Lower-Risk ETFs With High Risk-Adjusted Returns

Summary

• Risk, rank, return compared to peers, Ulcer Index, and Martin Ratio were used to select 25 funds in different Lipper categories.
• Different measures of risk and risk-adjusted return are compared.
• Three portfolios are compared in Portfolio Visualizer using these lower-risk, higher risk-adjusted return ETFs.

With any method for computing investment risk and performance, it is important to use data covering as long a time period as possible. In particular, the measurement period must include both bull and bear markets for the investments of interest.

- Peter G. Martin

How Do You Measure Risk in ETFs?

I discuss seven different measures of risk and three measures of risk-adjusted returns. A summary for portfolios from 1932 to 2017 is shown below for Maximum Drawdown, Standard Deviation, Sharpe Ratio and Martin Ratio. The table is from "How Bad Can It Get?" by Charles Boccadoro. It is a comparison of the Sharpe Ratio to the Martin Ratio for the past 85 years by stock-to-bond ratio. For interested readers, I described the Ulcer Index and Martin Ratio in The Great Owl Portfolio last October.

The table shows that the portfolio with the highest Martin Ratio consists of around 30% stocks, compared to 40% stocks for the Sharpe Ratio. I also like that the Martin Ratio shows a wider range of values, as I believe that it captures a better estimate of risk. For example, the Martin Ratio for a 70% stock/30% bond portfolio is 35% lower than a 30% stock/70% bond portfolio, while the Sharpe Ratio is only 4% lower. Also noteworthy is that the 70% stock/30% bond portfolio has twice the drawdown of the 30% stock/70% bond portfolio. Over a very long period of time, portfolios with higher allocations to stock outperform, but my investing time horizon is not 85 years.

(Source: MFO.)

High Risk-Adjusted Return ETFs

I selected 25 funds from 24 Lipper categories based on risk, risk-adjusted returns, and performance compared to peers using Mutual Fund Observer Premium Search. Many of these are appropriate for a late business cycle stage leading into a recession.

Symbol	Name	Lipper Category
GSY	Invesco Ultra Short Duration	Ultra-Short Obligations
BSCJ	Invesco BulletShares 2019 Corporate Bond	Short Invest Grade Debt
XMLV	Invesco S&P MidCap Low Volatility	Mid-Cap Core
USMV	BlackRock iShares Edge MSCI Min Vol USA	Multi-Cap Core
SPHD	Invesco S&P 500 High Dividend Low Vol	Equity Income
FVD	First Trust Value Line Dividend Index	Multi-Cap Value
SDY	State Street SPDR S&P Dividend ETF	Equity Income
MGV	Vanguard Mega Cap Value Index ETF	Large-Cap Value
XSLV	Invesco S&P SmallCap Low Volatility ETF	Small-Cap Core
DON	WisdomTree US MidCap Dividend	Mid-Cap Value
PZA	Invesco National AMT-Free Municipal Bond	Muni Gen & Insured Debt
BAB	Invesco Taxable Municipal Bond ETF	General Bond
FXU	First Trust Utilities AlphaDEX	Utility
FCOM	Fidelity MSCI Communication Services	Telecommunication
AOK	BlackRock iShares Core Conserv Alloc	Mixed-Asset Target Alloc Consv
GQRE	FlexShares Global Qual Real Estate Index	Global Real Estate
PSR	Invesco Active US Real Estate	Real Estate
AOM	BlackRock iShares Core Moderate Alloc	Mixed-Asset Target Alloc Moderate
AOR	BlackRock iShares Core Growth Alloc	Mixed-Asset Target Alloc Growth
TOK	BlackRock iShares MSCI Kokusai ETF	Global Large-Cap Core
IYLD	BlackRock iShares Morningstar Mult-Asst Inc	Alternative Global Macro
ACIM	State Street SPDR MSCI ACWI IMI	Global Multi-Cap Core
CGW	Invesco S&P Global Water Index	Global Natural Resources
DGT	State Street SPDR Global Dow	Global Large-Cap Value
DBEU	Deutsche Xtrackers MSCI Europe Hdgd Eq	European Region

Click to enlarge

The Measures of Risk and Risk-Adjusted Returns

One of the most common measure of risk-adjusted return is the Sharpe Ratio, which is the return above a risk-free treasury divided by the Standard Deviation (STDEV) that measures whether the price of a fund is rising, falling, or both. It is not a good measure of risk. The Sortino Ratio is the return above a risk-free investment divided by the downside deviation (DSDEV). The Martin Ratio is the return above a risk-free investment divided by the Ulcer Index, which measures the magnitude and duration of drawdowns.

