- Comments (10)
- Text Size: Down Up
moneysense.ca, 29/01/08
Currency Effect on Foreign Equity Holdings
Many investors are interested in investing in currency neutral versions of foreign equity index funds given the recent rapid appreciation of the Canadian dollar. The Brandes Institute has researched the effects of currency on a portfolio and in a report titled Currency Hedging Programs: The Long-Term Perspective concludes that:
We believe that it’s appropriate for investors to choose either a hedged or an unhedged benchmark, and then stick with it for the long term (a 10-year horizon or longer).
You may also want to check out an earlier report published by the institute on the same topic that looked at the results of hedging for Canadian investors from the start of the floating exchange rate era to the end of 2005. The study found that the impact of hedging to be strongly cyclical and each cycle, on average lasting just under three years.
moneysense.ca, 29/01/08
I never thought about it before, but the “sticking to it” part is probably more important than which way you go. Just like selling stocks when they’re low and buying when they’re high again, switching to whatever looks best each year is probably the quickest way to ensure below-market returns no matter what currency you’re looking at.
>>”The study found that the impact of hedging to be strongly cyclical and each cycle, on average lasting just under three years.”
As I alluded to in yesterday’s post, their “three year cycles” are simply noise. They have a totally radical way of defining what a cycle is: “three successive quarterly numbers of the opposite sign”. Huh?
The chart they provided in the report isn’t a currency exchange chart. It is their own one-year rolling return chart defined as “annualized returns for a series of overlapping, smaller time periods within a single, longer-term period.” In this case, the short period is a quarter, and the long period is one year. So, all exchange rates you paid prior to 12 months ago are erased from their equation. For example, if the currency is down 50% over the past year and up 5% over the next 3 quarters, it’s a cycle.
I kept scratching my head wondering what’s the rationale for the new definition. It seems so arbitrary. Why not 4 successive quarters? Or 5, or 6?
Anyway, they can define a cycle however way they wish, but most of us would agree that a cycle is from peak to peak. I’ll re-post the Canadian/US currency chart below. Hopefully it’s obvious that a currency cycle (in the traditional definition) lasts about 15 years.
http://research.stlouisfed.org/fred2/data/EXCAUS_Max_630_378.png
It’s also interesting that they mentioned this in their conclusion:
“Currency moves and the related hedging impact tended not to wash-out completely over time, and even for 5- or 10-year periods, the range of results remained wide.”
FJ: The study looked at the advantage of hedging over an equivalent unhedged portfolio (positive if hedging wins and negative if the unhedged portfolio wins). Check out the first chart in the second report – it clearly suggests that the impact is cyclical.
The second comment you point to is from Exhibit 1 of the first report. It does shows that currency effects (positive or negative) are significant over a 10-year period. Look at the median effect in that chart – the ten-year median effect is close to zero. Compare this chart to that of returns of equities over 1-year, 5-year and 10-year periods, the median of which is strongly positive (not close to zero) and the lowest bound also turns positive as the holding period increases.
I did look at your chart – firstly, I’m not looking to invest in one currency but interested in finding its effect on a portfolio of stocks. Second, year-to-year fluctuations matter a great deal because it influences where new money is added.
SP: I totally agree with you that there is a large element of performance chasing in switching to a hedged portfolio after a rapid appreciation in the dollar against other currencies. A few years back, it was the other way around: investors were clamouring for RRSP foreign content restrictions to be removed because of the then recent outperformance of foreign equities boosted by currency gains. It’s not a new phenomenon that investors are always fighting the last war.
The wording in the reports is such that somehow an unhedged portfolio is the benchmark to beat, and whether a hedged portfolio will outperform.
However, Most people looking to hedge aren’t doing it to gain an advantage, but they simply want to remove currency bets off their portfolio. They don’t care if they outperform or not. To them, a hedged portfolio *is* the benchmark, and leaving currency open subject their portfolio to currency risk which they don’t want.
>>”The study looked at the advantage of hedging over an equivalent unhedged portfolio (positive if hedging wins and negative if the unhedged portfolio wins). ”
That doesn’t answer my post about noise. I liken this sort of measurement to technical analysis in stock investing, too fixated in short-term, small impact and random movements. If you sit down and work out the different scenarios (I have), you’ll discover many flaws in their definition. The report is correct based on a “cycle” definition that they invented. Most rational people would argue that a cycle is a path where it ends at the same point as it began. There’s no need to complicate things. If we must redefine what a cycle is, we have to apply it to the stock market as well. Apple-to-apple comparison, right? And I’m sure the random noises will shorten stock market “cycles” significantly down from 7 or 8 years.
>>”Look at the median effect in that chart – the ten-year median effect is close to zero.”
I’m not surprised, but risk isn’t about where the median is. It’s about probability of deviating from the median or mean.
>>”Compare this chart to that of returns of equities over 1-year, 5-year and 10-year periods, the median of which is strongly positive (not close to zero) and the lowest bound also turns positive as the holding period increases.”
The expected return of the stock market is ~10%, therefore the median should be positive. 10% is the baseline for stock market.
>>”I’m not looking to invest in one currency but interested in finding its effect on a portfolio of stocks.”
Fair enough. We should invest in a collection of currencies if we intent to travel abroad or if we import a lot of foreign goods. That’s a good reason to buy foreign currencies, but too often, people claim that currency exchange is a wash over the long-term without realizing that long-term is beyond 10-years, and there’s enough volatility in currency exchange to tower over the mere 15 basis point saving.
>>”Second, year-to-year fluctuations matter a great deal because it influences where new money is added.”
These 3-year cycles are insignificant mini-cycles/noises sitting inside a monsterous super cycle. You can attempt to “time” the bottom in these mini-cycles, but in the grand scheme of things, the super cycle will render the exercise futile.
FJ: If an investor needs money in less than 10 years, why would he put it 100% in foreign stocks in the first place? Even if he decides he wants a portion in equities, he’d first have to decide how much of the equity portion should be in foreign stocks. Only now, does the question of taking on currency risk arises.
Taking on currency risk isn’t a bad idea: Link
You’re right, of course, that while the median might be near zero, currency could have a negative impact of as much as 10% over 10 years. But that’s only true if I invest solely in US stocks and never add to it or rebalance over the worst ten-year period. I’ll wager that not many investors are like that (edit: or at least, should not invest like that).
Even if I agree that hedging is beneficial, there are significant costs attached to it. The 0.15% extra charge is the cost of managing the hedge, the actual costs of the hedge shows up in the tracking error, which is significant for currency-neutral funds. Any potential benefit you may obtain with hedging could be lost in the cost of the making the hedge.
(Sorry for the numerous typos in my previous post. I rushed it before leaving the house.)
Anyway, the bottomline is that it’s a good idea to adjust your foreign currency exposure accordingly depending on how much foreign purchasing power you want to protect.
If for some reasons, someone invests 60% of his overall portfolio outside of Canada because “Canada only represents 3% of the world economy”, he has to figure out if the 15 basis points risk-premium is enough to justify taking on currency exchange risk, which in my opinion is far riskier than the stock market.
[...] many US and international equity funds are now available in currency-neutral flavours. There are two schools of thought on currency hedging: one holds that currency fluctuations “cancel out” for a long-term investor and the [...]
[...] Currency Effect on Foreign Equity Holdings [...]