November Meetup: Financial Data & Money Laundering

Another month, another PyData Amsterdam meetup! This time hosted by Rabobank in the Mondriaantower next to the Amsterdam Amstel station; that means a meetup with a view, obviously good pizza to start, drinks, awesome talks, and more drinks plus warm snacks afterward!

For this evening we have two awesome financial talks. In the first talk, we'll hear about how network analytics are applied to detect money laundering in combination with a hands-on demo on graph analytics. The second talk will be on demystifying financial statement data. We're looking forward to the meetup and hope to see you there!

SCHEDULE
18:00-19:00: Welcome (🍕 & 🍺/🧃!)
19:00-19:45: Talk 1 - Applying Network Analytics in KYC - Salomon Tetelepta
19:45-20:00: break
20:00-20:45: Talk 2 - Econometrics meets accounting: what can PCA teach us about financial statement data? - Cor Zuurmond & Mick van Rooijen
20:45-21:30: Networking / drinks & warm snacks!

TALK 1
“Applying Network Analytics in KYC”
At Rabobank, we started using network analytics to detect money laundering. In this presentation we will show how we did a pilot project where we created network features to detect risk. We will discuss the graph data model, how we apply graph queries, and which algorithms can be run on these graphs. Besides, we will go through a more hands-on demo that hopefully shows a bit more in depth how you can get started with applying graph analytics yourself using python and Neo4j.

Bio Salomon Tetelepta
Salomon Tetelepta is a Data Scientist at the Rabobank with a focus on the KYC domain. He has experience with applying Machine Learning for fraud detection and customer risk scoring. Current focus is to find Money Laundering at scale using graph data science with Neo4j.

TALK 2
“Econometrics meets accounting: what can PCA teach us about financial statement data?"
Financial statement data covers nearly 3,000 different variables, which calls for a dimension reduction technique. During this talk, Mick and Cor explain what principal component analysis (PCA) can teach us about financial statement data.

The research
First, we construct several factors from the high-dimensional dataset using Principal Component Analysis. After that, we use the PCS in a factor-augmented predictive regression model to forecast movements in stock prices.

We find several interesting features. First, the 6 PCs can explain about two-thirds of the variation in the entire dataset, which is relatively high for PCs constructed from financial data. Second, the Root Mean Squared Error (RMSE) of the predictions made by our model is about 35% lower than the RMSE of a benchmark random walk model, which indicates that the factors could be associated with changes in stock prices.

The data
Ritoku's Japanese Company Fundamentals data feed contains fundamentals data for over 4,500 Japanese companies, including over 3,750 publicly listed companies and over 2,500 unique fundamental indicators.

This data can be used to make investment decisions, gain strategic insight or make policy-related decisions related to Japanese companies. This data feed originates from the Japanese Financial Services Agency (a governmental supervisory body for financial markets in Japan). History starts with financial statements published after February 2014.

Bio Mick van Rooijen
Mick currently works at one of the biggest Dutch financial institutions. He hold a Master's in Banking & Finance from Utrecht University and another in Econometrics from VU Amsterdam. He is passionate about applying econometric principles to economic and financial data to understand how stuff works.

Bio Cor Zuurmond
Cor is the founder of Ritoku. With Ritoku, he sells the Japanese company fundamentals on Nasdaq's data platform. In addition, Cor built the data system required to extract high-quality financial data originating from XBRL documents published by companies.

Besides Ritoku, Cor is demystifying data at GoDataDriven. At GoDataDriven, he builds Azure data platforms and implements data use cases with our clients.

DIRECTIONS
Hosted by Rabobank in the Mondriaantower, which is located next to the Amsterdam Amstel station at Amstelplein 8. Please enter via the main entrance (revolving doors) and announce yourself at the reception. The security host will probably ask you to take place in the lounge area, after which you'll be picked up in small groups by one of the Rabobank hosts, who will guide you to the 12th floor, where the meetup is organised.