Skip to content
All library documents

QuantNet Transfers Market Trends Across Equity Trading Strategies

Article arXiv papers · Author: Adriano Koshiyama et al.
Read the original

Summary

QuantNet is a learning architecture for building trading strategies across multiple equity markets. Each market has its own encoder and decoder: the encoder represents local market data, a shared global model processes information across markets, and the decoder combines local and global signals to produce a market-specific strategy. This design aims to capture transferable patterns while letting each market’s model adapt to its own characteristics.

The paper evaluates the approach on historical data for 3,103 assets across 58 equity markets. It reports higher Sharpe and Calmar ratios than both its best-performing baseline and a version without transfer learning. The abstract does not describe the evaluation period, transaction costs, or implementation details, so the reported comparisons alone do not establish how the strategy would perform live or in other settings.

Key ideas

  • QuantNet combines market-specific encoders and decoders with a global model shared across markets.
  • The global component is intended to learn common trading patterns from multiple markets.
  • Market-specific parameters can adapt the strategy to local data.
  • The study reports improved Sharpe and Calmar ratios against a baseline and a non-transfer variant.
  • The abstract does not specify costs, evaluation period, or live-trading results.

Tags

EquitiesMachine learningStatisticsBacktesting
Full text
# QuantNet: Transferring Learning Across Systematic Trading Strategies


# QuantNet: Transferring Learning Across Systematic Trading Strategies









Systematic financial trading strategies account for over 80% of trade volume in equities and a large chunk of the foreign exchange market. In spite of the availability of data from multiple markets, current approaches in trading rely mainly on learning trading strategies per individual market. In this paper, we take a step towards developing fully end-to-end global trading strategies that leverage systematic trends to produce superior market-specific trading strategies. We introduce QuantNet: an architecture that learns market-agnostic trends and use these to learn superior market-specific trading strategies. Each market-specific model is composed of an encoder-decoder pair. The encoder transforms market-specific data into an abstract latent representation that is processed by a global model shared by all markets, while the decoder learns a market-specific trading strategy based on both local and global information from the market-specific encoder and the global model. QuantNet uses recent advances in transfer and meta-learning, where market-specific parameters are free to specialize on the problem at hand, whilst market-agnostic parameters are driven to capture signals from all markets. By integrating over idiosyncratic market data we can learn general transferable dynamics, avoiding the problem of overfitting to produce strategies with superior returns. We evaluate QuantNet on historical data across 3103 assets in 58 global equity markets. Against the top performing baseline, QuantNet yielded 51% higher Sharpe and 69% Calmar ratios. In addition we show the benefits of our approach over the non-transfer learning variant, with improvements of 15% and 41% in Sharpe and Calmar ratios. Code available in appendix.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.