Financial Portfolio Management is one of the most applicable problems in Reinforcement Learning (RL) by its sequential decision-making nature. Existing RL-based approaches, while inspiring, often lack scalability, reusability, or profundity of intake information to accommodate the ever-changing capital markets. In this paper, we design and develop MSPM, a novel Multi-agent Reinforcement learning-based system with a modularized and scalable architecture for portfolio management. MSPM involves two asynchronously updated units: Evolving Agent Module (EAM) and Strategic Agent Module (SAM). A self-sustained EAM produces signal-comprised information for a specific asset using heterogeneous data inputs, and each EAM possesses its reusability to have connections to multiple SAMs. A SAM is responsible for the assets reallocation of a portfolio using profound information from the EAMs connected. With the elaborate architecture and the multi-step condensation of the volatile market information, MSPM aims to provide a customizable, stable, and dedicated solution to portfolio management that existing approaches do not. We also tackle data-shortage issue of newly-listed stocks by transfer learning, and validate the necessity of EAM. Experiments on 8-year U.S. stock markets data prove the effectiveness of MSPM in profits accumulation by its outperformance over existing benchmarks.

Financial markets are highly non-linear and non-equilibrium systems. Earlier works have suggested that the behavior of market returns can be well described within the framework of non-extensive Tsallis statistics or superstatistics. For small time scales (delays), a good fit to the distributions of stock returns is obtained with q-Gaussian distributions, which can be derived either from Tsallis statistics or superstatistics. These distributions are symmetric. However, as the time lag increases, the distributions become increasingly non-symmetric. In this work, we address this problem by considering the data distribution as a linear combination of two independent normalized distributions - one for negative returns and one for positive returns. Each of these two independent distributions are half q-Gaussians with different non-extensivity parameter q and temperature parameter beta. Using this model, we investigate the behavior of stock market returns over time scales from 1 to 80 days. The data covers both the .com bubble and the 2008 crash periods. These investigations show that for all the time lags, the fits to the data distributions are better using asymmetric distributions than symmetric q-Gaussian distributions. The behaviors of the q parameter are quite different for positive and negative returns. For positive returns, q approaches a constant value of 1 after a certain lag, indicating the distributions have reached equilibrium. On the other hand, for negative returns, the q values do not reach a stationary value over the time scales studied. In the present model, the markets show a transition from normal to superdiffusive behavior (a possible phase transition) during the 2008 crash period. Such behavior is not observed with a symmetric q-Gaussian distribution model with q independent of time lag.

Machine learning models are increasingly used in a wide variety of financial settings. The difficulty of understanding the inner workings of these systems, combined with their wide applicability, has the potential to lead to significant new risks for users; these risks need to be understood and quantified. In this sub-chapter, we will focus on a well studied application of machine learning techniques, to pricing and hedging of financial options. Our aim will be to highlight the various sources of risk that the introduction of machine learning emphasises or de-emphasises, and the possible risk mitigation and management strategies that are available.

We introduce a prototype agent-based model of the macroeconomy, with budgetary constraints at its core. The model is related to a class of constraint satisfaction problems (CSPs), which has been thoroughly investigated in computer science. The CSP paradigm allows us to propose an alternative price-setting mechanism: given agents' preferences and budgets, what set of prices satisfies the maximum number of agents? Such an approach permits the coupling of production and output within the economy to the allowed level of debt in a simplified framework. Within our model, we identify three different regimes upon varying the amount of debt that each agent can accumulate before defaulting. In presence of a very loose constraint on debt, endogenous crises leading to waves of synchronized bankruptcies are present. In the opposite regime of very tight debt constraining, the bankruptcy rate is extremely high and the economy remains structure-less. In an intermediate regime, the economy is stable with very low bankruptcy rate and no aggregate-level crises. This third regime displays a rich phenomenology:the system spontaneously and dynamically self-organizes in a set of cheap and expensive goods (i.e. some kind of "speciation"), with switches triggered by random fluctuations and feedback loops. Our analysis confirms the central role that debt levels play in the stability of the economy. More generally, our model shows that constraints at the individual scale can generate highly complex patterns at the aggregate level.

Using the generalized extreme value theory to characterize tail distributions, we address liquidation, leverage, and optimal margins for bitcoin long and short futures positions. The empirical analysis of perpetual bitcoin futures on BitMEX shows that (1) daily forced liquidations to out- standing futures are substantial at 3.51%, and 1.89% for long and short; (2) investors got forced liquidation do trade aggressively with average leverage of 60X; and (3) exchanges should elevate current 1% margin requirement to 33% (3X leverage) for long and 20% (5X leverage) for short to reduce the daily margin call probability to 1%. Our results further suggest normality assumption on return significantly underestimates optimal margins. Policy implications are also discussed.