The top section in the table below shows where the Sharpe, Sortino, and Martin ratios are in agreement. Over the past five years, these have been top funds on a risk-adjusted basis. GSY and BSCJ are both bond funds. The returns (APR) have been low, but the risk is even lower. XMLV and USMV are both low volatility funds. SPHD, FVD, and SDY are dividend funds.

Seven measures of risk and volatility are shown, including the minimum 1-year rolling return, Downside Capture and Price-to-Earnings Ratio. I use these more heavily in a late business cycle stage to reduce risk. I include P/E as a risk because investors have been purchasing low volatility funds, and as a result, valuations have risen. For more information, I refer readers to the article "Low Volatility Factor: High Valuation".

(Source: Created by the Author with Data from Mutual Fund Observer)

The bottom section consists mostly of global and mixed asset funds, where the Sharpe, Sortino, and Martin ratios are in agreement that these funds have not had as high of a risk-adjusted return. I like the low P/E of global funds, and they may be suitable in a portfolio for diversification.

The middle section consists of value funds, small- and mid-cap ETFs, municipal bonds, and Utility and Sector ETFs, where the risk-adjusted return has not been as high.

Portfolio Visualizer

The Efficient Frontier from Portfolio Visualizer is shown below. I have roughly grouped them as they are in the table above.

(Source: Portfolio Visualizer)

The next three charts show the returns for the past 12 months for the same groupings as above. It can be clearly seen that the Highest Risk-Adjusted Returns are less volatile than the other two categories.

(Source: Created by the Author with Data from Portfolio Visualizer)

Of the High Risk-Adjusted Returns, the First Trust Utilities AlphaDEX ETF (FXU) and the Invesco National AMT-Free Municipal Bond Portfolio ETF (PZA) have clearly been more stable investments.

(Source: Created by the Author with Data from Portfolio Visualizer)

The Moderate Risk-Adjusted Funds have not fared well during the past 12 months.

(Source: Created by the Author with Data from Portfolio Visualizer)

Morningstar

The Morningstar ratings are shown below. The risk ratings are low-to-average.

Symbol	Fund Category	Rating	Risk	Return	Return 3 Mon
GSY	Ultrashort Bond	4	Below Average	Above Average	0.9
BSCJ	Short-Term Bond	5	Above Average	High	0.6
XMLV	Mid-Cap Value	5	Low	High	2.8
USMV	Large Blend	4	Low	Above Average	2.4
SPHD	Large Value	5	Below Average	Above Average	2.4
FVD	Large Value	5	Low	Above Average	2.1
SDY	Large Value	5	Below Average	High	2.2
MGV	Large Value	4	Average	Above Average	-0.4
XSLV	Small Value	5	Low	High	2.0
DON	Mid-Cap Value	5	Average	High	2.3
PZA	Muni Nation Long	4	Above Average	Above Average	2.8
BAB	Long-Term Bond	4	Below Average	Average	3.9
FXU	Utilities	3	Average	Average	-0.5
FCOM	Communications	4	Average	Above Average	1.7
AOK	Alloc - 30% to 50% Eq	2	Low	Below Average	2.8
GQRE	Global Real Estate	4	Average	Above Average	5.5
PSR	Real Estate	4	Below Average	Above Average	4.5
AOM	Alloc - 30% to 50% Eq	3	Below Average	Average	2.6
AOR	Alloc - 50% to 70% Eq	3	Below Average	Average	2.6
TOK	World Large Stock	4	Average	Above Average	3.3
IYLD	Alloc - 30% to 50% Eq	4	Above Average	Above Average	4.5
ACIM	World Large Stock	4	Average	Above Average	2.5
CGW	Misc. Sector			Below Average	6.7
DGT	World Large Stock	3	Average	Average	2.7
DBEU	Europe Stock	5	Low	High	3.7

Click to enlarge

(Source: Morningstar)

My Favorite ETFs

I created three portfolios of the funds that I liked most, to maximize returns over the past 13 months for 4%, 6%, and 8% target standard deviations. While the returns are close for the three portfolios, my preference is towards less risk.

