Piątek, 14 kwietnia 2023
- Prowadzący: Philippe J.S. De Brouwer, PhD., HSBC
Temat: Do we matter?
Opis: A lecture that brings chaos theory to life. We demonstrate how simple systems can lead to complex chaotic behaviour and use this knowledge to explore concepts such as the impact of small actions (aka the butterfly effect). We also illustrate this on business. - Prowadzący: Mateusz Pabian, PhD., UBS
Temat: A few things I wish somebody told me before I became a Data Scientist
Opis: What’s the difference between a Data Engineer, Data Scientist, Machine Learning Engineer? What is the scope of our role in the process of creating a data-driven product? In this lecture you will learn about the stages of development & maintenance of a data-driven product, the role of model testing & periodic validation that guide the model towards robustness and the technology stack & skill set used in day-to-day activities, in the context of speech recognition for smart-home applications. - Prowadzący: Marcin Jaskowski, PhD., HSBC
Temat: Machine Learning and Ethics
Opis: In this presentation, I argue that Machine Learning (ML) is supplementary to Econometrics – it is a new toolbox but at the same time, the domain of its applicability is different than for the Econometrics.ML appears now to be easy to use, as convenient packages are readily available. One can now easily estimate random forests, LASSO regressions and artificial neural networks on any type of data. But, this raises the risk that ML methods might be applied naively and their output might be misunderstood. We will try to explain where ML methods are more likely to excel (cross-sectional problems where data is abundant) and where ML methods might be less applicable (parameter estimation) or might lead to over-fit (time-series). - Prowadzący: Gabriele Giraldo i Sofia-Allegra Minuti, UBS
Temat: Credit Risk
Opis: The presentation will focus on Credit Risk Modelling. We will look at its fundamentals, definition and its purpose in the past and in the present times. We will present what is a credit risk model and how we can deal with it, starting from the data sample, passing through the model implementation finally getting to the analysis and interpretation of the output. - Prowadzący: Wojciech Tarnowski, PhD., State Street Global Advisors
Temat: The Zillow Failure
Opis: Zillow was a platform connecting real estate buyers with the sellers by listing their offers. In 2018 Zillow entered housing market as a market maker and started flipping houses using a machine learning model for estimation of house price. In the talk we will review the challenges and pitfalls in using ML models and discuss key factors that contributed to Zillow failure in 2021. - Prowadząca: dr hab. Elżbieta Kubińska, prof. UEK
Temat: Homo Heuristicus w świecie matematyki
Opis: Wykład jest próbą naszkicowania postaci Homo Heuristicus, który nie zawsze poprawnie interpretuje napływające informacje, w szczególności informacje określające parametry inwestycji, będącymi alternatywami w warunkach ryzyka. Neoklasyczne teorie finansów zakładają, że zarówno rynek jak i inwestorzy są doskonale racjonalni. Wynki finansów behawioralnych wskazują, że inwestorzy ulegają błędom poznawczym, które mogą prowadzić do systematycznie błędnych decyzji, co jest wyzwaniem stojącym przed modelami matematyki finansowej. - Prowadzący: Konrad Maślankiewicz, BM PKO BP
Temat: Ochrona kapitału w czasie wysokiej inflacji
Opis:
Sobota, 15 kwietnia 2023
- Prowadzący: Paweł Fronczak, NBP
Temat: Zarządzanie portfelem obligacji okiem praktyka
Opis: Prezentacja ma na celu przybliżenie praktycznych aspektów związanych z budową i zarządzaniem portfelem obligacji w USD. Podczas spotkania omówione zostanie ryzyko stopy procentowej i miary z nim związane. Ponadto przedstawione zostaną sposoby analizy strategii inwestycyjnych oraz rynku. Na koniec krótkie umówienie ostatnich wydarzeń na rynku amerykańskim i zachowań rynkowych z tym związanych. - Prowadzący: Artur Zając, HSBC
Temat: Stress Testing in Investment Banking
Opis: The presentation will describe recent developments in Stress Testing regulatory framework, and what are its capital implications for large investment banks. We will discuss example regulatory scenarios that have been considered in Stress Testing exercises. Moreover, some aspects of Stress Testing methodology and execution will be covered. - Prowadzący: Marcin Izbrandt i Tomasz Jachymiak, UBS
Temat: Risk management based on stress testing
Opis: The presentation will focus on stress testing exercises. Going through the overview of the history of stress testing and its possible future enhancements, we’ll look at the usage of the stress testing results and its importance in the risk management and financial stability. We’ll show how UBS is making use of stress testing results and how daily work of model developer is impacting business decisions of senior management. During the presentation we’ll also show a general example of Expected Tail Loss (ETL) simulations in a stress model, explain what issues may arise when it comes to the conceptual soundness, and share examples on how to tackle them. - Prowadzący: Katarzyna Piechnik i Dawid Skórka, Accenture
Temat: Job in consulting – EWI, FRTB case studies
Opis: During the presentation we will reveal how a usual day of a risk consultant looks like and how this job differs from working as a risk specialist in a bank or other company. Based on two examples, we will shed light on what kind of topics we help our clients with and most importantly where quantitative analysts play a key role. - Prowadzący: Urszula Imosa i Jacek Gronowski, State Street
Temat: State Street Portfolio & Risk Analytics Department – introduction
Opis: Have you ever wondered how market risk analytics are calculated and reported for the World’s biggest asset owners and asset managers? State Street Portfolio & Risk Analytics Department in Poland in cooperation with our colleagues in Dublin, Singapore, Frankfurt, Toronto and New York is doing that for clients across the Globe. State Street Risk Analytics combines unique capabilities in data and analytics to better serve our clients and build relationships with new customers. State Street Risk Analytics’ truView platform is a web-based application providing a full suite of analytics that give clients insight into the risks and performance of their investment portfolios. Supporting all asset classes the platform produces analytics such as: stress testing, Value at Risk (VaR), position specific analytics for derivatives and fixed income characteristics.