Text regression has traditionally been tackled using linear models. Here we present a non-linear method based on a deep convolutional neural network. We show that despite having millions of parameters, this model can be trained on only a thousand documents, resulting in a 40% relative improvement over sparse linear models, the previous state of the art. Further, this method is flexible allowing for easy incorporation of side information such as document meta-data. Finally we present a novel technique for interpreting the effect of different text inputs on this complex non-linear model.

Stock price movements are claimed to be chaotic and unpredictable, and mainstream theories of finance refute the possibility of realizing risk-free profit through predictive modelling. Despite this, a large body of technical analysis work maintains that price movements can be predicted solely from historical market data, i.e., markets are not completely efficient. In this paper we seek to test this claim empirically by developing a novel stochastic trading algorithm in the form of a linear model with a profit maximization objective. Using this method we show improvements over the competitive buy-and-hold baseline over a decade of stock market data for several companies. We further extend the approach to allow for non-stationarity in time, and using multi-task learning to modulate between individual companies and the overall market. Both approaches further improve the predictive profit. Overall this work shows that market movements do exhibit predictable patterns as captured through technical analysis.

Peer-to-peer lending is a new highly liquid market for debt, which is rapidly growing in popularity. Here we consider modelling market rates, developing a non-linear Gaussian Process regression method which incorporates both structured data and unstructured text from the loan application. We show that the peer-to-peer market is predictable, and identify a small set of key factors with high predictive power. Our approach outperforms baseline methods for predicting market rates, and generates substantial profit in a trading simulation.