A PDF version of this commentary can be downloaded here.
It took less than 3 hours for Jeremy Schwartz – Director of Research at WisdomTree – to email me and tell me he thinks I am wrong.
On Friday night, I posted a brief article on Forbes.com titled The Ups and Downs of Currency Hedged ETFs, which was meant to address both the growing interest in currency hedged ETFs as well as the implications of currency hedging. With the U.S. dollar appreciating nearly 23% against the euro over the last year – requiring European stocks to return 23% just for a U.S. investor to break even – currency hedging has become a popular topic.
My view in the article was that currency hedging can cut both ways, and that without a particular view on where currencies are headed, there may be diversification benefits of remaining unhedged.
The post went live at 6:18PM. Jeremy’s email was sitting in my inbox by 8:50PM. Now, Jeremy has been on the forefront of the currency-hedged movement for the last 5 years, so it comes as no surprise that he disagrees with my conclusion. But Jeremy is also an intellectual titan for whom I have a tremendous amount of respect: if he disagrees with me, I’m going to deeply consider his argument.
In the interest in sharing his wisdom, I thought I would address his arguments here:
1. Equity risk comes in two forms: valuation risk and fundamental risk (e.g. earnings or dividend per share growth). A strong dollar should serve as a headwind to U.S. fundamentals (as it hurts both exports as well as revenue and earnings coming from overseas) but a tailwind to foreign fundamentals. To diversify away from deteriorating U.S. equity fundamental risk, we would want to hold European equities – not the euro itself.
For a simplified thought experiment, this is trivially true. If the dollar appreciates against the euro, it should be a tailwind to European companies that export to the U.S. Their shares should also appreciate in value. To ensure we capture this full appreciation, we would want a hedged position – otherwise the gains in equity would be stifled by the losses in currency exposure. For export-driven economies, currency and equity prices should be inversely correlated.
Stepping back from a simplified world, however, we have to ask two questions. (1) “How sensitive is EPS/DPS growth to fluctuations in currency?” and (2) “what are the relative sizes of valuation risk and fundamental risk?”
To look at question #1, we can grab EPS and DPS for the S&P 500 from Robert Shiller and download the Real Trade Weighted U.S. Dollar Index (against major currencies) from the Federal Reserve Bank of St. Louis. From 1971 to 2000 (I’ve selected to ignore the 2000s due to the account-driven implosion of EPS in 2003 and 2009), monthly EPS and DPS changes had correlations of -0.06 and 0.02 to monthly changes in the U.S. Dollar Index. Looking at year-over- year changes instead of month-to-month doesn’t improve the numbers much. While certainly not conclusive, evidence suggests that fundamental risks may not be heavily tied to currency fluctuations, at least in somewhat “normal” environments. How closely currencies and fundamental measures are tied will certainly change country-to-country, based on idiosyncratic risks like inflation, interest rates, public indebtedness, terms of trade, et cetera. But the case certainly isn’t as clear cut.
To explore question #2, we will decompose price into fair value and deviations thereof. Fair value will be defined as a constant multiple of earnings based on the long term Shiller CAPE average over the period. Monthly changes of the fair value model had a volatility of 7.76% while the changes in valuation deviations had a volatility of 13.99%. In other words, from a sheer magnitude perspective, valuation risk trumps fundamental risk.
If we choose a more modern and moderate era – say 1960 through 2000 – the magnitude of these volatilities falls – at 2.14% and 8.48% respectively – but their size relative to one another clearly increases.
So here’s my take: there is no conclusive evidence, empirically, that changes in the dollar lead to changes in fundamental factors. Plus, valuation risk trumps fundamental risk anyway.
2. Euro volatility is about 10% annualized and both hedged and unhedged foreign equities have nearly the same correlation to the S&P 500. In fact, some evidence suggests that unhedged positions actually have increased volatility and beta to the S&P 500. An unhedged position is taking on uncompensated volatility – unless we believe the euro is going to appreciate.