		Portfolio Allocations
Name	Ticker	4% Vol	6% Vol	8% Vol	SPY
Invesco Ultra Short Duration	GSY	30%	25%	30%
Invesco BulletShares 2019 Corp Bd	BSCJ	25%	20%	5%
Invesco S&P 500 Hi Div Lo Vol	SPHD	5%	10%	20%
First Trust Value Line Div	FVD	17%	25%	25%
Invesco Nat. AMT-Free Muni Bd	PZA	13%	10%	5%
iShares Core Cons Alloc	AOK	5%	5%	5%
iShares MSCI Kokusai	TOK	5%	5%	10%
	Portfolio Metrics
	Return	2.8%	3.1%	3.4%	5.5%
	Max. Drawdown	-2.1%	-3.1%	-4.6%	-13.5%
	Sharpe Ratio	0.22	0.24	0.23	0.29
	Sortino Ratio	0.32	0.33	0.32	0.41
		Portfolio Exposure
	US Stocks	24%	36%	49%
	Intl. Stocks	5%	6%	7%
	US Bonds	41%	34%	24%
	Intl. Bonds	9%	8%	6%
	Cash	21%	17%	13%

Click to enlarge

(Source: Created by the Author with Data from Portfolio Visualizer)

Here is how the portfolios and funds have performed over the past 13 months. Each of the three portfolios performed in a low volatile manner, with close returns in spite of the large difference in stock allocation.

(Source: Portfolio Visualizer)

Below is the fund performance over the past 12 months.

Name	Ticker	Avg Rtn	Stdev	Max. Drawdown	Sharpe	Sortino
Invesco Ultra Short Duration	GSY	2.6%	0.3%	0.0%	2.2	6.1
Invesco BulletShares 2019 Corp Bd	BSCJ	1.8%	0.7%	-0.2%	-0.2	-0.2
Invesco S&P 500 Hi Div Lo Vol	SPHD	3.3%	13.9%	-7.5%	0.2	0.2
First Trust Value Line Div	FVD	5.6%	13.1%	-7.6%	0.3	0.5
Invesco Nat AMT-Free Muni Bd	PZA	1.0%	3.3%	-2.5%	-0.3	-0.4
iShares Core Cons Alloc	AOK	0.8%	5.1%	-4.1%	-0.2	-0.3
iShares MSCI Kokusai	TOK	2.5%	15.7%	-13.5%	0.1	0.2

Click to enlarge

(Source: Created by the Author with Data from Portfolio Visualizer)

The following table shows the correlations between the funds. The equity funds are most correlated to each other. PZA is the least correlated to other funds.

(Source: Created by the Author with Data from Portfolio Visualizer)

Summary

Each month, I evaluate whether I should make a small change to my portfolio in order to reduce risk and increase risk-adjusted returns. This month, I traded a volatile health care fund for a less volatile dividend and income fund following the concepts in this article.

Analyst’s Disclosure: I am/we are long FXU. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

I am an engineer with an MBA nearing retirement and not an economist nor an investment professional. The information provided is for educational purposes and should not be considered as advice. Investors should do their due diligence research and/or use an investment professional.

Comments

[exeditor]

Hello, Charles. I know you are invested in VFMF, so the following link might interest you. I suspect there are Dow ETFs that could be as cheap at the Vanguard Multi-Factor.

www.fortunefinancialadvisors.com/...

[Charles Bolin]

Thank you, @exeditor. I did own VFMFX. It was more volatile than the S&P500 during the recent correction. I sold it when it recovered. I am focusing on funds that do well in down turns. By the way, I was not able to open the link. Regards.

[Charles Bolin]

Another excellent article, @exeditor, thank you.

[exeditor]

www.etf.com/...

Hello, Charles. You might enjoy this.

[Baldy2000]

Charles, Thanks for your comments, Cheers.

[Baldy2000]

Charles, have you looked at SPLV. It has only 5 years of data, so has not been recession tested. I will have to study you list more closely. I currently have $$ in 6 mo treasuries to provide protection for my DGI portfolio. Cheers

[Charles Bolin]

@Baldy2000, I own a small amount of SPLV, but no, I have not written about it in an article. It is a good fund. What I like about Mutual Fund Observer is it's risk analysis of funds. The Ulcer Index is 3.2 for SPLV and 3.5 for USMV compared to 5.8 for SPY. The drawdown of these two low volatility funds is only 40 to 50% of SPY. The P/E are comparable to SPY.

[RettW]

Standard deviation also assumes a normal distribution for these funds, which is not the case. Multiplying s.d. by 6/(6+skew) seems to probabilistically estimate their fat tails well.
How might you use Relative Strength from ETF screener to further protect your portfolio from downside, and still ride upside?

[6264541]

Could you clarify your comment? Standard deviation Is defined independent of any probability distribution. It’s significance depends on the probability distribution.

