The Market Price Database (MPDB) – The foundation of data-driven energy processes.

One of the highlights of BelVis⁺ PFM is the new market price database. Product manager Alla Prokhorova and bid manager Rouven Voß explain in an interview how it gives users new freedom in handling market data.

What exactly is the market price database (MPDB) and what is the idea behind it?

Alla Prokhorova: The MPDB is the central hub for price and market data within the company – it consolidates all relevant information in a single, reliable database that is modelled along the typical processes in the energy market. This allows us to avoid duplication, inconsistencies and media breaks and create a ‘single source of truth’ for portfolio, trading, risk management and reporting processes – both within and outside the BelVis+ world. The design goal was a transparent, easily integrable market price management system that makes prices, quotas, EoD and live PFCs manageable for relevant energy markets.

Rouven Voß: Architecturally, the MPDB is an API-driven, highly automated data platform. It imports, validates, versions and merges time series; it shifts PFCs, monitors jobs and exports data to connected systems. The open role of the MPDB is important: it is available as a central company database. BelVis+ PFM can be ‘just’ one of several data consumers of the MPDB – as can SAP, optimisation systems, BI tools, market models in Matlab or other target systems.

Which MPDB USPs do you consider particularly crucial for customers?

Prokhorova: Our customers repeatedly emphasise the same points:

• Centralised, audit-proof database: Everything is historicised, versioned and stored in a traceable manner. No ‘Excel history’, no uncertainties.
• Technical logic is already built in: products, markets, trading calendars, time series types – the system understands energy relationships instead of just storing numbers.
• Automation across the entire process: import, PFC shifting, plausibility checks, replacement value creation, versioning – everything can be automated.
• Thanks to standard interfaces to all important data sources, the MPDB can be set up and ready to use in just a few hours.
• Easy expandability: Additional data sources or in-house systems can be easily connected via Python scripts, both for data import and export.

Voß: My customers and I are particularly impressed by the user-friendliness of MPDB: in addition to all the great API technology for importing and exporting data, every user can quickly and easily export any data as a CSV file and thus make it available for Excel or other target systems. It sounds like a quick and dirty solution, but sometimes that’s exactly what you need in your day-to-day work.

How do imports work – and why do you combine API and scripts?

Prokhorova: Our standard method is API-based connection to data sources. At the same time, we offer standardised and customised import scripts (e.g. for futures, spot and PFCs) that can be configured and automated via the UI. There is a reason for this: formats – for example, at EEX/EPEX – change occasionally. With scripts, customers remain decoupled from the MPDB release cycle and can compensate for format deviations at short notice. However, MPDB also contains standard scripts that do not need to be customised, as they allow users to configure the desired import or export format themselves via the GUI.

Voß: In operational terms, this means time- and event-triggered processes, transparent activity logs with the results of each step, and document services for exported or downloaded data, including (re-)import. This ensures that the entire data flow is audit-proof and traceable – even in an audit.

What happens if data is incorrect – e.g. incorrect settlement prices?

Prokhorova: In the MPDB, documents/data files can be searched for, edited and released in a targeted manner; then the automation process is restarted. This correction process can be completed with just a few clicks. Users can set up an automated export of the relevant market data, which regularly sends them an overview of the imported quotas by e-mail for checking and correction.

Voß: In addition, plausibility checks and, if necessary, replacement values are applied automatically. The user can choose whether to generate a new version of the time series or overwrite the original time series. There are extended checks for PFCs (e.g. reference tag comparison); each run generates a PFC plausibility status, which is taken into account during release.

What role do shifting processes and PFC mark-ups play?

Prokhorova: PFC shifting makes it possible to transfer historical price trends to current market conditions – an essential tool for traders, portfolio managers and pricing analysts. In a dynamic market environment, yesterday’s PFC or even this morning’s PFC is outdated anyway. If you shift this PFC every five minutes in line with real-time prices and make it available to sales or risk management, you are simply closer to the market and have a clearer valuation or fairer prices.

Voß: PFC surcharges complement this logic – you can set surcharges for different time segments, whether fixed, percentage-based or using predefined and versioned surcharge time series. Together, these two features make the MPDB a real pricing building block for many customers.

How do users work with the market price database?

Prokhorova: At its heart is a very clear, configurable web dashboard. Here, you can monitor imports or visualise various price trends.

All processes are documented in a fully auditable manner: in the activity log, you can see every automated step – imports, plausibility checks, shifting, replacement calculations, exports.

Voß: The MPDB is an open system that can be easily integrated into data lakes or BI tools via REST or configured using Python scripts. This makes the MPDB accessible to all areas of the company, regardless of the software they use. Integration is extremely flexible.

What is your personal favourite feature?

Prokhorova: For me, it’s definitely the consolidated view of all versions of a PFC – with status, history, plausibility checks, approvals and curve comparisons. That’s something you want when you’ve been working with Excel PFCs for many years.

Voß: For me, it’s the automation stack: the fact that a customer can import, validate and distribute 20,000 time series per day without having to intervene manually shows how powerful this module has become.