In recent months I have written about defensive bond strategies that can be used to replace an individual risk-off asset employed in tactical strategies. For example, many tactical strategies utilize a long-term treasury as a risk-off asset; i.e. when equities turn bearish in a tactical strategy, money is moved into long-term treasuries. In the past 30 years, such an approach has been profitable. However, in the current rising rate environment, moving to a long-term treasury in a risk-off posture may prove to be disastrous. It is possible that equities and treasuries may trend downward together. January - February 2009 is such an example when both equity and treasury markets were bearish at the same time.
My solution is to use a low volatility defensive bond strategy in place of a long-term treasury (or any other individual bond asset) in risk-off situations, i.e. when equities are trending downward. In this way, the risk-off asset is selected based on short duration momentum principles, and is not just a fixed asset. To ensure safety, the out-of-market asset in the defensive bond strategy is cash (money market).
Previously, I have developed defensive bond strategies such as the Defensive Bond Strategy [DBS] and the Modified Defensive Bond Strategy [MDBS] that employ mutual funds. However, there are some tactical strategies that use ETFs instead of mutual funds, and they require an ETF defensive bond strategy. This article will present an ETF version of MDBS applicable to Schwab accounts, i.e. using Schwab commission-free ETFs. I will conclude this article by applying the ETF defensive bond strategy to a conservative volatility targeting strategy that employs equity sector ETFs.
ETF Version of Modified Defensive Bond Strategy
Developing an ETF version of the Modified Defensive Bond Strategy (that I will call ETF_MDBS) is somewhat problematic since bond ETFs are very short-lived. Backtesting can only be performed over a very limited timeframe. The solution is to find mutual funds with long histories that can be used as proxies for the ETFs. Then backtesting to the 1980s can be performed to ensure the strategy will perform well over various market conditions.
Like the MDBS, the ETF_MDBS uses three different classes of bond assets and selects the one with the highest momentum at the end of each month. The three classes of bonds are:
1. Senior floating rate bonds or high yield corporate bonds. These types of bond assets perform well in a rising rate environment. In the first case, I have selected PowerShares Senior Loan ETF (NYSEARCA:BKLN) as the Schwab commission-free ETF. In the second case, I have selected PIMCO 0-5 Year High Yield Corporate Bond ETF (NYSEARCA:HYS) as the Schwab commission-free ETF.
2. High yield municipal bonds. I have selected Nuveen High Yield Municipal Bond ETF (NYSEARCA:HYMB) as the Schwab commission-free ETF.
3. GNMA bonds. This class tends to do well when equities fall. Schwab does not have a commission-free GNMA ETF. Thus, I have selected iShares Mortgage-Backed Securities ETF (NASDAQ:MBB). A $9 fee will be paid for every transaction of MBB on Schwab. This cost will be relatively small for a reasonable account size.
I decided to use the dual momentum timing tool in Portfolio Visualizer [PV] with a 3-month look-back period. Changing the look-back period between 2 - 4 months did not have a significant effect on backtest results.
The final Schwab ETF defensive bond version uses HYS, HYMB and MBB as its universe. The correlations between these ETFs from 06/17/2011 and 11/23/2016 are shown below. It can be seen that the ETFs are not well correlated, as desired. The daily standard deviations [DSDs] are 0.38% for HYMB, 0.31% for HYS and 0.17% for MBB. Although these DSDs are not as low as their corresponding mutual fund counterparts, they are sufficiently low to be used in the ETF_MDBS with a short duration look-back period.
Here is a PV link of ETF_MDBS. The PV backtest results from 2012 - October 2016 are shown below. The ETF_MDBS results are listed as "Timing Portfolio" in the figures. Also included in the figures are equal weight buy & hold portfolio results (HYS, HYMB and MBB rebalanced annually) as well as Schwab US Aggregate Bond ETF (NYSEARCA:SCHZ) for reference.
It can be seen that ETF_MDBS has a Compounded Annual Growth Rate [CAGR] of 8.4%, a Maximum DrawDown [MaxDD] of -2.5%, a Worst Year of +2.0% (2015), and an annualized Standard Deviation [SD] of 3.8%. The monthly win rate is 72.4% and the adjusted return-to-risk ratio MAR (CAGR/MaxDD) is 3.4. There are only two drawdowns greater than -2.0%. These results are much better than the results from the equal weight portfolio or SCHZ.
As a comparison, PV was run with proxy mutual funds: Eaton Vance Income Fund (MUTF:EVIBX), MFS High Yield Municipal Bond Fund (MUTF:MMHYX), and Vanguard GNMA Fund (MUTF:VFIIX). The results are presented below. The timeframe is 2012 - October 2016, i.e. the same timeframe as the ETF results. The CAGR = 8.2%, MaxDD = -3.1%, Worst Year = +4.8%, and SD = 3.3%. These results compare favorably with the ETF version shown above for the same timeframe. This means the ETFs are well represented by the mutual fund proxies.
