Tail Risk Management Collaring Multiple-Asset Exchange-Traded Funds

Feature

By Philip H. Gocke

ear and greed have long been FIGURE 1: GROWTH OF $100 SPY ONE-MONTH CALL/SIX-MONTH PUT used to characterize market BALANCED COLLARS sentiment. What has changed in F June 1, 2007–December 30, 2011 today’s investment climate is the speed at which the market vacillates between $130 +22% $120 +14% the two extremes. The shortened cycles $110 led to a new sentiment description— $100 risk on, risk off, or “ro ro.” One $90 –9% $80 consequence of this bipolar attitude $70 toward risk has been increased asset $60 class correlations, negating the effective $50 $40 benefit of many traditional equity diversifiers. Meanwhile investors have Jan 08 Jan 09 Jan 10 Jan 11 Jun 07 Jun 08 Jun 09 Jun 10 Jun 11 Sep 07 Sep 08 Sep 09 Sep 10 Sep 11 Nov 07 Nov 08 Nov 09 Nov 10 Nov 11 been forced to investigate strategies that May 07 May 08 May 09 May 10 May 11

offer downside protection in addition to SPY SPY 2% OT Collar SPY 5% OT Collar upside participation. One solution to these competing desires is an option-based equity collar. TABLE 1: SUMMARY STATISTICS SPY ONE-MONTH CALL/SIX-MONTH PUT BALANCED COLLARS A collar can provide portfolios with SPY Collar Summary ATM 2% OTM 5% OTM greater downside risk protection than Statistics 1–Month 1–Month Call, 1–Month Call, standard multi-asset diversification pro- June 1, 2007 to SPY Total Call, ATM 2% OTM 5% OTM grams while also allowing for profits dur- December 30, 2011 Return 1–Month Put 6–Month Put 6–Month Put ing risk on rallies. The Options Industry Annualized Return –2.14% 3.48% 4.47% 3.03% Annualized Standard Council (OIC), as part of its mission 19.46% 6.68% 8.37% 10.42% to provide education and research to Deviation institutional investors, helped sponsor Mean Monthly Return –0.02% 0.30% 0.39% 0.29% Median Monthly research on the performance of the col- 0.01% 0.41% 0.57% –0.03% Return lar strategy against a range of exchange- Period Cumulative –9.45% 16.95% 22.17% 14.64% traded funds (ETFs) across multiple asset Return classes, including equity, currency, com- Sharpe Ratio –0.16 0.36 0.41 0.19 modity, fixed income, and real estate. In Maximum Drawdown –50.80% –8.99% –11.13% –19.81% “Option-Based Risk Management in a Maximum Run Up 93.38% 20.76% 35.85% 48.81% Multi-Asset World,” Edward Szado and Thomas Schneeweis (2012) show that % Down Months 47% 40% 36% 51% for most of the asset classes considered, % Up Months 53% 60% 64% 49% an option-based collar strategy, using Number of Months 55 55 55 55 six-month put purchases and consecutive one-month call writes, provides the Over the 55-month study period ending superior returns with less than half the holy grail of investing: improved risk- December 30, 2011, the 2-percent out- risk as measured by the standard devia- adjusted performance and significant of-money (OTM) SPY collar returned tion (8.4 percent for the collar versus risk reduction. more than 22 percent (4.5 percent annu- 19.5 percent for SPY). One of the most Figure 1 and table 1 illustrate the ally), while the long SPY experienced impressive statistics supporting the benefit of an equity collar strategy on a loss of more than 9 percent (‒2.1 potential benefit of equity collar protec- the popular SPDR® S&P 500® (SPY) ETF. percent annually). The collar earns its tion is the maximum drawdown. During

50 Investments&Wealth MONITOR

TABLE 2: SUMMARY STATISTICS 5% OTM, ONE-MONTH CALL/SIX MONTH PUT COLLARS (STUDY PERIOD 55 MONTHS EXCEPT GLD) 5% OTM 5% OTM ETF Collar 5% OTM ETF Collar Annualized Annualized ETF Collar ETF Cumulative Cumulative Std Std Maximum Maximum Symbol Return Return Deviation Deviation Drawdown Drawdown S&P 500 – SPDR SPY –9.45% 14.64% 19.46% 10.42% –50.80% –19.81% Russell 2000 – iShares IWM –6.72% –0.42% 25.13% 14.17% –52.42% –23.87% NASDAQ 100 – PowerShares QQQ 21.01% 8.28% 22.79% 13.16% –49.74% –28.80% EAFE Index – iShares MSCI EFA –29.43% –15.73% 24.52% 12.24% –57.38% –31.84% Emerging Market Index – iShares EEM –1.97% 29.88% 30.95% 14.05% –60.44% –17.61% Australian Dlr Tr. – CurrencyShares FXA 48.16% 26.36% 18.24% 10.92% –31.75% –18.82% British Pound Tr. – CurrencyShares FXB –16.99% –12.13% 10.78% 7.00% –28.09% –18.41% Canadian Dlr Tr. – CurrencyShares FXC 9.26% 13.67% 12.66% 8.19% –23.74% –9.13% Euro Tr. – CurrencyShares FXE 1.43% 3.40% 13.62% 9.27% –21.19% –13.20% Swiss Franc Tr. – CurrencyShares FXF 31.10% 34.20% 14.70% 9.51% –17.65% –10.37% Japanese Yen Tr. – CurrencyShares FXY 55.69% 47.00% 10.37% 8.13% –9.36% –7.79% Hi Yield Corp. Bond – iBoxx HYG 26.63% 2.50% 17.01% 7.26% –30.28% –15.30% Barclays 20+ Treasury Bond – iShares TLT 70.12% 29.76% 16.85% 10.97% –21.80% –17.02% DJ US Real Estate – iShares IYR –17.14% –6.51% 33.27% 13.33% –67.89% –32.58% GSCI Commodity Tr. – S&P GSG –19.95% 12.59% 28.83% 11.37% –67.85% –19.34% Gold Tr. – SPDR * GLD 66.29% 34.35% 22.82% 13.65% –21.95% –11.30% U.S. Oil Fund – U.S. Commodity Funds USO –22.16% 26.11% 37.79% 17.03% –76.20% –33.76%

