The law of one price is a fundamental building block of modern financial theory. Indeed, several well-known pricing relations in finance depend on the notion that two traded or synthesized instruments with the same future cash flows should trade at the same price. In practice, of course, market microstructure aspects may cause a temporary deviation of prices from their no-arbitrage values. Academic work has long recognized that deviations from no-arbitrage relations are related to the frictions associated with transacting, in particular to illiquidity and trading costs such as the bid-ask spreads and tick sizes. In the stock and stock futures market, deviations from no-arbitrage relations should be related to market liquidity because liquidity facilitates arbitrage. At the same time, a wide futures-cash basis may trigger arbitrage trades and, in turn, affect liquidity. The aim of the project is to study the dynamic relationship between stock market liquidity and the index futures basis.

Vincent Lars Wolff

The Law of One Price and and Stock Market Liquidity

Financial Economics

Text classification models are often fooled by adversarial examples, which are small changes to the original text that are not noticeable to humans but cause the model to make incorrect predictions. Previous research in natural language processing has used rule-based techniques to generate adversarial examples by replacing words with synonyms. However, these methods can result in unnatural and confusing changes to the text that are easily detected by humans. The BAE (https://arxiv.org/abs/2004.01970) approach is a black box method for creating adversarial examples by masking parts of the text and using a BERT masked language model to generate alternative words, producing adversarial examples that are more grammatically correct and semantically coherent than previous methods. The student is asked to develop an approach that may lead to adversarial examples in the financial domain, particularly exploring the vulnerability of FinBERT. For this thesis, prior knowlege on natural language processing, transformer methods and the GPT family is required. 