International Journal of Knowledge Based Computer System

Apoorva Uday Nayak ¹, Megha Ugalmugale ¹, Vishwanjali Gaikwad ¹, Vahida Z. Attar ²

Affiliations
1 College of Engineering, Pune, Maharashtra, IN
2 Department of Computer Engineering, IT at the College of Engineerig, Pune, Maharashtra, IN

Abstract

The stock market is one of the most important sources for companies to raise money. The prediction of stock prices has always been a challenging task. Investors are always in search of profitable trading strategies which provide accurate trading signals. Pair trading is an investment strategy of matching a long position (buy) with a short position (sell) in two historically correlated stocks. In pair trading, the use of correlation or co-integration as a dependence measure is central. Some traditional techniques also assume symmetric distribution of data along the mean zero. However, in case of financial assets the occurrence of symmetric distributions is very rare and, hence, the use of these techniques may lead to erroneous results. Copula is a relatively new pair trading technique which describes dependence structures for linear and non-linear distributions without making rigid assumptions. Copula, thus, overcomes the shortcomings of traditional pair trading techniques and gives accurate trading signals. This paper is a study of different types of copulas and their application in finance to fit different types of data.

Keywords

Copula, Pair Trading, Stock, Linear Dependence, Normal Distribution, Kendall Tau.

Full Text