Options‐based benchmark indices—A review of performance and (in)appropriate measures

• DOI: http://doi.org/10.1002/fut.21865
• Author: Markus Natter
• Published date: 01 February 2018
• Date: 01 February 2018

Received: 8 April 2017

|

Accepted: 12 July 2017

DOI: 10.1002/fut.21865

RESEARCH ARTICLE

Options-based benchmark indices—A review of performance

and (in)appropriate measures

Markus Natter

University of Augsburg, Chair of Finance

and Banking, Augsburg, Germany

Correspondence

Markus Natter, University of Augsburg,

Chair of Finance and Banking,

Universitaetsstr. 16, 86159 Augsburg,

Germany. Homepage: http://www.wiwi.uni-

augsburg.de/bwl/wilkens/team/mitarbeiter/

natter_markus/

Email: natter.markus@t-online.de

This paper reviews the performance and profitability of different option strategy

benchmark indices provided by the CBOE. Using different performance approaches,

I show that performance measurement of these indices is highly complex and

sensitive to the model choice. Moreover, this study controls for time-varying delta

exposure via linear timing approaches and uses a linear option-factor model that is

independent from the portfolio composition. Splitting the sample, I find that

outperformance reported by previous studies is mostly driven by limited data.

Moreover, the profitability of option strategies for private investors is evaluated based

on easily investable investment products.

JEL CLASSIFICATION

G11, G20, G23

1

|

INTRODUCTION AND LITERATURE OVERVIEW

In 2002, the CBOE introduced the first option-based benchmark strategy index—the BuyWrite® Index (BXM). Since then, the

palette of these indices has experienced dramatic growth. More than 20 strategy indices with different underlying indices and

different properties are available on CBOE’s homepage.

1

Scientific researchers and practitioners both analyze the performance

of these hypothetical portfolios, especially for option-writing indices.

2

This study reviews the performance of strategy

benchmark indices using common and novel approaches to measuring profitability. To the best of my knowledge, I am the first to

measure the performance of a large number of benchmark indices instead of exclusively one strategy. Moreover, this paper uses a

long time horizon as well as different time windows for each strategy index, and introduces novel approaches, for example,

conditional factor models, to measure the performance of portfolios containing options.

So far, there are many studies attesting superior performance forCBOE’s strategy benchmark indices (e.g., Ungar & Moran,

2009; Whaley,2002). The first index published was the BXM BuyWriteIndex, which is a simple passive coveredcall strategy that

is long the S&P 500 and sells 1-month call options on the underlying index. The original purpose of the BXM was to provide a

sufficient benchmark for investorswhose portfolios contain options.

3

Whaley (2002) describesthe construction of this index and

finds morethan 20 basis points outperformance on a monthlyand risk-adjusted basis compared to the S&P 500.

4

At first glance,this

result seems surprising since the BXM strategy theoretically solely invests in the S&P 500 and the risk-free rate, because the

1

http://www.cboe.com/products/strategy-benchmark-indexes

2

A collection of contributions on option strategy benchmark indices can be found at http://www.cboe.com/products/strategy-benchmark-indexes/

bibliography

3

Natter et al. (2016) construct a risk-factor derived from BXM’s return.

J Futures Markets. 2018;38:271–288. wileyonlinelibrary.com/journal/fut © 2017 Wiley Periodicals, Inc.

|

271

replicatingportfolio of a short call option consistsof a short position in the respectiveunderlying and a long position ina zero bond.

Following a covered call strategy, therefore, means being long the underlying andsimultaneously selling some part of the same

underlying, whereasthe remaining amount of money is investedin a riskless bond. An investment in an index mixedwith the risk-

free rate should not generate any risk-adjusted outperformance. Nonetheless, there is a vast stream of literature showing the

profitability of option writingstrategies in the past. Among others, prominent representativesare the studies of Pounds (1978) as

well as Bookstaber and Clarke(1981, 1984, 1985). Several studies also find outperformancefor CBOE’s covered call strategies.

Feldman and Roy (2005),for example, report an annualized Jensen’s alphaof almost 3% p.a. for the BXM over a 16-year period.

Kapadia and Szado(2007) analyze a similar strategy with the Russell2000 as the underlying index and report an annualizedalpha

of 2%. Ungar and Moran(2009) measure the performance of the CBOE PutWrite indexwhose payoff is the analog to a covered call

strategy and find an outperformance of over 6% p.a. on a risk-adjusted basis.

This paper contributes to the literature on performanc e measurement of option-bas ed strategies as follows. ( i) So far,

most of the recently introduce d indices have not been analyze d in detail. (ii) This paper appli es and discusses more accura te

methods to measure the per formance of portfolios co ntaining options by allo wing time-varying delta e xposure in linear

models.

5

(iii) The volatility as the most impor tant determinant of option values is co nsidered in performance measur ement

via a simple option-factor approa ch. (iv) By splitting my sample into two s eparate parts, I show that outperfo rmance

documented by previous studi es is driven by the first-half (1990–2003)ofobservations.

6

In addition (v), I examine th e

profitability of option strat egies in different crisis scen arios. While Natter, Rohleder , Schulte, and Wilkens (2016) fi nd

superior performance amon g actively managed equity funds investing in options; I ex amine whether this result is transferable

to passively managed investmen ts and evaluate benefits of dire ct investments for private inves tors. Moreover, I show that

replicating theoreticall y calculated indices is non-tr ivial, as it is associated wit h costs (vi). Overall, I review an d validate

previous studies and critica lly assess their results.

The remainder of the paper is organized as follows. Section 2 introduces the data. Section 3 describes the performance

models and section 4 reports the results of the empirical analysis. In section 5, I highlight the possibility to gain exposure toward

option strategy benchmark indices. Section 6 concludes.

2

|

DATA AND INDEX DESCRIPTION

CBOE’s strategy benchmark indexes analyzed in this paper can be distinguished into six subgroups

7

:

2.1

|

BuyWrite indexes

A BuyWrite Index or covered call strategy is a passive hypothetical strategy that is long a specific underlying and writes call

options on that underlying. Hence, these call options consequently are considered to be covered. The indices of this subgroup

differ in their underlyings and as well as in the characteristics of the written call options.

2.2

|

PutWrite indexes

A Putwrite Index writes put options on different underlyings, whereas proceeds of the received option premia are invested into

riskless T-bills. Hence, the put options are cash-secured and, therefore, also covered. The payoff of this investment strategy is

similar to that of a covered call strategy.

4

The annualized outperformance is, therefore, 2.76% p.a.

5

Israelov and Nielsen (2015) and Israelov and Klein (2016) develop an approach to decompose the return of option strategies adequately and consider

time-varying delta exposure as well. However, their approach requires the knowledge of the exact portfolio composition and data on deltas of actually

traded options. The models in this paper shall be translatable to any portfolio return time-series.

6

Constantinides et al. (2009) find overpricing in S&P 500 options in this period.

7

A detailed description of the construction of benchmark indices can be found on CBOE’s homepage.

272

|

NATTER