There’s no doubt that Turkey has dominated the headlines in recent weeks. And, title of this blog notwithstanding, we don’t make light of the topic. It’s serious for a country and its populace when its economy gets into the type of precarious situation in which Turkey currently finds itself, and to the extent that market contagion can sometimes spread to other countries, we are certainly keeping a close eye on the situation. However, it’s also important to be able to discern between the extent of a sell-off, and its potential impact on U.S. investors.
In the case of Turkey, it’s key to remember that, for U.S. investors who access their emerging market (EM) exposures through a broad basket, the country is likely to account for only a small proportion of their investment. Indeed, its weight in the MSCI Emerging Market index is only around 50 basis points currently, making it almost an afterthought compared to the top ten countries that account for around 90% of the index (the smallest of which is Malaysia at around 2.33% of the index) as of 8/27/18.
Of course, if a U.S. investor accesses the emerging markets on an unhedged basis, then they are long not only the stocks of these countries, but also the currencies. And, if it has been a rough year so far for emerging market equities, it has also been pretty bad for their currencies too.
Figure One shows the details, listing all the countries and currency hedges, and their weights, held by one of our currency hedged EM exchange traded funds (ETFs), (note, these hedge weights will differ slightly due to a monthly foreign exchange (FX) hedge frequency). Also shown are the year-to-date currency returns versus the U.S. dollar, and their “Impact.”
| Figure One: Countries and currency hedges and weights, held by one of our currency hedged EM ETFs (weights as at 08/23/18, returns versus the U.S. Dollar Year-to-date) | |||
| Currency | Weight | YTD Return | Impact |
| Hong Kong Dollar | 25.8% | -0.4% | -0.11% |
| Korean Won | 15.5% | -3.9% | -0.60% |
| Taiwanese Dollar | 13.2% | -3.3% | -0.43% |
| Indian Rupee | 10.0% | -8.8% | -0.89% |
| South African Rand | 6.7% | -13.5% | -0.91% |
| Brazilian Real | 6.5% | -19.2% | -1.25% |
| Russian Ruble | 3.6% | -14.4% | -0.52% |
| Mexican Peso | 3.4% | 4.2% | 0.14% |
| Malaysian Ringgit | 2.6% | -1.3% | -0.03% |
| Thai Baht | 2.6% | 0.0% | 0.00% |
| Indonesian Rupiah | 2.1% | -7.2% | -0.15% |
| Polish Zloty | 1.3% | -5.5% | -0.07% |
| Chilean Peso | 1.2% | -6.9% | -0.08% |
| Philippine Peso | 1.1% | -6.6% | -0.07% |
| Qatari Riyal | 1.0% | 0.0% | 0.00% |
| UAE Dirham | 0.7% | 0.0% | 0.00% |
| Turkish Lira | 0.7% | -38.8% | -0.26% |
| Colombian Peso | 0.5% | 0.9% | 0.00% |
| Chinese Renminbi | 0.4% | -4.3% | -0.02% |
| Greek Euro | 0.4% | -3.3% | -0.01% |
| Hungarian Forint | 0.3% | -7.3% | -0.02% |
| Czech Koruna | 0.2% | -4.1% | -0.01% |
| Egyptian Pound | 0.1% | -0.8% | 0.00% |
| Total | 100.00% | -5.31% | |
| Source: DWS FactSet, Bloomberg as of 8/23/18. Past performance may not be indicative of future results. | |||
What is this last metric, “Impact”? Well, it’s one that we find helpful in understanding what have been the main drivers of performance between currency hedged, and un-hedged performance year-to-date. Allow us to explain. So far in 2018, the unhedged emerging market index is down -7.73%. That represents the sell-off both in stocks and currencies that a U.S. investor could have been exposed to. However, they also had the option to hedge out that EM currency exposure, and, had they done so, they would be down by less, -2.95% so far over the same period (YTD).
That leaves a positive performance difference of 4.79% to the hedged investor that needs to be accounted for. Well the majority of it is composed of two things – the weighted currency return (shown in the table as “impact”), and the cost to hedge emerging market currencies (not shown, but currently around -1.40% per annum). Impact measures the year to date return of each currency, weighted by its size in the unhedged fund. So, it controls for the currencies where investors should care more. And, while it’s clear from the table that the Turkish lira is having the worst year, by far, of all the currencies in the fund in return terms, it should also be clear that it’s relatively small weight mutes the impact. One can see at a glance that, despite being down less than the lira, it is in fact the Brazilian real, the South African rand, the Indian rupee, the Korean won, the Russian ruble, and the Taiwanese dollar that have all been more influential so far this year, because of their heavier weights.
In fact, we can be more specific, and conclude that, of the -5.31% return to the weighted currency basket, the Turkish lira only accounts for around 5% of this performance gap (ignoring cost to hedge, and the math is simply -0.26% as a percentage of -5.31%). The “top five” currencies by impact are the real, the rand, the rupee, the won, and the ruble. These five have accounted for almost 80% of the differential this year between hedged and unhedged performance. Indeed, it’s the Brazilian real has been most critical, alone accounting for around a quarter of the unhedged investor’s underperformance.
Finally, you may be wondering why, in eliminating exposure to a currency basket that is down -5.31% year-to-date, a hedged investor “only” has outperformance of 4.79%? Well, there will be other things going on – including transaction costs and tracking issues – but a large part of the difference also comes from the cost to hedge EM currencies. That cost is currently around -1.40% per annum to the U.S. investor, but is has come down over recent years as the U.S. Federal Reserve (the Fed) has put rates up and other countries (notably Brazil in fact) have lowered their interest rates (rate differentials are the main component of hedging costs).
So, amongst all the headlines, always ask yourself not just what an individual stock or currency’s absolute performance has been, but also its relative weight in the portfolio. The two together show the impact on a portfolio and, in the case of Turkey, it should be clear that the column inches are at odds with the bottom line.