Emerald Management Review:
Portfolio selection under DEA-based relative financial strength indicators: case of US industries

Abstract:

Purpose - To use data envelopment analysis (DEA) to generate a relative financial strength indicator (RFSI) for firm stock performance analysis.

Design/methodology/approach - Touches on stock price prediction studies using historical data, and how DEA gauges firms' relative performance using various input/output indicators. Formulates a modified CCR-DEA model where firms have an efficiency of zero if an input parameter is non-positive. Identifies 18 public-domain quarterly financial statement parameters. Selects nine of these as inputs and seven as outputs for a basic selection model (BS-DEA). Incorporates two excluded valuation parameters in an augmented model (AS-DEA). Generates a parsimonious BS-DEA version (RB-DEA), and a direct selection model (DS-DEA) based on raw financial data. Assesses each and the RIV valuation model (Ali. et al., 2002) using publicly traded US firms' 1996-2002 data from http://biz.yahoo.com/p/8conameu.html and Wharton Research Data Service (six industry sectors). Develops a refinement to rank 100% efficient firms.

Findings - Presents comparative bar charts, goodness-of-fit and other results in a detailed performance analysis. Concludes, inter alia, that BS and RB models give better correlations over one- and two-month time lags than AS, so their efficiency scores qualify as RFSI proxies. Uses these for strong-performing firm selection against the RIV. Finds BS-DEA RFSPIs outperform the others (Sharpe ratios) for one- and, two-month time lags, but RIV is best at a three-month lag.

Research limitations/implications - None stated.

Originality/value - Provides a DEA-based financial indicator proxy approach for portfolio selection that is a 'serious contender' to RIV-based fundamental analysis.

Keywords:

Article Type:

Research paper

Reference:

Top