SUMMARY

• Using price-to-book (PB) or price-to-earnings (PE) results in similar Value factor performance
• Some sectors are perpetually expensive while others are always cheap
• Sector rotation is higher with PE than with PB

INTRODUCTION

In 2016 the Value factor in the US and globally showed strong performance after years of random factor performance and dedicated value investors hoped for a continuation of this trend. In this short research note we’re going to analyse what sectors were in the long and short portfolios in US Value portfolios from 2000 to the beginning of 2017, effectively having a look of what’s under the hood.

METHODOLOGY

Value can be defined in many ways with the most common differences arising from the selection of the valuation methodology and the definition of the investable universe.

Over the last few years there has been some empirical evidence that cashflow based valuation metrics produce superior performance compared to more traditional price-to-book (PB) or price-to-earnings (PE) ratios. Intuitively this make sense as PB in the US uses antiquated book value numbers, which are not particular relevant for buying or selling a company. Having said this, the investible universe can shrink significantly when using more cashflow based metrics, e.g. using EBITDA makes the Financials sector ineligible. Therefore we will use PB and PE for constructing long and short portfolios.

The investible universe is usually defined by sectors or taken across sectors. Higher Sharpe ratios are typically achieved via intra-sector factor exposure, however, often at much lower absolute returns. Given that we want to analyse the sector rotation within the Value factor, the portfolios are created across sectors.

VALUE US USING PE

The chart below shows the performance of a long-short portfolio of US stocks selected on PE, i.e. buying cheap and selling expensive companies. The portfolios represent the bottom and top 10% of the stock universe of US companies with a market capitalization larger than $1 billion. Stocks are rotated monthly and 10bps of transaction costs are included.

The analysis starts in 2000 and the performance was exceptionally strong until 2006. The chart is somewhat misleading as the performance before 2000 was strongly negative, i.e. we’re only seeing the recovery, not the drawdown before that. Value crashe