Do Low-Volatility ETFs Do What They Say on the Tin?

Low-volatility ETFs aim to deliver lower volatility than market-wide ETFs. They can lose less when the market drops and lag when it rises.

To measure performance, I used a broad-market ETF as a benchmark in each country. For Canada, I used the iShares Core S&P/TSX Capped Composite Index ETF (XIC.TO) because it covers most of the Canadian stock market. For the U.S., I used the iShares Core S&P 500 Index ETF (XUS.TO) as a large-cap benchmark because the funds in this comparison generally pick from large and mid-sized U.S. companies.

The Canadian funds performed similarly despite the differences in methodology I discuss below. Over the three years I looked at, their annualized volatility ranged from 9.1% to 9.4%, compared with 12.8% for XIC.TO. They returned 18.8% to 21.3% a year, versus 25.1% for the benchmark.

The U.S. funds were more spread out. Their annualized volatility ranged from 10.2% to 12.0%, compared with 14.8% for XUS.TO. They returned 11.8% to 15.6% a year, versus 21.5% for the benchmark.

Over the three years I looked at, all seven funds did what they said on the tin and delivered lower volatility than their benchmarks.

Low-volatility funds differ in how they define and measure the volatility they seek to minimize. Some seek to reduce how much the portfolio moves with the wider market. Others seek to reduce day-to-day swings in the portfolio’s value or focus on how far it falls on down days. Here are three ways funds define stock volatility.

Beta: How much does the stock move with the market?

The first measure is beta, which tells you how sensitive a stock is to the wider market. A stock with a beta of 1 has tended to move about 1% in the same direction for every 1% move in the market. A stock with a beta of 0.1 has tended to move about 0.1% for the same market move.

Standard deviation: How much does the stock move on its own?

Standard deviation measures how far a stock’s returns tend to stray from their own average. Say a stock falls 1% one day and rises 3% the next. The average of those two returns is 1%. Each day’s return deviates from the average by two percentage points.

Downside volatility: How big are the moves below a threshold?

An index can use zero as the threshold and focus on the size of negative returns. Take the same stock that falls 1% one day and rises 3% the next. The 1% loss counts toward downside volatility. The 3% gain contributes zero.

Once I know how a fund measures volatility, the next question is how it decides how much of each stock to hold.

Weight inversely to volatility

One approach is to give more weight to stocks with lower beta or lower volatility. Suppose three stocks have volatilities of 10%, 20% and 40%. If weights are inversely proportional to volatility, the stock that swings half as much gets twice the weight. Give the 40% stock one part, the 20% stock two parts and the 10% stock four parts. That’s seven parts in total. The three stocks get 4/7, 2/7 and 1/7 of the portfolio, or about 57%, 29% and 14%, respectively.

Look at the whole portfolio

The second approach uses an optimizer, a computer model that looks at how stocks move together. Take three made-up stocks. Bank A and Bank B each have annual volatility of about 15%, and they usually rise and fall together. Utility C has volatility of about 10% and often moves on different days. The optimizer might put most of the money in the utility and some in the banks. If the banks and the utility move differently enough, the mix can move less than any of them on its own.

Start with company size and adjust the weights

The third approach starts with company size, as a regular market index would, then uses risk scores to tilt the weights. A big company can still get a big weight. The fund can give it less weight than it has in the market index if it scores as more volatile, or more weight if it scores as less volatile.

Canadian investors can find low-volatility ETFs for their home market, the U.S. and overseas. I’ve picked examples to show how the approaches differ. You can find more through the low-volatility factor filter on DecodeETF’s Browse page.

1. Beta + Inverse weighting

BMO ranks stocks by beta and gives more weight to the ones that have been less sensitive to the market. ZLB.TO applies that approach to Canadian stocks. ZLU.TO applies it to U.S. stocks.

The same approach is used by the BMO Low Volatility International Equity ETF (ZLI.TO) and the BMO Low Volatility Emerging Markets Equity ETF (ZLE.TO).

2. Return volatility + Inverse weighting

ULV.C.TO follows an index that picks the 100 eligible S&P 500 stocks with the smallest price swings over the previous year. The lowest volatility stocks get the biggest weights. There’s no sector cap, so a sector can take up a large share of the fund if many of its stocks qualify.

3. Return volatility + Optimizer

iShares funds such as XMV.TO and XMU.TO follow MSCI indices that use an optimizer to look for a combination of stocks that together produce lower volatility. The optimizer limits how far the fund’s sector weights can stray from the broader index.

Other iShares examples in the chart are the iShares MSCI Min Vol Global Index ETF (XMW.TO), the iShares MSCI Min Vol EAFE Index ETF (XMI.TO) and the iShares MSCI Min Vol Emerging Markets Index ETF (XMM.TO).

TD uses its own optimizers to build low-volatility portfolios for the TD Q Canadian Low Volatility ETF (TCLV.TO), the TD Q U.S. Low Volatility ETF (TULV.TO) and the TD Q International Low Volatility ETF (TILV.TO).

4. Downside volatility + Optimizer

The CI Global Minimum Downside Volatility Index Fund, ETF C$ Unhedged Series (CGDV.B.TO) follows a Solactive index that looks for a mix of developed-market stocks with smaller swings on the downside. The index sets limits on individual holdings and on how far sector and regional weights can stray from its starting universe.

5. Combined measures + Score and tilt

Fidelity funds such as FCCL.NE combine price swings, beta and earnings stability into a score. The index uses those scores to help pick stocks within sectors, then shifts some weight from more volatile sectors toward steadier ones. That makes sector weights another way this approach can differ from an optimizer with tighter sector limits.

The three low-volatility funds that invest in Canadian stocks had similar results. They had annualized volatility between 9.1% and 9.4%. Their worst drops ranged from 7.2% to 8.4%. The funds that invest in U.S. stocks were more spread out on both measures.

Why is this? XIC.TO holds a little over 200 stocks, while XUS.TO gives you exposure to roughly 500. A smaller pool of large companies may help explain why the three funds land on some of the same Canadian stocks despite different methodologies.

Indeed, the average pair of funds investing in Canadian stocks shared 57% of its holdings by weight. For funds investing in U.S. stocks, the average was 30%.

Financials were the largest sector in all three funds that invest in Canadian stocks, at 26.5% to 35.0% of each portfolio. They differ in energy exposure. XMV.TO held 15.4% in energy, compared with 10.2% for FCCL.NE and 1.7% for ZLB.TO. XMV.TO’s index limits how far its sector weights can stray from the broader MSCI Canada index.