., 2020), highlighting the importance of stages and complementarities across practices. Nevertheless, IPM adoption is often measured using aggregate indices rather than distinguishing between its individual principles (e.g. Bagheri et al., 2019; Creissen et al., 2019). This limits understanding of which components are under-implemented by farmers and why. Existing research has identified several behavioural drivers of IPM adoption, including environmental attitudes, perceived control, and risk preferences (e.g. Despotović et al., 2019; Abdollahzadeh et al., 2024). However, important behavioural dimensions remain underexplored. This thesis will investigate how behavioural factors, including non-cognitive skills, risk and time preferences, objectives, and innovativeness, influence the adoption of IPM practices across different principles: prevention, monitoring, intervention, and evaluation. The analysis will be based on data from eight crops across ten EU countries, allowing for comparisons across perennial and annual systems and across national contexts. To account for interdependencies between IPM principles, seemingly unrelated regressions (SUR) will be employed. **References** Abdollahzadeh, G., Damalas, C. A., & Sharifzadeh, M. S. (2024). Integrated pest management adoption among citrus growers in Iran: an application of the protection motivation theory. Pest Management Science. https://doi.org/10.1002/ps.8358 Bagheri, A., Bondori, A., & Damalas, C. A. (2019). Modeling cereal farmers’ intended and actual adoption of integrated crop management (ICM) practices. Journal of Rural Studies, 70, 58–65. https://doi.org/10.1016/j.jrurstud.2019.05.009 Barzman, M., Bàrberi, P., Birch, A. N. E., Boonekamp, P., Dachbrodt-Saaydeh, S., Graf, B., Hommel, B., Jensen, J. E., Kiss, J., Kudsk, P., Lamichhane, J. R., Messéan, A., Moonen, A. C., Ratnadass, A., Ricci, P., Sarah, J. L., & Sattin, M. (2015). Eight principles of integrated pest management. Agronomy for Sustainable Development, 35(4), 1199–1215. https://doi.org/10.1007/s13593-015-0327-9 Creissen, H. E., Jones, P. J., Tranter, R. B., Girling, R. D., Jess, S., Burnett, F. J., Gaffney, M., Thorne, F. S., & Kildea, S. (2019). Measuring the unmeasurable? A method to quantify adoption of integrated pest management practices in temperate arable farming systems. Pest Management Science, 75(12), 3144–3152. https://doi.org/10.1002/ps.5428 Despotović, J., Rodić, V., & Caracciolo, F. (2019). Factors affecting farmers’ adoption of integrated pest management in Serbia: An application of the theory of planned behavior. Journal of Cleaner Production, 228, 1196–1205. https://doi.org/10.1016/j.jclepro.2019.04.149 Helepciuc, F. E., & Todor, A. (2021). Evaluating the effectiveness of the EU’s approach to the sustainable use of pesticides. PLoS ONE, 16(9), e0256719. https://doi.org/10.1371/journal.pone.0256719 Kreft, C., Huber, R., Wuepper, D., & Finger, R. (2021). The role of non-cognitive skills in farmers’ adoption of climate change mitigation measures. Ecological Economics, 189. https://doi.org/10.1016/j.ecolecon.2021.107169 Lefebvre, M., Langrell, S. R. H., & Gomez-y-Paloma, S. (2015). Incentives and policies for integrated pest management in Europe: a review. Agronomy for Sustainable Development, 35(1), 27–45. https://doi.org/10.1007/s13593-014-0237-2 Sharifzadeh, M. S., & Abdollahzadeh, G. (2024). Do typologies of pesticide risk knowledge influence the adoption of IPM strategies? Evidence from rice farmers’ behavior in northern Iran. Pest Management Science, 80(9), 4427–4436. https://doi.org/10.1002/ps.8151 Steiro, Å. L., Kvakkestad, V., Breland, T. A., & Vatn, A. (2020). Integrated Pest Management adoption by grain farmers in Norway: A novel index method. Crop Protection, 135. https://doi.org/10.1016/j.cropro.2020.105201 Vatn, A., Kvakkestad, V., Steiro, Å. L., & Hodge, I. (2020). Pesticide taxes or voluntary action? An analysis of responses among Norwegian grain farmers. Journal of Environmental Management, 276, 111074. https://doi.org/10.1016/j.jenvman.2020.111074 Wuepper, D., Roleff, N., & Finger, R. (2021). Does it matter who advises farmers? Pest management choices with public and private extension. Food Policy, 99. https://doi.org/10.1016/j.foodpol.2020.101995

The goal of this master thesis is to examine how behavioural factors influence farmers’ adoption of Integrated Pest Management practices across its different principles, using multi-country crop-level data and accounting for interdependencies between principles.

