Indexology^® Blog

The S&P/BMV Mexico Target Risk Index Series comprises four multi-asset class indices that are designed to serve as benchmarks for the Mexican pension system based on the risk tolerance levels of plan participants. Generally, younger individuals with longer time horizons until retirement will have greater risk tolerance and therefore higher exposure to risky assets such as equities, while older individuals will allocate to more conservative assets such as short-term nominal and inflation-linked bonds. As outlined in our paper, “Benchmarking Lifecycle Investment Strategies: Introducing the S&P/BMV Mexico Target Risk Indices,” we used liquid, investable indices to represent each asset class.

Designed to represent investment strategies with varying levels of risk appetite, the indices allocate across various asset classes to meet their respective risk target. Therefore, it is important to understand the sources of portfolio risk. In this blog, we decompose the risk of each portfolio by asset class so that we can identify and attribute the portion of realized risk coming from each index.

In order to assess the contribution of each asset class to portfolio risk,[1] we first calculate the weighted covariance matrix of all the indices representing each asset class in the index series. As seen in the equations in the footnote, the contribution of a given asset class to portfolio volatility is computed by the correlation between that asset class and the index portfolio multiplied by the volatility of the asset class in question and its weight in the index portfolio. Correlation and volatility are estimated using daily returns over one-year periods from 2009 through 2017.

We then get the percentage of contribution to total portfolio risk by dividing the asset class (or the index) contribution by the total portfolio volatility. The sum of all the contributions must be equal to the total portfolio risk and the sum of the asset class percentages of contributions to portfolio risk must be equal to 1 (or 100%).

Exhibit 1 charts the realized portfolio risk of each S&P/BMV Mexico Target Risk Index, the risk contribution to the index for each asset class, and the percentage of contribution to the total risk for each asset class on an annual basis. For example, in 2017, the total risk of the S&P/BMV Aggressive Target Risk Index was 3.65%, of which 1.36% was from the equity portion, contributing to 38% of the overall risk. The contribution to portfolio risk by the equity portion rose after declining to 34% in 2016.

As we noted in our paper, there are periods, such as 2013, when the aggressive portfolio underperformed its growth, moderate, and conservative counterparts due to asset allocation mix. A close examination of asset class contribution to risk provides some insight. For example, in 2013, the biggest contributor to the total risk of the S&P/BMV Mexico Target Risk Aggressive Index came from fixed income, namely the S&P/BMV Sovereign MBONOS 20+ Year Bond Index. At roughly 28% of portfolio total risk, its risk contribution was higher than that of the equity portion (see Exhibit 2).

In this blog, we highlighted the importance of understanding the sources of portfolio risk. Asset allocation strategies, such as the S&P/BMV Mexico Target Risk Indices, require decomposition of portfolio risk so that market participants are aware of where the risk is coming from and can assess the impact of each asset class on the overall portfolio.

[1] To calculate each asset class contribution to risk, we start by defining the portfolio volatility as 

From this equation, the risk contribution of each asset class can be derived as

Where we stand in the economic cycle can have a measurable effect on sector performance

There are many determinants of stock performance. Corporate earnings, fiscal policy and interest rates can all influence the equity markets. But equity returns are also dependent on where we stand in the economic cycle.

Some sectors, such as industrials and financials, tend to display strong performance early in the economic cycle when economic growth is accelerating. Other sectors, like utilities and consumer staples, tend to be strongest very late in the economic cycle when economic growth is weakest.

How do we know this?

The correlation between excess returns and economic cycles

Consider the table below, which displays the excess price return of the 10 sectors that make up the S&P 500 Index over a roughly 17-year period, from October 2001 through April 2018. (Here, I am defining excess returns as the price of each sector minus the price of the S&P 500 Index.) I’ve segregated returns by economic cycle, defined by the position of the Institute of Supply Management Manufacturing Index (ISM Manufacturing Index), which is a broad barometer of manufacturing activity in the US.

