Given the wide range of sector products available to traders via the futures, equity and options markets, the ability to create a portfolio tilt is easier than ever. However, the abundance of products may also make an approach to constructing such a portfolio more unwieldy. This article makes use of the Standard & Poors Depositary Receipts (SPDRs®) and the Select Sector SPDRs® to present a modified tilt approach.
Portfolio tilts combine a passive investment vehicle that tracks a broad market benchmark and adds positions to overweight securities that are expected to outperform the benchmark while underweighting securities expected to underperform it. The approach presented here is identified as “modified” because it also incorporates a basic market timing technique. There are a countless number of combinations that can be explored by technicians and this article is geared towards providing food for thought for newer practitioners who seek an allocation approach for a portion of their funds to track and beat a specific market benchmark.
Construction Approach
The sample portfolio uses $10,000 as a base unit with no commissions or slippage included in the calculations. S&P 500 Index data is used to analyze returns prior to the introduction of SPY, which has a +1.0 correlation with the index since 2/1998. No total portfolio review or rebalancing was completed in this analysis since the data available for the “tilt” components was limited.
The base of the tilt is constructed using the S&P 500 SPDR® (SPY). In this low volatility market environment, deep in the money SPY Long Term Equity Anticipation Securities (LEAPs) with December expirations can offer some leverage to the trader, if desired. The nine Select Sector SPDRs® can be used to overweight and underweight the portfolio by sector. Industry and style alternatives are also available through other exchange traded fund (ETF) families. The primary ETF discussed here is the Select Sector SPDR®—Technology (XLK).
The portfolio tilt allocation is completed as followed in Table A.
Basic Market Timing
Rather than maintaining a $6,000 investment in the S&P 500 (SPX), the modified portfolio approach used a basic 50-200 daily exponential moving average (EMA) crossover system for entry and exit signals. When the 50-day EMA crosses above the 200-day EMA, 100% of the allocation is invested at the start of the following month. A cross of the 50-day EMA down below the 200-day EMA triggers removal of the allocation at the start of the following month. When not invested in the SPX, the appropriate T-Bill rate of return is applied to the funds. Using SPX and T-Bill returns from January 1976 through December 2005, the modified approach required an immediate investment in the SPX since the 50-day EMA was above the 200-ay EMA at the end of December 1975. (Table B)
Sector Allocation
In the spirit of keeping it simple, the sector approach makes use of the 10-year pattern in the stock market discussed by Larry Williams in his 2003 Wiley book, The Right Stock at the Right Time: Prospering in the Coming Good Years. He credits observations made by Edson Gould, along with work from Yale Hirsch in identifying preferred years for market investments. The approach used for this portion of the tilt primarily makes use of bullish periods in the 10-year pattern and a high beta sector.
According to the work performed, it is not uncommon for strong market rallies to begin in years ending in 2 or 3 and continue through years ending in 5. Similarly, rallies beginning in years ending in 7 generally realize their best gains prior to the end of a year ending in 0. However, since years ending in 7 can also see significant losses, an investment made in the month of November of that year is less likely to suffer initial, quick losses.
The allocation for the portfolio included two investment periods: 1) three years beginning January 1st for years ending in 3 (Jan x3 – Dec x5), and 2) twenty-six months beginning November 1st for years ending in 7 (Nov x7 – Dec x9). When not invested in the SPX, the funds receive the same rate of return as T-Bills. Once again, the period from January 1976 through December 2005 was used. (Table C)
Rather than using an SPX based product such as SPY, the trader could instead go long a higher beta sector security during bullish periods, or go long during bullish periods and short during the other market periods. Using XLK as the trading vehicle with a more limited period (January 1999 through December 2005), the following results were obtained: (Table D)
Although the long-short results for a high beta sector are strong, the review period is short and includes a particularly bearish period. Significant work is needed to assess drawdowns and risk. Since this is a 100% directional approach, such drawdowns could be substantial—an assessment into specific bearish returns in a 10-year cycle would be more appropriate. The use of long-term, in the money put options during bearish periods will cap risk; however, volatility levels will affect the cost of these securities when entering and exiting the position. Use of deeper in the money options helps alleviate the volatility impact.
The NASDAQ 100 Index can be used as a proxy for XLK (correlation for returns is +0.96) to obtain a longer review period. The short risk is better highlighted in the results that follow: (Table E)
Discretionary Allocation
Analysis of relative strength (RS) comparisons to SPY to identify outperforming and underperforming sectors can be used as part of a discretionary approach, along with other traditional technical analysis techniques. When using relative strength analysis, the trader needs to incorporate the impact of beta on the rate of change for the ratio line since conditions for more volatile sectors can change more quickly.
A second approach uses with Sam Stovall’s sector observations for the economic business cycle (Standard & Poor’s Guide to Sector Investing)[1]. (Table F)
The National Bureau of Economic Research (NBER) is the entity that declares the state of the business cycle, on a lagging basis. The most recent information from the group came in July 2003, announcing that a November 2001 trough ended the last economic contraction. NBER breaks out business cycle metrics for three periods (Table A) with the entire cycle lengthening in the third period. This lengthening is definitely attributable to longer recoveries, as the term for contractions declined successively in each period [from “Business Cycle Expansions and Contractions.” NBER Website. Tuesday, July 11, 2006]. (Table G)
Using the November 2001 low, we are currently 56 months into an expansion. This value is slightly lower than the average period for an expansion in the last 60 years, so it’s reasonable to look for signs of slowing. If we are truly in the later stages of an expansion, we’d expect to see relative health in Basic Materials and Energy, with some signs of life in the Consumer Staples and Utilities sectors. The charts for the corresponding Select Sector SPDR® seem to confirm the late expansion/early contraction period we are entering. Although the Basic Materials group (XLB) has shown recent weakness after a nice ten month rise, the Energy group (XLE) continues to remain strong. (Figures 1a & 1b)
The Utilities group (XLU) is in a price consolidation, but has shown strength relative to the S&P 500 since late spring. Similarly, Consumer Staples (XLP) turned in late spring after a multi-year decline in relative strength. (Figures 2a & 2b)
Summary
Although the techniques used for this portfolio construction were very basic, sometimes keeping it simple has its benefits. Since the tilt includes a significant investment in a security that is strongly, positively correlated to the underlying benchmark, the other allocations in the account require modest outperformance to meet the goal of beating the benchmark over time. Any discretionary approach that allows significant losses within the individual trades will make that goal elusive. Portfolio construction should include an analysis of sufficient historical data prior to implementation, along with a review of brokerage costs, for a better comparison to a passive investment.
Endnote
- John Murphy, Intermarket Analysis Profiting from Global Market Relationships, [John Wiley & Sons, 2004]