Gaëlle Leduc (gleduc@ethz.ch) & Robert Finger (rofinger@ethz.ch)

In unserer Gruppe Agrarökonomie und Agrarpolitik kommen unterschiedliche Optimierungsmodelle zum Einsatz, welche beispielsweise die Produktionsentscheide der Schweizer Landwirtinnen und Landwirte simulieren. Im Rahmen dieser Bachelorarbeit definieren Sie selber die Fragestellung aus aktuellen betriebswirtschaftlichen oder agrarpolitischen Herausforderungen für die Schweizer Landwirtschaftsbetriebe. Anschliessend erstellen Sie ein kleines Optimierungsmodell, welches exemplarisch die Fragestellung beantwortet. Sie dokumentieren und interpretieren die Resultate und diskutieren die Ergebnisse mit Bezug auf die methodischen Grundlagen. Durch die Erstellung und Nutzung des Modells lernen Sie den Umgang mit Optimierungs- und Simulationstools kennen.

Erstellung und Anwendung eines kleinen Optimierungsmodells im Kontext einer eigenen Forschungsfrage.

Robert Huber rhuber@ethz.ch

Society and policymakers have paid particular attention to sustainable pest management and pesticide use. A key aspect of this is the economics of pest management and pesticide use. This includes the financial implications of different policies. Furthermore, how farmers respond to fluctuating pesticide prices is essential, particularly when considering pesticide taxation (e.g., Finger & Pedersen, 2025; Ayouba & Vigeant, 2020). Such taxes on pesticides are currently under discussion in Swiss agricultural policy and in other countries (Nielsen et al., 2023; Devilliers et al., 2025). However, the available evidence on farmers' price inelasticity to changing pesticide prices is outdated and limited in scope (e.g., Böcker & Finger, 2017). This bachelor's thesis will provide new and updated evidence on the price elasticity of pesticides. It will offer valuable insights for agriculture, industry, and policy. To this end, a systematic review was conducted following the PRISMA guidelines (Page et al., 2023) to identify relevant studies in Europe and comparable regions. Then, the thesis will use the identified studies to conduct a meta-analysis with the identified price elasticity estimates. Additionally, the thesis will perform tests on the elasticity of demand for pesticides to determine how it differs across crops, types of pesticides, countries, and time. References Ayouba, K., & Vigeant, S. (2020). Can we really use prices to control pesticide use? Results from a nonparametric model. Environmental Modeling & Assessment, 25(6), 885-900. Böcker, T. G., & Finger, R. (2017). A meta‐analysis on the elasticity of demand for pesticides. Journal of Agricultural Economics, 68(2), 518-533. Devilliers, E., Koutchadé, O.P. and Carpentier, A., 2025. Pesticide taxation and the adoption of low-input cropping management practices using a hidden Markov model. European Review of Agricultural Economics, 52(4), pp.818-847. Finger, R., & Pedersen, A. B. (2025). Input taxes in agriculture: Experiences and perspectives for European agriculture. Ecological Economics, 233, 108575. Nielsen, H. Ø., Konrad, M. T. H., Pedersen, A. B., & Gyldenkærne, S. (2023). Ex-post evaluation of the Danish pesticide tax: a novel and effective tax design. Land Use Policy, 126, 106549. Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., ... & Moher, D. (2021). The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. bmj, 372.

This bachelor's thesis will provide new and updated evidence on the price elasticity of pesticides.

Robert Finger (rofinger@ethz.ch) & Gaelle Leduc (gleduc@ethz.ch)

Der Anbau von Zwetschgen spielt im Schweizer Obstbau eine wichtige Rolle, wobei sich die Produktion bisher stark auf früh- und mittelfrühe Sorten konzentriert. Der Anbau von späten bzw. späteren Zwetschgensorten könnten jedoch neue Marktchancen eröffnen, indem sie Angebotslücken gegen Ende der Saison schliessen. Gleichzeitig sind mit dem Anbau später Sorten spezifische agronomische Risiken verbunden, etwa durch Witterungseinflüsse, Krankheitsdruck oder einen ungünstigen Erntezeitpunkt. Vor diesem Hintergrund ist es das Ziel der Bachelorarbeit, das Marktpotenzial später Zwetschgensorten in der Schweiz systematisch zu analysieren und zu untersuchen, unter welchen wirtschaftlichen Szenarien sich eine Investition lohnt. Somit wird sowohl die Perspektive des landwirtschaftlichen Betriebs als auch die generelle Marktperspektive eingenommen. Die Arbeit liefert eine wichtige Entscheidungsgrundlage für Produzenten und die Beratung im Steinobstbau der Schweiz.