As you can see from the chart below, the ISM Manufacturing Index has historically tracked year-over-year changes in earnings per share (EPS) growth for the S&P 500 Index. In fact, the correlation between year-over-year EPS growth and the ISM Manufacturing Index over the 17-year period was a robust 0.60, indicating a strong correlation between economic cycles and corporate profit growth. This correlation can be useful in framing sector performance relative to economic cycles, represented by the swells and dips in the ISM Manufacturing Index.

Where are we in the current economic cycle?

The ISM Manufacturing Index also displays a vibrant relationship to interest rates, further highlighting its ability to capture economic cycles. Historically, the one-year change in the 10-year US Treasury yield has coincided with the one-year change in the ISM Manufacturing Index, as evidenced by a 0.52 correlation. As of April, the ISM Manufacturing Index was 57.3 below its February 2018 peak of 60.8. In my view, this suggests that US economy is mid-cycle, but on its way to being late-cycle.

• Materials have tended to underperform when the ISM Manufacturing Index is falling from its cycle peak, and have tended to outperform when the index is increasing. The excess return of the materials sector now looks quite cyclical following the direction of the ISM Manufacturing Index.
• Utilities also display cyclicality, but their excess return moves opposite to the excess return of materials. Utilities have historically generated excess returns when the ISM Manufacturing Index was falling and have underperformed when the index was rising.
• Consumer staples stocks are cyclical as well, displaying past weakness during periods when the ISM Manufacturing Index was rising through the early phase of a decline in the index. However, this sector has historically outperformed later in the cycle when the ISM Manufacturing Index was below 55 and falling.
• Health care has historically performed somewhat like consumer staples — performing best very late in the cycle and poorly early in the cycle.
• Industrials have historically been weakest very late-cycle when the ISM Manufacturing Index was under 55 and falling, but have shown strength through the upward cycle of the index (under 50 and rising, 50 to 55 and rising, and over 55 and rising). Moreover, industrials have a history of generating excess return very early in the downside cycle (ISM Manufacturing Index over 55 and falling). Conversely, industrials have historically performed most poorly very late in the cycle when the index was under 55 and falling.
• Financials have historically shown the greatest strength early-cycle when the ISM Manufacturing Index was rising from under 50 through the 50-to-55 range. They were weakest very late-cycle when the index was falling. Fallout from the housing bubble and global financial crisis may have distorted some of the cyclical influences since 2008.
• The performance of information technology has historically been choppy throughout economic cycles. The potential for technology to provide innovation may have muted the influence of the economic cycle on performance. However, technology has historically shown to perform very poorly very late-cycle and most strongly very early-cycle.
• Energy stocks are likely more sensitive to the price of oil, oil products and natural gas than the economic cycle. Their past performance relative to the ISM Manufacturing Index seems mixed, which may simply be random, rather than indicative of a clear pattern.
• Real estate — particularly real estate investment trusts (REITs) — has historically displayed strength mid- to late-cycle (over 55 and falling, and 50 to 55 and falling), and relative weakness very late-cycle (under 50 and falling) and very early cycle (over 50 and rising). REITs can be tricky, as they are often influenced by both interest rates and rent patterns. Rents could be under pressure very late- and very early-cycle, and strong through periods of the middle cycle.

In the second blog of this series, we saw that the S&P 500^® Low Volatility Rate Response generally achieved similar levels of volatility reduction as the S&P 500 Low Volatility Index. In our paper Inside Low Volatility Indices (published in 2016), the low volatility index historically outperformed the S&P 500 during severe market downturns (Exhibit 5 in the paper) due to lower beta to market and lower realized portfolio volatility. In this final blog, we examine if the rate response strategy achieves a similar level of volatility reduction as the low volatility strategy in those down-market periods.

We first looked at monthly up-market and down-market hit rates, along with the average excess monthly returns (see Exhibit 1).

For all periods, the two strategies had similar hit rates of about 50%, with the rate response index having a slightly higher hit rate (51% to 49%). While these two indices underperformed the S&P 500 half of the time, the returns seen in the first blog show that the rate response and low volatility indices outperformed the S&P 500 on a cumulative basis over the long-term investment horizon. What could be the cause of this? We investigated by looking at the monthly returns broken down between up markets and down markets.