Analysis of policies for managing epidemics require simultaneously an economic and epidemiological perspective. We adopt a cost-of-policy framework to model both the virus spread and the cost of handling the pandemic. Because it is harder and more costly to fight the pandemic when the circulation is higher, we find that the optimal policy is to go to zero or near-zero case numbers. Without imported cases, if a region is willing to implement measures to prevent spread at one level in number of cases, it must also be willing to prevent the spread with at a lower level, since it will be cheaper to do so and has only positive other effects. With imported cases, if a region is not coordinating with other regions, we show the cheapest policy is continually low but nonzero cases due to decreasing cost of halting imported cases. When it is coordinating, zero is cost-optimal. Our analysis indicates that within Europe cooperation targeting a reduction of both within country transmission, and between country importation risk, should help achieve lower transmission and reduced costs.

We examine probabilistic forecasts for battleground states in the 2020 US presidential election, using daily data from two sources over seven months: a model published by The Economist, and prices from the PredictIt exchange. We find systematic differences in accuracy over time, with markets performing better several months before the election, and the model performing better as the election approached. A simple average of the two forecasts performs better than either one of them overall, even though no average can outperform both component forecasts for any given state-date pair. This effect arises because the model and the market make different kinds of errors in different states: the model was confidently wrong in some cases, while the market was excessively uncertain in others. We conclude that there is value in using hybrid forecasting methods, and propose a market design that incorporates model forecasts via a trading bot to generate synthetic predictions. We also propose and conduct a profitability test that can be used as a novel criterion for the evaluation of forecasting performance.

In this paper we develop models of asset return mean and covariance that depend on some observable market conditions, and use these to construct a trading policy that depends on these conditions, and the current portfolio holdings. After discretizing the market conditions, we fit Laplacian regularized stratified models for the return mean and covariance. These models have a different mean and covariance for each market condition, but are regularized so that nearby market conditions have similar models. This technique allows us to fit models for market conditions that have not occurred in the training data, by borrowing strength from nearby market conditions for which we do have data. These models are combined with a Markowitz-inspired optimization method to yield a trading policy that is based on market conditions. We illustrate our method on a small universe of 18 ETFs, using three well known and publicly available market variables to construct 1000 market conditions, and show that it performs well out of sample. The method, however, is general, and scales to much larger problems, that presumably would use proprietary data sources and forecasts along with publicly available data.

Risk capital allocations (RCAs) are an important tool in quantitative risk management, where they are utilized to, e.g., gauge the profitability of distinct business units, determine the price of a new product, and conduct the marginal economic capital analysis. Nevertheless, the notion of RCA has been living in the shadow of another, closely related notion, of risk measure (RM) in the sense that the latter notion often shapes the fashion in which the former notion is implemented. In fact, as the majority of the RCAs known nowadays are induced by RMs, the popularity of the two are apparently very much correlated. As a result, it is the RCA that is induced by the Conditional Tail Expectation (CTE) RM that has arguably prevailed in scholarly literature and applications. Admittedly, the CTE RM is a sound mathematical object and an important regulatory RM, but its appropriateness is controversial in, e.g., profitability analysis and pricing. In this paper, we address the question as to whether or not the RCA induced by the CTE RM may concur with alternatives that arise from the context of profit maximization. More specifically, we provide exhaustive description of all those probabilistic model settings, in which the mathematical and regulatory CTE RM may also reflect the risk perception of a profit-maximizing insurer.

In Japan, the increase in the consumption tax rate, a measure of balanced public finance, reduces the inequality of fiscal burden between the present and future generations. This study estimates the effect of grandchildren on an older person's view of consumption tax, using independently collected data. The results show that having grandchildren is positively associated with supporting an increase in consumption tax. Further, this association is observed strongly between granddaughters and grandparents. However, the association between grandsons and grandparents depends on the sub-sample. This implies that people of the old generation are likely to accept the tax burden to reduce the burden on their grandchildren, especially granddaughters. In other words, grandparents show intergenerational altruism.

Foreign Exchange (Forex) is the largest financial market in the world. The daily trading volume of the Forex market is much higher than that of stock and futures markets. Therefore, it is of great significance for investors to establish a foreign exchange forecast model. In this paper, we propose a Wavelet Denoised-ResNet with LightGBM model to predict the rate of change of Forex price after five time intervals to allow enough time to execute trades. All the prices are denoised by wavelet transform, and a matrix of 30 time intervals is formed by calculating technical indicators. Image features are obtained by feeding the maxtrix into a ResNet. Finally, the technical indicators and image features are fed to LightGBM. Our experiments on 5-minutes USDJPY demonstrate that the model outperforms baseline modles with MAE: 0.240977x10EXP-3 MSE: 0.156x10EXP-6 and RMSE: 0.395185x10EXP-3. An accurate price prediction after 25 minutes in future provides a window of opportunity for hedge funds algorithm trading. The code is available from https://mkhushi.github.io/