Not logged in.

Contribution Details

Type Master's Thesis
Scope Discipline-based scholarship
Title The Performance of the Lambda Star Strategy under New Stock Listings and Delistings - Backtested on Swiss Stock Market Data
Organization Unit
Authors
  • Qian Cao
Supervisors
  • Thorsten Hens
  • Sabine Elmiger
Language
  • English
Institution University of Zurich
Faculty Faculty of Economics, Business Administration and Information Technology
Number of Pages 45
Date April 2016
Abstract Text The purpose of this study is to backtest the performance of the theoretically successful Lambda Star strategy in evolutionary finance on real Swiss stock market data. The pool of assets has been hitherto assumed to be constant in the existing literature in evolutionary finance and does not consider the creation and disappearance of firms. However, since the evolutionary finance aims to seek for the winner strategy in a very long horizon, the birth and death of firms is an inevitable issue when we come to the real stock market. The main goal of our study is to investigate how the Lambda Star strategy behaves in the competition with other investment strategies when we allow for new stock listings and delistings of stocks in the sample consisting of stocks ever listed on SIX Swiss Exchange. Since the real-world dividends follow non-stationary processes, which are hard to be estimated in our case with a large sample of assets, we calculate the Lambda Star strategy with current relative dividends, implicitly assuming the dividends follow a Martingale process. Our backtesting results indicate that the relative dividend yield strategy (RDY) turns out to be the most successful strategy and can always outperform the Lambda Star strategy, whether the sample includes delistings or not and whatever delisting returns are assumed for the delisted stocks in the sample. We conclude that the current-dividend-based Lambda Star strategy is less effective than the RDY strategy in reflecting the fundamental value of firms’ dividends. The results also confirm the conclusions in Schweri (2011) that when dividends are not correctly estimated, the generalized Lambda Star strategy can be triumphed over by other investment strategies. Our results also show that the dividend-based strategies (i.e, the Lambda Star strategy and the RDY strategy) have better resilience towards delistings relative to other actively managed strategies and, therefore, suffer much less from delistings. The first reason for this result is that the dividend-based strategies automatically invest across a smaller pool of assets since most stocks in the baseline sample never pay dividends. This smaller pool of stocks includes fewer delistings and, therefore, the dividend-based strategies suffer less from delisting returns. Furthermore, firms usually stop paying dividends before they are actually delisted. The dividend-based strategies, which look at the current dividends, do not invest in such firms anymore when the firms’ current dividends are zero. In this way, the dividend-based strategies avoid the decreasing prices that often happen before the delisting day. The two reasons combine to make the dividend-based strategies much less affected by delistings relative to the non-dividend-based strategies, while the latter have no way to avoid either the delisting returns or the decreasing prices following stop of dividend payments. Accordingly, the non-dividend-based strategies fall behind the Lambda Star strategy when delisting returns are enormously large (e.g., -100%) or there is quite a large quantity of delistings in the sample.
Export BibTeX