In up markets, the low volatility and rate response indices underperformed the S&P 500 the majority of the time. This is not surprising, given their beta to market—Exhibit 2 shows the historical rolling 36-month betas of the two low volatility strategies compared with the S&P 500. The average 36-month beta for the rate response index was 69.9%, compared with 62.2% for the low volatility index. Nevertheless, the higher beta exhibited by the rate response index means upside participation of the strategy could be higher than that of the low volatility index. In fact, the average monthly underperformance for the low volatility index was 0.88% and 0.65% for the rate response index.

In down markets however, both indices outperformed the S&P 500 more than 80% of the time. The average excess return in down markets for the rate response index was 1.47%, slightly lower than the low volatility index (1.81%). Therefore, the relative outperformance of both indices versus the S&P 500 in down-market periods was markedly higher than the underperformance in up markets.

In addition, a simple compounding mathematical rule requires that for a given percentage portfolio decline, a higher percentage gain is required to get back to even. This compounding effect also helps explain the cumulative positive excess returns. Next, to study extreme bear markets, we looked at the three largest drawdown periods for the S&P 500 since 1991.

The rate response strategy performed similarly to the low volatility index—outperforming the S&P 500 in all three periods. The largest performance difference came during the tech bust in the early 2000s. While the S&P 500 dropped by over 47%, the rate response and low volatility indices had positive absolute returns. The cumulative outperformance impact relative to the S&P 500 can been understood by calculating the peak-to-recovery period return. For this date range, the low volatility index outperformed the S&P 500 by 91%, while the rate response index outperformed by 97%.

The analysis in this post shows that the rate response index was able to deliver similar historical performance as the low volatility index—more importantly, their down-market returns were similar. Along with the analysis shown in prior posts in this series, the S&P 500 Low Volatility Rate Response is a variation of the S&P 500 Low Volatility Index, designed for rising interest rate environments. The index reduces rising interest rate risk, while still delivering lower realized portfolio volatility—a salient characteristic of low volatility investing.

As advisors begin mid-year rebalancing of client portfolios, they face a dilemma.

While investors have enjoyed a 9-year Bull Market—including a 21.14% total return by the S&P 500 in 2017—their investment portfolios have built up more equity risk. Absent rebalancing, a standard 60/40 equity/bond portfolio might now resemble a riskier 75/25 allocation.

Given these sizable gains, many clients are reluctant to rebalance portfolios for fear of paying a big tax bill.

So how can advisors and their clients rebalance portfolios without incurring capital gains? Some are turning to a Buy-Write options strategy. The “Buy” part of the phrase includes stocks that an investor already owns, while “Write” refers to using options to rebalance the risk of client portfolios in a more tax-efficient manner.

Buy-Write in Four Steps

According to the CBOE, a Buy-Write strategy is “an investment strategy in which an investor buys a stock or a basket of stocks, and also writes covered call options that correspond to the stock or basket of stocks. Buy-Write strategies have an added attraction to some investors in that Buy-Writes can help lessen the overall volatility in many portfolios.”

Many investors execute this strategy around the CBOE S&P 500 Buy-Write Index (BXM), a benchmark index designed to track the performance of a hypothetical  Buy-Write strategy on the S&P 500 Index (SPX).

Here are some steps advisors consider when setting up a Buy-Write strategy for a client:

1. Determine equity beta for all holdings in a portfolio: This includes calculating the amount of equity market exposure inherent in their client’s underlying portfolio.
2. Calculate the difference between current and target exposure: This number will determine the magnitude of the overlay strategy that will run in tandem with the client’s portfolio.
3. Implement option overlay: Sell call options on the S&P 500 in the appropriate amount to bring current exposures in line with a client’s risk tolerance.
4. Monitor: Manage the option component in the existing account and make adjustments as markets move.

An illustrative example

The following scenario is familiar to many advisors: a client who just a few years ago had a $700,000 taxable account with a 60/40 equity/bond allocation now has a $1,000,000 nest egg. But outsized gains in the S&P 500 and other stock indices have pushed the portfolio to 75 percent equities and 25 percent bonds.