* The inception date for the GLD ETF was November 18, 2004, with option data available from June 20, 2008. Study period = 42 months the study period, SPY experienced a the six-month put expires, it is settled transaction costs by purchasing at the maximum loss of 50.8 percent while at intrinsic value and again rolled into offer price and selling at the bid price. the 2-percent OTM collar reduced this new puts and calls with the specified Conclusions negative performance to a maximum moneyness and time to expiration. loss of 11.1 percent. The authors wished to avoid In contrast to earlier studies that con- Szado and Schneeweis (2012) circumstances where the underlying centrated on equity markets, Szado and evaluated the impact of collar strategies prices decline significantly and new Schneeweis (2012) provide extensive across a wide range of asset classes calls would have been written at lower analysis of the performance of collar based on a set of trading rules (table 2). (crossed) strike prices than the existing strategies over a diverse set of asset For each of the ETFs, the study analyzes deep in-the-money (ITM) put. In these classes including equities, currencies, an array of strike prices with defined cases the long put and long underlying commodities, fixed income, and real initial moneyness where a six-month positions would counteract each other estate. Their study covers the period put is purchased and consecutive and the new call would essentially be from June 1, 2007, to December 30, 2011, one-month calls are written. At the written naked. Since writing naked calls except for the gold ETF (GLD), which close on the day before the Saturday is inconsistent with the risk-reduction begins on July 1, 2008, the first full month expiration of the calls and depending on goal of the collar strategy, Szado and after the June 20, 2008, inception of GLD the particular passive implementation, Schneeweis (2012) implemented a rule option trading. The period of study was the initial moneyness of the calls and to roll the puts to the strategy’s target chosen to capture both the financial crisis puts was set at: 25-percent, 10-percent, moneyness and maturity based on the as well as the following market recov- 5-percent, or 2-percent OTM, or at- current underlying price on the day the ery. The risk reduction and drawdown the-money (ATM). At expiration, the new short call position was initiated. protection ability is evident in the results calls were settled at intrinsic value and The put sale was rolled at the mid-point across all asset classes analyzed. new one-month calls with the specified between the bid and ask and the new Table 2 shows that the results are moneyness were sold. The longer-term put is purchased at the ask price. All somewhat mixed from a total-return put is held for another month. Once other executions in the study include Continued on page 56

November/December 2012 51

Kitces Continued from page 55

to import data from outside providers live in a client meeting instead of via a automation will impact or change your (although not necessarily comprehen- multi-day back-and-forth data-gather- value proposition with clients?

sively or fully automated—at least not ing process followed by data input to yet). Some proactive financial-plan the planning software and an “updated Michael Kitces, MSFS, MTAX, CFP®, monitoring tools have been built into plan presentation” meeting. CLU, ChFC, is a partner and the the new (and free) software platform So what do you think? Is your firm director of research for Pinnacle inStream. an early adopter of the new technologies Advisory Group, a private wealth But ultimately, expect far more inte- that will dominate plan monitoring management firm in Columbia, gration than even what these tools have and updating in the future? Do you or Maryland. He is publisher of the achieved so far, which in turn will make your clients use account aggregation e-newsletter The Kitces Report and many aspects of the financial-planning software? Imports to your financial the blog Nerd’s Eye View. Contact process easier. Imagine if financial-plan planning software? Have you tried him at [email protected]. updates could be done in minutes, inStream? Are you concerned that this

Surz Continued from page 49

that there are no generally accepted standards to guide these Ron Surz, CIMA®, is president of PPCA Inc. and its decisions. Without standards we cannot differentiate between division, Target Date Solutions. He earned an MBA in good and bad. Accordingly, plan sponsors need to adopt finance from the University of Chicago and an MS in TDF standards and, in our opinion, these standards should applied mathematics from the University of Illinois. emphasize safety, especially during the critical transition Contact him at [email protected]. period. Plan sponsors need to drive this rocket ship during the accumulation phase.

Gocke Continued from page 51

perspective. Two currencies (Australian dollar and Japanese Philip H. Gocke is managing director of research and yen), the two bond ETFs (HYG & TLT), and the QQQ and GLD educational strategy for institutional investors at the had higher cumulative returns than any of the collar iterations. Options Industry Council (OIC), where he is respon- The authors suggest that with respect to total returns, option- sible for educating pension funds, hedge funds, and based collar strategies tend to outperform when market declines money management firms about the benefits and risks are aggressive. However, the results indicated that the strategy of exchange-traded options. He earned a BA in econom- tends to underperform in periods of extreme market rallies ics from Lafayette College and an MBA from New York when the written call caps the gains of the underlying long University. Contact him at [email protected]. position. While option-based collars may not provide complete Reference protection for all products and in all market conditions, collars can provide significant risk control across a wide range of asset Szado, Edward, and Thomas Schneeweis. Option-Based Risk classes, significantly reducing volatility, drawdowns, and, in Management in a Multi-Asset World. Working paper, University of certain market environments, they also can provide enhanced Massachusetts Amherst. http://www.optionseducation.org/documents/ returns relative to a stand-alone investment. literature/files/options-based-risk-mgmt.pdf.

56 Investments&Wealth MONITOR