If investors want to even out the volatility in the U.S. stock market by diversifying into international equities, they should consider a risk-aware approach, said speakers at “Upgrading Your Global ETF Equity Allocation Strategy,” a Monday panel discussion at the ETF Strategy Summit in Dallas.

By market capitalization, U.S. stocks make up approximately half of the investible public equity universe, said Joe Smith, senior market strategist at CLS Investments, but CLS tends to overweight international stocks.

“We’ve been fairly overweight towards international over U.S.,” said Smith. “Our benchmark internally for our equities sleeve is now 45 percent in the U.S., 55 percent international.”

Alex Piré, head of client portfolio management at Natixis affiliate Seeyond, explained that international investing looks differently depending on where an investor is domiciled. U.S. investors tend to have a larger concentration in the U.S., keeping about 60 percent of their portfolios focused on domestic stocks, with half of that U.S. allocation in large capitalization companies. The remaining 40 percent is typically split 50-50 between emerging and developed markets. Natixis’ European investors, on the other hand, tend to be more internationally focused and keep a smaller allocation to domestic stocks.

Seeyond is subadvisor for MVIN, the Natixis Seeyond International Minimum Volatility ETF, which uses a quantitative model to create an investment universe of international stocks with low volatility, then active management to select the best investments from a risk perspective, said Piré.

Both Smith and Piré argued that factors were a more effective method for thinking about international diversification than fundamentals, sector or geography.

“The way we’ve thought about building portfolios and managing risks has been along factor lines, where we want to harvest various risk premiums,” said Smith. “For us that’s a more important consideration than price.”

For one thing, sector classifications as defined in the U.S. do not easily translate to many global markets, said Piré.

Piré used Partners Group, a Switzerland-based company classified in the financial services sector, as an example. The bulk of Partners Group’s business involves buying concessions in toll roads and other infrastructure around the world. The company’s performance is not at all indicative of the Swiss market or Swiss financial services, he said.

“We tend to take for granted the precision of sectors in the U.S.,” said Piré. “We all understand that in the U.S. biotech companies, financial companies, technology companies can all be classified properly. When you go into developed markets and emerging markets, those classifications lose their ability to truly justify what a company’s doing.”

Yet investors should be careful when investing in factors, said Piré, because passive, smart beta approaches to factors like low volatility and momentum are unlikely to rebalance often enough to capture all of the benefits of holding an allocation to momentum or low volatility.

Though CLS uses some smart beta products in its ETF strategies, Smith agreed.

“We’re always trying to measure the degree of purity of a factor exposure,” said Smith. “What are the other consequences that you’re getting from exposure to a factor? The timing to a rebalance versus the purity of an exposure at a point-in-time, for something like momentum especially, can be a little worrisome. For factors more driven by fundamentals, the information is already there and isn’t really moved by price.”

Other factors operate over such long-term cycles that investors may not have the patience to hold an exposure long enough to achieve its performance benefits.

For example, the value premium has all but disappeared in U.S. equities since the 2008 financial crisis, said Smith, but when investors look across a longer time span, or globally, the benefits of value as a factor still exist.

“When you look at data over a much longer term, you see that the value premium has been persistent and pervasive across markets,” said Smith. “People tend to forget that the value premium takes a while to pay off. If you’re looking at a window of a year, or three years, there’s a lot of research to show that you’re not going to see value magically working over shorter periods of time.”

Smith also described ESG methodologies as risk mitigation for international investing. CLS tends to avoid countries with weak governments.

Seeyond uses a similar method to remove countries with high levels of government intervention or poor corporate governance, said Piré, before appyling a more formal negative ESG screen to weed out specific companies with low ESG scores

“We think about performance in an ESG framework,” said Piré. “Our studies on our own strategies show that if we have this layer excluding ESG worst offenders embedded in the strategy, we end up taking on less risk. There is no drag on performance, and even if ESG doesn’t enhance performance it can potentially be an additional risk management layer in our process that isn’t picked up by the quantitative model.”