Anatomy of a portfolio optimizer under a limited budget constraint

Abstract

Predicting the market’s behavior to profit from trading stocks is far from trivial. Such a task becomes even harder when investors do not have large amounts of money available, and thus cannot influence this complex system in any way. Machine learning paradigms have been already applied to financial forecasting, but usually with no restrictions on the size of the investor’s budget. In this paper, we analyze an evolutionary portfolio optimizer for the management of limited budgets, dissecting each part of the framework, discussing in detail the issues and the motivations that led to the final choices. Expected returns are modeled resorting to artificial neural networks trained on past market data, and the portfolio composition is chosen by approximating the solution to a multi-objective constrained problem. An investment simulator is eventually used to measure the portfolio performance. The proposed approach is tested on real-world data from New York’s, Milan’s and Paris’ stock exchanges, exploiting data from June 2011 to May 2014 to train the framework, and data from June 2014 to July 2015 to validate it. Experimental results demonstrate that the presented tool is able to obtain a more than satisfying profit for the considered time fram

Balancing the equity-efficiency trade-off in personal income taxation: an evolutionary approach

“…As an example, Figure 1 shows the net effect τ − χ for an employee without dependent individuals before and after the tax reform. 6 For what concerns c S 2 , the formula applied by the Italian tax code is a little bit different. In particular, it considers higher values than c Sp 2 − u in the income range 29,000-35,200: instead of 690, values range between 700 and 720.…”

“…Here we show that a set of other feasible tax structures, all reducing the tax revenue by the same amount, involving all the tax parameters and all taxpayers as well as better balancing equity and efficiency (and in the meanwhile keeping under control the re-ranking effect), are possible. Our methodology exploits an evolutionary algorithm, as such heuristics approaches have been demonstrated well suited to tackle complex real-world scenarios [4,5,6]. We then compare our results both with the effect of the reform recently implemented by the Italian government, and the one estimated in the article by Morini and Pellegrino [7], which is also based on an evolutionary algorithm framework.…”

“…where u is equal to 110, w to 40,000, k to 80,000, c Sp 2 is equal to 800, whilst c Op 2 to 750 6 . Tax credits for items of expenditure c 3 can be classified in several groups according to the percentage of the expense the tax law admits as a tax credit.…”

Input significance analysis: feature ranking through synaptic weights manipulation for ANNS-based classifiers

“…This weight initialization function has been specifically chosen and applied in the works of [42][43]. It has been claimed as the best function generating weights and biases that increase the speed of the training [44].…”