I think this is trivially true when risk is defined as volatility. When a U.S. investor buys an unhedged foreign share, they’ve taken their dollar, converted it into a foreign currency, and purchased a foreign currency denominated equity share. In effect, the dollar is now invested in two assets: a foreign currency and a foreign equity – an action nearly equivalent to taking two dollars and investing the first in a hedged foreign equity position and the other in just the foreign currency. In a very naïve sense, an unhedged position more or less levers our portfolio by 100% and invests in two asset classes.That’s the magic of currencies: you can retain the exposure while still invested in a real asset.So unless the currency fluctuations and the foreign equity are negatively correlated, your portfolio volatility is going to be higher than a hedged position.This happens for all of our U.S. dollar denominated investments as well. The reality is, however, we do most of our spending in dollars and our rate of inflation has been so low and steady, we barely notice it.Which highlights an important point: as valuable as investing abroad is to hedging valuation and fundamental risks, foreign currencies help us hedge localized risks like inflation. And if we are going to hold the currency to help us hedge that risk, we might as well do something with it: like invest.
In his August 2014 white paper, Don’t Layer Currency Risk on Top of Equity Risk, Jeremy writes that investors who take on FX risk “must also recognize they are implicitly expressing a bearish view of the U.S. dollar.”Breaking down foreign returns into “nominal share price changes” + “currency fluctuations,” our return is certainly maximized with a bearish dollar move.However, holding this position does not mean we have to have a bearish disposition.Consider that when we invest in a foreign economy through broad market exposure, we are investing in assets with real earning power that we expect to grow over time. This growth may be realized in nominal share price increases, or it may be realized in an appreciating foreign currency. How the split will actually occur is difficult to predict.This is the case we are seeing in the U.S. right now: a stronger economic positioning has led to a rally in the U.S. dollar. For a foreign investor, much of the U.S. economic strength is being realized in an appreciating U.S. dollar.
3. Over long periods of time, currency fluctuations are not independent of stock returns. Some of the best times to buy foreign equities were when their currency was depreciating.
This is an interesting point that certainly warrants more exploration on my part.
A Counter-Point for Jeremy
Let’s consider a very simplified example of a European company that exports 100% of its goods to the U.S.If the dollar appreciates by 100% against the euro, the company has effectively doubled its revenue and. For simplicity, we’ll assume this translates into a 100% increase in nominal share value. The company is implicitly 100% long the dollar.As a U.S. investor, if we’ve invested in this company in an unhedged manner, we end up with a 0% return – the increased share value cancels out with the depreciating euro exposure.If we’ve hedged, then we end up with a 100% return.But the nominal share price return was due entirely to currency fluctuations.
Ironically, by hedging, we’ve actually exposed ourselves to fluctuations in share prices due to currency fluctuations.
So if want to diversify our fundamental risk, shouldn’t we choose to only expose ourselves to nominal share prices due to things other than currency? In which case, shouldn’t we remain unhedged?
One of the wonderful things about being an ETF strategist is the ever-expanding palette of investible exposures. Currency-hedged ETFs are one of the newest additions and have not only provided a way for investors the ability to express their views about currency fluctuations, but have also raised awareness to investors about the role that currency plays in returns.While Iremain unswayed in my initial viewpoint, Jeremy’s email and writing this commentary sent me down a rabbit hole of research that I am grateful for. Unjustified assumptions are a dangerous thing to have in the investment world.
I know Jeremy isn’t done with me – at the end of his email, he mentioned an asset/liability matching argument. I look forward to being proven wrong.
In Our Models
There were no changes in our models this week.The momentum picture remains mixed from both a domestic and global perspective. Globally, six sectors remain positive, two are neutral, and three are negative. Domestically, seven sectors are positive and two are negative.Recent downside pressure has inched the portfolios closer to building a cash position. However, we can see that the internal strength of each sector is vastly different. For the U.S. sectors, we estimate that enough sectors would turn off after another 5-7% drop in markets to warrant a short-term U.S. Treasury position – but the portfolio wouldn’t go fully to cash until a 15-17% drop.
Globally, we see a bit of a steeper move to cash. We would expect an initial introduction of cash after another 7.5% sell-off and rapidly increase the cash position up to 100% at a 12.5% sell-off. This is indicative that internally, the momentum exhibited by the positive global sectors is much weaker than their domestic counter-parts.