Why Currency Hedging Doesn’t Work in Canada

In the last two years, Canadian ETF providers have finally launched US and international equity ETFs that do away with currency hedging. Yet the strategy remains hugely popular: the hedged versions of Vanguard’s international and US total market ETFs remain much larger than their unhedged counterparts, while investors have more than $2 billion in the iShares S&P 500 Hedged to CAD (XSP), making it the third largest ETF in Canada.

None of my model portfolios include currency-hedged funds: I’ve long argued the strategy is expensive and imprecise. Even when the Canadian dollar appreciates strongly, the high tracking error of currency-hedged funds often reduces any potential benefit. In one dramatic example, Justin Bender looked at the period from 2006 through 2011, when the US dollar depreciated by almost 13% and hedging should have produced a huge boost: in reality, XSP lagged its US-listed counterpart.

This leads to an interesting question. If currency hedging were free and precise—with an expected tracking error of zero—would it be worth considering?

Does hedging lower volatility?

The most common argument in favour of currency hedging is that it lowers volatility. In Unconventional Success, David Swensen—the legendary manager of the Yale Endowment—explains that currency diversification is a benefit, but only to a point. He says one should hedge foreign currency once it exceeds 25% of the portfolio’s assets: “Beyond a quarter of portfolio assets, the currency exposure constitutes a source of unwanted risk.”

Other studies also suggest too much foreign currency exposure creates more volatility with no increase in expected returns—which is, of course, a lousy combination. A research paper prepared for the IMF looked at data for France, Germany, Japan, the United Kingdom and the US, concluding that “currency hedging appears to effectively reduce the variance of foreign investment returns not only at short investment horizons but also at horizons of up to 5 years in most cases.”

These findings will likely raise the eyebrows of investors who use my model portfolios, which have as much as 40% exposure to foreign currency. Wouldn’t it make sense to hedge at least part of the US or international equity holdings? No, it wouldn’t. It turns out the idea that hedging lowers volatility simply doesn’t hold up in Canada.

A couple of industry white papers—one Canadian and one American—have provided evidence to support this idea. Exhibit A comes from Pyramis Global Advisors: their research looked at data from 1990 through 2009 and found that for Canadians a fully unhedged portfolio had the lowest volatility: hedging half the foreign currency increased the annual standard deviation of returns, and hedging all the currency bumped it even higher.

In another analysis, researchers at J.P. Morgan Asset Management looked at the effect of hedging on investors in the US, Japan, the United Kingdom, Germany, Switzerland, Australia and Canada. They found hedging at least part of an equity portfolio’s foreign currency risk lowered volatility in the first five countries, but in Canada and Australia the strategy was counterproductive. “A currency hedge is not a diversifier within these two countries,” they wrote.

Loonie tunes

Why isn’t hedging effective in Canada? The Pyramis researchers explain that the US dollar, euro and Swiss franc tend to have negative correlation with the global equity markets. (Recall that when all risky assets plummeted in 2008, the US dollar soared.) Negative correlation is what diversification is all about: any part of your portfolio that goes up when equities go down is a welcome addition, so exposure to these currencies is a benefit, and hedging wipes it out. As the J.P. Morgan authors write: “The hedge will tend to produce profits at the same time that equity markets are advancing, and produce losses when equities are falling.” In other words, it magnifies volatility rather than reducing it.

Still not convinced? The Canada Pension Plan offers another commentary in its 2013 annual report. The CPP Investment Board sees “no compelling reason to hedge equity-related currency exposure,” largely because “hedging would unduly tie Fund returns to the price of oil and other commodities as they drive the foreign exchange value of the Canadian dollar.” (Incidentally, that explanation helps explain why hedging is equally dubious in Australia, where the dollar is also correlated with commodity prices.)

The evidence seems clear that hedging is a costly strategy that actually increases risk and frequently fails to offer a benefit even when the Canadian dollar appreciates. How long before the popularity of currency-hedged index funds finally wanes?

 

52 Responses to Why Currency Hedging Doesn’t Work in Canada

  1. Shaun February 1, 2016 at 4:03 pm #

    @CCP
    Hi Dan,
    Another solid post here. I agree with everything you have written in this one.

    Since the foreign withholding tax is a real and fairly consistent drag on returns, would it not be appropriate to incorporate some estimate for these costs in your Model Portfolio backdated returns? If I read it right you are only assuming a 0.28% total drag for your estimation of what VXC would have done prior to inception in your model portfolio return methodology.
    http://canadiancouchpotato.com/wp-content/uploads/2016/01/CCP-Model-Portfolios-Vanguard-2015.pdf

    The longest running foreign unhedged ETF in Canada that I found is XEM (iShares emerging markets index). It started in June 2009. It has lagged the benchmark by 14.02% cumulative, or 1.55% annualized since inception through to the end of 2015. The Management fee and MER are 0.82% and 0.78% respectively. So there is about 0.73% of annualized drag unaccounted for by fees. I assume that this is mostly related to foreign withholding taxes as it lines up closely with your 0.7% estimate, plus emerging markets have even higher yields.
    https://www.blackrock.com/ca/individual/en/products/239636/ishares-msci-emerging-markets-index-etf

  2. Canadian Couch Potato February 1, 2016 at 4:59 pm #

    @Shaun: The index returns we use for the international and emerging markets equities are already reported net of foreign withholding taxes (at least since 2000) so this accounted for already. For the record, ETFs report their returns net of foreign withholding taxes, too, so if you are able to recover the tax in a non-registered account, your personal rate of return could be higher than the published return.

    I have not drilled into the numbers, but your estimates for XEM seem very reasonable. Foreign withholding taxes on “wrap” ETFs like XEM can be quite significant indeed.

Leave a Reply