Heuristic methods for stock selection and allocation in an index tracking problem

Index tracking is one of the most popular passive strategy in portfolio management. However, due to some practical constrains, a full replication is difficult to obtain. Many mathematical models have failed to generate good results for partial replicated portfolios, but in the last years a data driven approach began to take shape. This paper proposes three heuristic methods for both selection and allocation of the most informative stocks in an index tracking problem, respectively XGBoost, Random Forest and LASSO with stability selection. Among those, latest deep autoencoders have also been tested. All selected algorithms have outperformed the benchmarks in terms of tracking error. The empirical study has been conducted on one of the biggest financial indices in terms of number of components in three different countries, respectively Russell 1000 for the USA, FTSE 350 for the UK, and Nikkei 225 for Japan.

Do central counterparties reduce counterparty and liquidity risk? Empirical results

A central counterparty (CCP) interposes itself between buyers and sellers of financial contracts to extinguish their bilateral exposures. Therefore, central clearing and settlement through a CCP should affect how financial institutions engage in financial markets. Though, financial institutions’ interactions are difficult to observe and analyze. Based on a unique transactional dataset corresponding to the Colombian peso non-delivery forward market, this article compares—for the first time—networks of transactions agreed to be cleared and settled by the CCP with those to be cleared and settled bilaterally. Networks to be centrally cleared and settled show significantly higher connectivity and lower distances among financial institutions. This suggests that agreeing on central clearing and settlement reduces liquidity risk. After CCP interposition, exposure networks show significantly lower connectivity and higher distances, consistent with a reduction of counterparty risk. Consequently, evidence shows CCPs induce a change of behavior in financial institutions that emerges as two distinctive economic structures for the same market, which corresponds to CCP’s intended reduction of liquidity and counterparty risks.

Portfolio selection in non-stationary markets

We consider the task of portfolio selection as a time series prediction problem. At each time-step we obtain prices of a universe of assets and are required to allocate our wealth across them with the goal of maximizing it, based on the historic price returns. We assume these returns are realizations of a general non-stationary stochastic process, and only assume they do not change significantly over short time scales. We follow a statistical learning approach, in which we bound the generalization error of a non-stationary stochastic process, using analogues of uniform laws of large numbers for non-i.i.d. random variables. We use the learning bounds to formulate an optimization algorithm for portfolio selection, and present favorable numerical results with financial data.

Market Reaction to iPhone Rumors

The paper studies the effects of new product rumors about the iPhone on the stock price of the Apple company. We scrape iPhone rumors from Macrumors.com, and obtain a dataset covering 1,264 articles containing 180 words on average between January 2002 and December 2015. Moreover, we construct a market-decided lexicon to transform qualitative information into quantitative data, and analyze what type of words and what information embedded in the rumors are apt to impact on Apple’s stock price. Unlike previous studies, we do not rely on the widely-adopted Harvard-IV-4 dictionary, as the coefficients of the words from the dictionary are neither significant nor consistent with their polarities, compared with our results. The paper obtains three main findings. First, the spread of rumors has a significant impact on the stock price. Second, positive words, rather than negative words, play an important role in affecting the stock price. Third, the stock price is highly sensitive to the words related to the appearance of the iPhone.

A novel algorithm for clearing financial obligations between companies - An application within the Romanian Ministry of economy

The concept of clearing or netting, as defined in the glossaries of European Central Bank, has a great impact on the economy of a country influencing the exchanges and the interactions between companies. On short, netting refers to an alternative to the usual way in which the companies make the payments to each other: it is an agreement in which each party sets off amounts it owes against amounts owed to it. Based on the amounts two or more parties owe between them, the payment is substituted by a direct settlement. In this paper we introduce a set of graph algorithms which provide optimal netting solutions for the scale of a country economy. The set of algorithms computes results in an efficient time and is tested on invoice data provided by the Romanian Ministry of Economy. Our results show that classical graph algorithms are still capable of solving very important modern problems.

Portfolio optimization for cointelated pairs: SDEs vs Machine learning

With the recent rise of Machine Learning (ML) as a candidate to partially replace classic Financial Mathematics (FM) methodologies, we investigate the performances of both in solving the problem of dynamic portfolio optimization in continuous-time, finite-horizon setting for a portfolio of two assets that are intertwined. In the Financial Mathematics approach we model the asset prices not via the common approaches used in pairs trading such as a high correlation or cointegration, but with the cointelation model in Mahdavi-Damghani (2013) that aims to reconcile both short-term risk and long-term equilibrium. We maximize the overall P&L with Financial Mathematics approach that dynamically switches between a mean-variance optimal strategy and a power utility maximizing strategy. We use a stochastic control formulation of the problem of power utility maximization and solve numerically the resulting HJB equation with the Deep Galerkin method introduced in Sirignano and Spiliopoulos (2018). We turn to Machine Learning for the same P&L maximization problem and use clustering analysis to devise bands, combined with in-band optimization. Although this approach is model agnostic, results obtained with data simulated from the same cointelation model gives a slight competitive advantage to the ML over the FM methodology1.

Deep prediction of investor interest: A supervised clustering approach

We propose a novel deep learning architecture suitable for the prediction of investor interest for a given asset in a given time frame. This architecture performs both investor clustering and modelling at the same time. We first verify its superior performance on a synthetic scenario inspired by real data and then apply it to two real-world databases, a publicly available dataset about the position of investors in Spanish stock market and proprietary data from BNP Paribas Corporate and Institutional Banking.1,2

Sector categorization using gradient boosted trees trained on fundamental firm data

We examine to what extent the GICS sector categorization of equity securities may be systematically reconstructed from historical quarterly firm fundamental data using gradient boosted tree classification. Model complexity and performance tradeoffs are examined and relative feature importance is described. Potential extensions are outlined including ideas to improve feature engineering, validating internal consistency and integrating additional data sources to further improve classification accuracy.

An endogenous mechanism of business cycles

This paper suggests that business cycles may be a manifestation of coupled real economy and stock market dynamics and describes a mechanism that can generate economic fluctuations consistent with observed business cycles. To this end, we seek to incorporate into the macroeconomic framework a dynamic stock market model based on opinion interactions (Gusev et al., 2015). We derive this model from microfoundations, provide its empirical verification, demonstrate that it contains the efficient market as a particular regime and establish a link through which macroeconomic models can be attached for the study of real economy and stock market interaction. To examine key effects, we link it with a simple macroeconomic model (Blanchard, 1981). The coupled system generates nontrivial endogenous dynamics, which exhibit deterministic and stochastic features, producing quasiperiodic fluctuations (business cycles). We also inspect this system’s behavior in the phase space. The real economy and the stock market coevolve dynamically along the path governed by a stochastically-forced dynamical system with two stable equilibria, one where the economy expands and the other where it contracts, resulting in business cycles identified as the coherence resonance phenomenon. Thus, the incorporation of stock market dynamics into the macroeconomic framework, as presented here, allows the derivation of realistic behaviors in a tractable setting.