nach oben

2012 | Buch

Kapitel lesen Erstes Kapitel lesen

Practical Applications of Evolutionary Computation to Financial Engineering

Robust Techniques for Forecasting, Trading and Hedging

verfasst von: Hitoshi Iba, Claus C. Aranha

Verlag: Springer Berlin Heidelberg

Buchreihe : Evolutionary Learning and Optimization

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Einloggen, um Zugang zu erhalten

Über dieses Buch

“Practical Applications of Evolutionary Computation to Financial Engineering” presents the state of the art techniques in Financial Engineering using recent results in Machine Learning and Evolutionary Computation. This book bridges the gap between academics in computer science and traders and explains the basic ideas of the proposed systems and the financial problems in ways that can be understood by readers without previous knowledge on either of the fields. To cement the ideas discussed in the book, software packages are offered that implement the systems described within.

The book is structured so that each chapter can be read independently from the others. Chapters 1 and 2 describe evolutionary computation. The third chapter is an introduction to financial engineering problems for readers who are unfamiliar with this area. The following chapters each deal, in turn, with a different problem in the financial engineering field describing each problem in detail and focusing on solutions based on evolutionary computation. Finally, the two appendixes describe software packages that implement the solutions discussed in this book, including installation manuals and parameter explanations.

Inhaltsverzeichnis

Frontmatter

Introduction to Genetic Algorithms

Abstract

We can take from the above expert that a large part of technological and social innovations come from improvements on already existing ideas. It could not be in any other way: the human being has an irresistible urge to explore the world around it and modify it, and this includes both the natural world, and that created by his ancestors.

However, just as recent technologies have been developed upon previous discoveries by great minds of the past, a new trend has come into being. What if we tried to improve our technology by watching how nature solves its own problems? Bio-inspired Computing is a new discipline where the inspiration for new algorithms comes from the observation of animals and natural systems.

Hitoshi Iba, Claus C. Aranha

Advanced Topics in Evolutionary Computation

Abstract

In the first chapter we saw that Evolutionary Computation use the concepts of natural evolution to efficiently search the solution of an optimization problem.

In the previous chapter we described the basic concepts of Evolutionary Computation. We discussed the natural underpinnings of evolution that were used as motivation and guidelines for EC. Then we described its main paradigms: Genetic Algorithms,Genetic Programming, Evolution Strategies and Evolutionary Programming. Finally, we introduced the essential components of an Evolutionary Computing method, showing how to put them together in a simple evolutionary optimization system.

However, Evolutionary Computing as a field already exists for more than 30 years. In this period, many issues in the basic methods were identified, and solutions for those issues were developed. These solutions have developed into full subsystems of the evolutionary method themselves, each worthy of a whole volume for detailed explanations.

Hitoshi Iba, Claus C. Aranha

Financial Engineering

Abstract

Financial Engineering is the term used to describe the use of engineering and mathematical methods and tools to solve financial problems. This includes, but it is not limited to, the mathematical analysis of the market, the modeling of its behavior, and eventually the use of optimization methods on this model.

The goal of the financial engineer is to generate a system that uses such models and optimization methods to aid a trader in his decisions for trading in the market. More recently, the generation of completely automated systems that are able to make complex investment decisions, and act on these decisions independently, has also become popular.

Hitoshi Iba, Claus C. Aranha

Predicting Financial Data

Abstract

Let us try to predict a time series. The goal here is to establish a function based on observed values (time series data). Using

$$x_1,x_2,x_3,...,x_t,~~~~~~~(4.1)$$

we attempt to obtain a function

$$x_t = f (x_{t-1},x_{t-2},x_{t-3},x_{t-4},...,x_{t-M})~~~~~~~ (4.2)$$

that can be used to predict current data x _t from previously observed data. The reader should note that the arguments of this function do not include a time variable t. This is to avoid deriving a time series prediction function that is dependent on absolute time.

Hitoshi Iba, Claus C. Aranha

Trend Analysis

Abstract

In the previous chapter, we saw how to predict the future prices of an asset, based on its past prices. This problem was called “time series prediction”, and there are many different techniques, both traditional and evolutionary, to perform this task. All these techniques use the information provided by the past prices of the stock, called the historical data, to forecast the future price.

However, sometimes to make a financial decision, we don’t need to know the exact price of the asset in the future. For some policies, knowing only that the asset’s price rise or fall in the short term may be enough to decide whether to hold a position or to close it.

Hitoshi Iba, Claus C. Aranha

Trading Rule Generation for Foreign Exchange (FX)

Abstract

In the previous chapters we described how we can use Evolutionary Computation to perform forecasting in financial data and trend analysis. In both cases, computational intelligence, in the form of EC, processes large amounts of financial data, and transforms it into information that can be used by a human trader.

But what if we want to design a computational agent that is able to perform trades from end to end? The artificial trader would be able to receive raw technical data, such as the price of stocks or exchange rates, and analyze it. Based on the information from this analysis, it can autonomously make a trading decision, such as buying or selling.

Hitoshi Iba, Claus C. Aranha

Portfolio Optimization

Abstract

One of the first concepts that a person learns when dealing with the market is that of “Buy Low, Sell High”. In other words, if the trader buys a stock which is expected to go up in price, he will make a profit if he sells that stock later. For example, let’s say that a car company A is planning to build a new factory this year. In a simple interpretation of the “buy low, sell high” concept, it would make sense to buy some stocks of A.We expect that, when the new factory opens, A’s productivity will increase and its stock’s price will go up accordingly.

However, imagine that during the construction of the factory a natural accident happens and destroys it, making A lose much of the time and money invested in that project. A’s stock, which was predicted to go up actually decreases, resulting in a loss to the traders who bought the stock when the new factory was announced (See Figure 7.1).

Hitoshi Iba, Claus C. Aranha

Backmatter

Titel: Practical Applications of Evolutionary Computation to Financial Engineering
verfasst von: Hitoshi Iba
Claus C. Aranha
Verlag: Springer Berlin Heidelberg
Electronic ISBN: 978-3-642-27648-4
Print ISBN: 978-3-642-27647-7
DOI: https://doi.org/10.1007/978-3-642-27648-4

Springer Professional

Über dieses Buch

Inhaltsverzeichnis

Frontmatter

Introduction to Genetic Algorithms

Advanced Topics in Evolutionary Computation

Financial Engineering

Predicting Financial Data

Trend Analysis

Trading Rule Generation for Foreign Exchange (FX)

Portfolio Optimization

Backmatter

Premium Partner