[Charles Bolin]

Thank you for you comment, @Sanya Robert. Standard Deviation, Sharpe Ratio and the other metrics are time dependent. Standard Deviation typically gets larger during bear markets. Secondly, the time period was short making the metrics more reflective of conditions instead of long term performance. I used the last 13 months to show the recent volatile trends. Returns were lower and standard deviation higher. Sharpe is the return above treasuries divided by standard deviation.

[Sanya Robert]

Thank you for the work and the excellent series. What is puzzling to me is the Sharpe ratio for your portfolio is quite low.

[keltus 1952]

Holding SPHD. It pays monthly. Fees a little high @ .30%. Nice piece of my pie though.

[6264541]

Charles,

I did a backtest on your 4% volatility portfolio back to Nov 2012

https://bit.ly/2Hf3egr

and got downside capture ratio of 21.14 %. That could indicate be a significant drawdown in a coming recession. Your thoughts on this.

Jeff

[Charles Bolin]

Isn't Portfolio Visualizer great, @6264541? If the S&P500 goes down 40% and the portfolio captures 20% then the drawdown would be 8% which is good for a portfolio of 35% stocks. If one is more risk adverse than this then they could reduce stock exposure to 20%, but this allocation will not keep up with a 4% withdrawal rate and inflation. I believe many if not most investors want a portfolio to buy and hold and are okay with an 8% drawdown. I try to set my allocation according to the business cycle.

[6264541]

I agree that the Portfolio Visualizer optimization to is great. Thank you for pointing it out in your articles. I agree that your 4 % portfolio is good, but I am trying to do a little better by using some more aggressive assets. So by using PONAX, SPLV, QQQ, and BTAL I got some interesting results. I limited PONAX below 40 % and got these results:

1) Maximize return at 4 % volatility

CAGR of 8.16 %, MDD of 2.59 %, upside capture 34.9 %, downside capture of 10.7 %.

https://bit.ly/2tV4yxe

2) Minimize variance

CAGR of 7.29 %, MDD of 2.85 %, upside capture 28.9 %, downside capture of 3.4 %.

https://bit.ly/2Tox751

Essentially these portfolios have comparable capture ratios to PONAX (31 % up/ 5 % down) with better returns in 2018.

Jeff

[Charles Bolin]

Nice, Jeff. You can set the Optimization Goal to "Maximize Return subject to" and the "Compared Allocation" to "Maximum Sharpe Ratio Weights" and get the results in one run.

These are interesting results. Looking at the Efficient Frontier or Assets, BTAL has lost nearly 2% on an annual basis with a standard deviation of 11%. It had a maximum drawdown of 22%. I see that it moves in the opposite direction of the other funds which will reduce volatility. I substituted MINT for BTAL and the return went down and the MDD increased.
Charles

[jackcollinsjr]

Charles,
Once again, thanks for your work. I assume you're using the Target Volatility Market Timing Model in PV to arrive at the portfolio weights in the article. This model requires 'seed' values for the portfolio (asset symbol and %), what values did you use to seed it? I've entered various values and target Vols, but get different numbers from yours. It might be easiest if you could export the link (in the Market Timing line) and copy it in a reply.
Thanks in advance, Jack.

[Charles Bolin]

Thank you, @jackcollinsjr. I used 4%, 6%, and 8% as target volatility and they will be close to the results. I also use fund constraints and category constraints to ensure diversification and to avoid allocating too much to a non-core fund. Finally, I round the allocations up to the nearest 5% for the article. Sorry, but I only saved the PDF of the output and do not have a link.
Regards

[Charles Bolin]

Thank you, @tmow. I am planning an article on Global Funds soon and it will include VMNVX and ACWV.

[jgedmond]

Looking forward to that article as I have used these funds for international exposure along with EFAV.

[tmow]

Super article Charles! You covered a lot of my favorite ETFs. I would add two global funds for your consideration (VMNVX and ACWV) and as far as risk-adjusted returns go, PIMIX is a bond fund that has superior risk-adjusted returns.

[meowy]

very cool article wjth lots of data. i am comfortable with Vanguard Wellesley and Global Minimum Volatility funds in terms of risk/reward. Lots of diversification and strong exposure to developed and emerging markets. high quality bonds and stable stocks and decent yield between the two of them. cash is Vanguard Prime MMF.

[Charles Bolin]

Thank you. My strategy is to reduce risk without being bearish prior to the next recession.

[goyum]

Thanks Charles. Excellent analysis and very clearly written. Provides me with several new investment alternatives. Am still reinvesting and building three low-moderate tax/non taxable portfolios that will not require tactical monthly management, for about 10-20 MFs and ETF's. Still have to re-read your latest articles again.

[Charles Bolin]

You are welcome.