To rebalance back to 60/40, an advisor would need to reduce the portfolio’s equity exposure by $150,000, or 15 percent of the total portfolio value. Selling could mean a big tax hit and clients are often wired to avoid such actions.

Assuming the equity portion of the sample portfolio above is similar to the S&P 500, executing the buy-write strategy with S&P 500 call options can be a relatively simple and tax-efficient way to bring a portfolio back in line with client objectives.

To do so, an advisor could write, or sell, a call option on the S&P 500 Index (SPX) in relation to the client’s equity allocation that is overweight. In this case, implementing a 30 percent overlay of at-the-money calls would bring the portfolio’s risk profile back to its original 60/40 mandate.

In a previous blog post, I described some benefits of having an income goal and a method of estimating one’s retirement income liability. I reviewed a hypothetical example to illustrate steps 1-2 from the list of benefits below, and I calculated a net estimated retirement liability (the portion of retirement income one must fund with personal savings). This post will focus on the remaining steps, 3-5.

1. Income levels are intuitive because they provide built-in budgeting guidelines.
2. Estimating required future income can be tailored to individual circumstances and does not require specialized financial knowledge.
3. Estimating a future income level (that would be attained upon conversion of one’s savings into income risk hedging assets) does not require guesstimating future market returns the way that estimating one’s future wealth level would.
4. As a result of step 3, there can be more certainty about one’s future income than about one’s future wealth as long as some of one’s assets are managed to hedge income risk and more savings are devoted to those assets over time.
5. Once a desired future income level has been attained through the allocation to income risk hedging assets (essentially locking in future income), if other capital is available it can remain invested for long-term growth without great risk of income impairment during market downturns.

The reason step 3 is true is that a future income level is a future stream of cash flows, and we can find the present value of future cash flows with certainty by using the current, observable yield curve. Related to its S&P STRIDE Indices, each month, S&P DJI publishes the present value of specific inflation-adjusted cash flows (called “Income Cost”), commencing at particular points in the future. The cash flows begin at target years spaced over five-year increments from 2005 to 2060, and each one lasts 25 years from its respective target year. For example, the 2035 cash flow starts in January 2035 and continues until 2060. These hypothetical cash flow streams are designed to represent inflation-adjusted retirement income.

Using the Income Cost for 2035 (assuming that is when I want to retire), and continuing the example from my last post, if I have to fund $58,000 in annual spending, then I can measure the current value of income-hedging assets that would be required to lock in that cash flow in real terms, beginning in January 2035:

$58,000 X $19.29 (cost of 2035 income as of April 2018) = $1,118,670

If I invested $1.12 million in a portfolio of income-hedging assets, I would essentially lock in my inflation-adjusted retirement cash flow over a period of 25 years – which is the reason step 4 is true. The more of my assets that I dedicate to income-hedging assets, the more certain I can be of my future cash flows. The opportunity cost of doing so is that those assets cannot be invested for growth. However, the benefit is lowering risk to future income.

Finally, step 5 flows from step 4. Once I have an amount invested in income-hedging assets that, given the current income cost, secures future cash flows sufficient to meet my net estimated retirement liability, then any remaining assets can be invested for growth without risk to income. In other words, I can sleep soundly without worrying about market risk. Where does one find income-hedging assets? S&P STRIDE Indices incorporate income-hedging into their framework by allocating index weight to S&P U.S. TIPS sub-indices designed to hedge income risk.

Putting steps 1-5 into practice allows one to periodically check progress toward an income goal. In other words, you can measure your funded status, the currently funded portion of your net estimated retirement liability. A funded ratio, expressed as a percentage, assuming you have $500,000 of income-hedging assets, would be calculated as:

$500,000 ÷ $1,118,670 = 44.7%

Now that you can measure your funded status, it can be utilized to adjust current spending, saving, and portfolio allocations in order to manage future income risk. If you periodically add income-hedging assets, your funded ratio will grow over time. When it gets close to 100%, you have effectively hedged income risk over a period of 25 years, commencing at retirement.