As a final step in validating the ETF_MDBS, I present the PV results from 1987 - October 2016 using the proxy funds. They are shown below as the "Timing Portfolio." Also shown are the results of an Equal Weighted Portfolio (rebalanced annually), and the Vanguard Total Bond Market Index Fund (MUTF:VBMFX).
The ETF_MDBS using proxy mutual funds has a CAGR of 10.4% and a MaxDD of -5.2%. The Worst Year is 0.0% (2008). In comparison, VBMFX has a CAGR of 6.2%, a MaxDD of -5.9%, and a Worst Year of -2.6% (2013). Thus, the ETF_MDBS has better performance and lower risk than VBMFX.
Application to Volatility Targeting Strategy
This section will apply the ETF_DBS to a volatility targeting tactical strategy that features defensive sector equity ETFs. I have presented a similar volatility targeting strategy in a previous SA article. The risk-on sector equities are taken from the healthcare, utilities and consumer staples sectors. These are considered defensive equities since they have lower volatility than other equity sectors and they tend to do better when the overall market (S&P 500) is bearish. I selected Schwab commission-free ETFs from these sectors, and will use mutual funds as their proxies for longer backtesting. The ETFs and mutual fund proxies are listed below:
I selected a 2.5% Target Volatility parameter, and I only used downside volatility (i.e. "Yes" in the PV box "Use Downside Volatility Only"). The look-back volatility period is one calendar month.
Here are the PV results for the mutual fund version that goes back to 1990. The timeframe is limited by the risk-off strategy I used [MDBS] that only goes back to 1990. So for these backtest results I have used VGHCX, FSUTX and FDFAX (each with a 33.3% weighting) as the risk-on assets, and MDBS as the risk-off asset. Please refer to my previous article on MDBS to understand the details of MDBS.
It can be seen that this defensive volatility targeting strategy has a CAGR of 10.2%, a MaxDD of -6.3% (second highest drawdown is -3.9%), a Worst Year of -2.4%, and a SD of 5.2%. These numbers can be compared to buying and holding a Fidelity Balanced Fund (MUTF:FBALX) that gives a CAGR of 8.8%, a MaxDD of -40.5%, a Worst Year of -31.3%, and a SD of 9.9%.
Next, I ran PV with the ETF version of the volatility targeting strategy. The risk-on ETF assets are RYH, RYU and RHS (equal weighting), and the risk-off asset is ETF_MDBS. Each month the split between the equities and ETF_MDBS is determined based on the downside volatility of the equities over the previous month.
The PV results are shown below for 2012 - October 2016.
As to be expected, the results for the ETF version of the volatility targeting strategy over a bullish equity market period indicate better performance and lower risk than shown previously for the longer timeframe of the mutual fund version. For the ETF version, CAGR = 11.7%, MaxDD = -3.4%, Worst Year = 3.0%, and SD = 5.6%. Over this shorter timeframe, FBALX produces better numbers than it did over the longer timeframe, but its CAGR is still about 2% less than the ETF version of the volatility targeting strategy, and the MaxDD is twice as high.
To show the benefits of using ETF_MDBS as a risk-off asset, SCHZ is used in place of ETF_MDBS for the volatility targeting strategy. The overall backtest results from 2012 - October 2016 using SCHZ as the risk-off asset is presented below.
It can be seen that the CAGR is greatly reduced (8.9% using SCHZ versus 11.7% using ETF_MDBS) and the MaxDD is increased (-4.1% using SCHZ versus -3.4% using ETF_MDBS) when SCHZ is used as the risk-off asset.
This article presents an ETF version of the Modified Defensive Bond Strategy [MDBS]. Previously, I developed MDBS using mutual funds, but it is wise to use an ETF version (ETF_MDBS) when a risk-off asset is needed in an ETF tactical strategy. This article shows that ETF_MDBS used as a risk-off asset in a volatility targeting strategy produces higher annual growth and lower risk compared to using a single aggregate bond ETF like SCHZ as the risk-off asset.
Although ETF_MDBS can only be backtested over a short timeframe (2012 - present), mutual funds are used as proxies for the ETFs, and the mutual fund version of MDBS is backtested to 1987. The mutual fund version shows good performance and low drawdown in the backtest results. This gives confidence that the ETF version of MDBS will perform well into the future, and can be used to select the risk-off asset in tactical ETF strategies.
Disclosure: I am/we are long HYS. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.