Für die Abschätzung des Marktpotenzials soll sich die Bachelorarbeit auf Import- und Produktionsdaten sowie auf eine Literaturrecherche stützen. Diese Marktanalyse wird anschliessend durch Investitionsrechnungen ergänzt, die Szenario- und Risikoanalysen beinhalten. Hier dient Reseau-lution (https://reseau-lution.agridea.ch/) von AGRIDEA als Grundlage. Die Szenario- und Risikoanalysen sollen durch die Literatur motiviert werden und können Aspekte wie Ertragsstabilität, Preiserwartungen, Krankheitsanfälligkeit, Erntefenster und Unterschiede Schweizer Anbaustandorte beinhalten. Grundlagen und Tools für die Analysen werden aus dem Kurs ‘Produktion, Investition und Risikomanagement in der Landwirtschaft’ https://www.vorlesungen.ethz.ch/Vorlesungsverzeichnis/lerneinheit.view?lang=de&semkez=2026S&lerneinheitId=198662& übernommen (Details teilen wir dann mit Ihnen).

Kontaktperson: Robert Finger (rofinger@ethz.ch) und Sergei Schaub (sergei.schaub@agroscope.admin.ch). Die Bachelorarbeit ist eine Zusammenarbeit zwischen der AECP Gruppe und Agroscope. Weitere Fachpersonen von Agroscope werden gezielten Input geben.

Accurate prediction of market volumes plays a critical role in agricultural markets by enabling producers, traders, and policymakers to better align supply with expected demand. This can help reduce price volatility, minimize food losses, and improve the overall efficiency of the supply chain. Additionally, it can support policy decisions about import quotas and timing of market protection measures. The importance of reliable forecasts is particularly pronounced in the market for berries and stone fruits, which is characterized by short harvest windows and high perishability. While studies have focused on on-field and -farm yield prediction for berries and stone fruits (e.g., Magro et al., 2023), research on market volume forecasting remains scarce. This thesis aims to address this gap by comparing different prediction approaches (such as machine learning methods and classical econometric time series models) for forecasting berry and stone fruit market volumes (see, e.g., Storm et al., 2020; Brignoli et al., 2024). These approaches will also be benchmarked against realized market volumes and currently used industry forecasts. Given the inherent riskiness and high economic value of berry production, particular attention will be paid to how well the models predict extreme outcomes (i.e., very high or very low market volumes) (see, e.g., Barbaglia et al., 2023). Moreover, the thesis may explore how specific determinants—such as weather forecasts and online indicators (e.g., Google Trends)—can improve prediction accuracy. The analysis will draw on weekly market volume data provided by the Schweizer Obstverband (SOV) such as weekly market volume and weekly market prognosis from https://reports.swissfruit.ch/, complemented by additional datasets including weather (forecast) data, Google Trends data, import statistics, holiday calendars, price information, and yield records. This thesis is jointly supervised by the AECP group and Agroscope.

References Barbaglia, L., Consoli, S., & Manzan, S. (2023). Forecasting with economic news. Journal of Business & Economic Statistics, 41(3), 708-719. Brignoli, P. L., Varacca, A., Gardebroek, C., & Sckokai, P. (2024). Machine learning to predict grains futures prices. Agricultural Economics, 55(3), 479-497. Storm, H., Baylis, K., & Heckelei, T. (2020). Machine learning in agricultural and applied economics. European Review of Agricultural Economics, 47(3), 849-892. Magro, R. B., Alves, S. A. M., & Gebler, L. (2023). Computational models in Precision Fruit Growing: reviewing the impact of temporal variability on perennial crop yield assessment. SN Computer Science, 4(5), 554.

Contact: Robert Finger (rofinger@ethz.ch) and Sergei Schaub (sergei.schaub@agroscope.admin.ch)

This project is set to limited visibility by its publisher. To see the project description you need to log in at SiROP. Please follow these instructions: - Click link "Open this project..." below. - Log in to SiROP using your university login or create an account to see the details. If your affiliation is not created automatically, please follow these instructions: http://bit.ly/sirop-affiliate