Synthetic data
A mix of Gaussian distributions can beat GenAI at its own game
Synthetic data is seen as the preserve of AI models. A new paper shows old methods still have legs
Filling gaps in market data with optimal transport
Julius Baer quant proposes novel way to generate accurate prices for illiquid maturities
Digging deeper into deep hedging
Dynamic techniques and GenAI simulated data can push the limits of deep hedging even further, as derivatives guru John Hull and colleagues explain
Predicting financial distress of Chinese listed companies using a novel hybrid model framework with an imbalanced-data perspective
In this paper a novel hybrid model framework is constructed to solve the problem of predicting the financial distress of Chinese listed companies using imbalanced data.
Quants turn to machine learning to unlock private data
Replication could allow financial firms to use – and monetise – data that was previously off-limits
Synthetic data enters its Cubist phase
Quants are using the theory of rough paths to distil the essence of financial datasets
Generating financial markets with signatures
Signatures can provide the synthetic data to train deep hedging strategies
Fake data can help backtesters, up to a point
Synthetic data made with machine learning will struggle to capture the caprice of financial markets
In fake data, quants see a fix for backtesting
Traditionally quants have learnt to pick data apart. Soon they might spend more time making it up
The economic cost of a fat finger mistake: a comparative case study from Samsung Securities’s ghost stock blunder
This paper quantifies the economic cost of Samsung Securities’s ghost stock blunder using the synthetic control method.
Podcast: Horvath and Lee on market generator models
Quants explain the application of the latest techniques
Podcast: Kondratyev and Schwarz on generating data
Market generator models may aid areas of finance where data is limited or sensitive
The market generator
A generative neural network is proposed to create synthetic datasets that mantain the statistical properties of the original dataset