Stock return predictability is an important research theme as it reflects our economic and social organization, and significant efforts are made to explain the dynamism therein. Statistics of strong explanative power, called "factor" have been proposed to summarize the essence of predictive stock returns. Although machine learning methods are increasingly popular in stock return prediction, an inference of the stock returns is highly elusive, and still most investors, if partly, rely on their intuition to build a better decision making. The challenge here is to make an investment strategy that is consistent over a reasonably long period, with the minimum human decision on the entire process. To this end, we propose a new stock return prediction framework that we call Ranked Information Coefficient Neural Network (RIC-NN). RIC-NN is a deep learning approach and includes the following three novel ideas: (1) nonlinear multi-factor approach, (2) stopping criteria with ranked information coefficient (rank IC), and (3) deep transfer learning among multiple regions. Experimental comparison with the stocks in the Morgan Stanley Capital International (MSCI) indices shows that RIC-NN outperforms not only off-the-shelf machine learning methods but also the average return of major equity investment funds in the last fourteen years.

We propose a novel approach to sentiment data filtering for a portfolio of assets. In our framework, a dynamic factor model drives the evolution of the observed sentiment and allows to identify two distinct components: a long-term component, modeled as a random walk, and a short-term component driven by a stationary VAR(1) process. Our model encompasses alternative approaches available in literature and can be readily estimated by means of Kalman filtering and expectation maximization. This feature makes it convenient when the cross-sectional dimension of the sentiment increases. By applying the model to a portfolio of Dow Jones stocks, we find that the long term component co-integrates with the market principal factor, while the short term one captures transient swings of the market associated with the idiosyncratic components and captures the correlation structure of returns. Finally, using quantile regressions, we assess the significance of the contemporaneous and lagged explanatory power of sentiment on returns finding strong statistical evidence when extreme returns, especially negative ones, are considered.

We analyze what can be learned from tests for p-hacking based on distributions of t-statistics and p-values across multiple studies. We analytically characterize restrictions on these distributions that conform with the absence of p-hacking. This forms a testable null hypothesis and suggests statistical tests for p-hacking. We extend our results to p-hacking when there is also publication bias, and also consider what types of distributions arise under the alternative hypothesis that researchers engage in p-hacking. We show that the power of statistical tests for detecting p-hacking is low even if p-hacking is quite prevalent.

Thousands of cryptocurrencies have been issued and publicly exchanged since Bitcoin was invented in 2008. The total cryptocurrency market value exceeds 300 billion US dollars as of 2019. This paper analyzes the prices, volumes, blockchain transactions, coin difficulties and public opinion popularities of 3607 actively exchanged cryptocurrencies. We aim to reveal and explain the homogeneity, i.e., the strong correlation of market performance, and the heterogeneity, i.e., the imbalance of popularities and sophistications, of the cryptocurrencies.

We study a dynamic portfolio optimization problem related to convergence trading, which is an investment strategy that exploits temporary mispricing by simultaneously buying relatively underpriced assets and selling short relatively overpriced ones with the expectation that their prices converge in the future. We build on the model of Liu and Timmermann (2013) and extend it by incorporating unobservable Markov-modulated pricing errors into the price dynamics of two co-integrated assets. We characterize the optimal portfolio strategies in full and partial information settings both under the assumption of unrestricted and beta-neutral strategies. By using the innovations approach, we provide the filtering equation that is essential for solving the optimization problem under partial information. Finally, in order to illustrate the model capabilities, we provide an example with a two-state Markov chain.

Hutchinson, Lo and Poggio raised the question that if learning works can learn the Black-Scholes formula, and they proposed the network mapping the ratio of underlying price to strike $S_t/K$ and the time to maturity $\tau$ directly into the ratio of option price to strike $C_t/K$. In this paper we propose a novel descision function and study the network mapping $S_t/K$ and $\tau$ into the ratio of time value to strike $V_t/K$. Time values' appearance in artificial intelligence fits into traders' natural intelligence. Empirical experiments will be carried out to demonstrate that it significantly improves Hutchinson-Lo-Poggio's original model by faster learning and better generalization performance. In order to take a conceptual viewpoint and to prove that $V_t/K$ but not $C_t/K$ can be approximated by superpositions of logistic functions on its domain of definition, we work on the theory of universal approximation on unbounded domains. We prove some general results which imply that an artificial neural network with a single hidden layer and sigmoid activation represents no function in $L^{p}(\RR^2 \times [0, 1]^{n})$ unless it is constant zero, and that an artificial neural network with a single hidden layer and logistic activation is a universal approximator of $L^{2}(\RR \times [0, 1]^{n})$. Our work partially generalizes Cybenko's fundamental universal approximation theorem on the unit hypercube $[0, 1]^{n}$.

Financial crime is a large and growing problem, in some way touching almost every financial institution. Financial institutions are the front line in the war against financial crime and accordingly, must devote substantial human and technology resources to this effort. Current processes to detect financial misconduct have limitations in their ability to effectively differentiate between malicious behavior and ordinary financial activity. These limitations tend to result in gross over-reporting of suspicious activity that necessitate time-intensive and costly manual review. Advances in technology used in this domain, including machine learning based approaches, can improve upon the effectiveness of financial institutions' existing processes, however, a key challenge that most financial institutions continue to face is that they address financial crimes in isolation without any insight from other firms. Where financial institutions address financial crimes through the lens of their own firm, perpetrators may devise sophisticated strategies that may span across institutions and geographies. Financial institutions continue to work relentlessly to advance their capabilities, forming partnerships across institutions to share insights, patterns and capabilities. These public-private partnerships are subject to stringent regulatory and data privacy requirements, thereby making it difficult to rely on traditional technology solutions. In this paper, we propose a methodology to share key information across institutions by using a federated graph learning platform that enables us to build more accurate machine learning models by leveraging federated learning and also graph learning approaches. We demonstrated that our federated model outperforms local model by 20% with the UK FCA TechSprint data set. This new platform opens up a door to efficiently detecting global money laundering activity.