If you read the financial press closely, it was hard not to learn a bit about the Icelandic krona over the past week. The krona pays a high interest rate, you see - enabling Iceland to attract the funds needed to cover its (significant) current account deficit. It is a mini-New Zealand, in other words. Everbank was offering Icelandic krona denominated accounts.
It is good to get into carry trades early - you get the high interest rate and watch as other investors bid up the currency too. But it is not so good to get in late - falls in the currency can offset the gains from higher interest rates.
And I guess some folks started to get a bit nervous. Fitch downgraded Iceland, the krona fell sharply, and downward movement in the krona led to downward moves in Brazil, South Africa, Indonesia, Poland, Mexico and Turkey.
Why, you might ask, does what happens in Iceland matter to Brazil? Or Turkey?
Their economies are not exactly any more similar than their cultures. The answer is that all are linked together by common set of carry-driven investors. Tony Northfield of ABN Amro:
"These countries are unrelated geographically, but they are not unrelated in portfolios."
The FT described last weeks dynamics in terms anyone who still remembers 1997 or 1998 would easily understand:
"For traders who has borrowed in euros, the krona's sharp slide wiped out more than a year's worth of carry trade profits in one fell swoop. Those very same speculators, many of them hedge funds, will have built long positions in a plethora [great word] of other high-yielding currencies as well. Hence, selling mushroomed, as positions in one country were closed to fund losses in another, prompting further losses."
No one wants to buy a high yielding currency just before it starts to fall. The goal always is to get out before the other guy ...
Does this have any broader implications? Iceland, after all, isn't the only country in the world with an unsustainable current account. And it at least offers investors willing to finance its deficit a nice 10.75% coupon. Another big borrower only pays 4.5% or so ...
Flat yield curves in the US (and rather slim credit spreads) make simple carry trades in the US hard. Borrowing short to lend long doesn't work. Borrowing short to take on long-term credit risk works - but it doesn't pay as much as it used to.
For investors who do not want to make fancy bets on correlation, my sense is that there are two kinds of carry trades that still work:
Borrowing in dollars (or euro) and buying the currencies of countries with higher interest rates than the US - whether Icelandic krona, New Zealand dollars, Brazilian real, Turkish lira or another currency.
Borrowing in yen and buying just about anything ... remember, yen interest rates are low. This pays off if the yen doesn't rise in value. And pays off big if the yen falls in value.
There is another trade that has worked: Selling insurance against bad outcomes. If nothing bad happens, you collect the premium. And recently, nothing bad has happened. Those selling insurance have made money; those buying, not so much.
Such trades are usually done with derivatives, but selling insurance against bad outcomes -as I understand it -- has certain similarities with a simple carry trade. Borrowing dollars to buy Brazilian real pays off so long as nothing bad happens to the Brazilian real. Borrowing yen to buy dollars pays off so long as nothing bad happens to the dollar. And selling insurance against big market moves pays off so long as nothing bad happens in the major markets.
Why does any of this matter to those who do not have money on the line?
Well, big bets gone awry are a potential source of systemic risk. Ragu Rajan of the IMF has warned that investors are forced to take on ever-increasing amounts of tail risk (the risk of really bad outcomes) to eke out the high returns needed to justify their management fees. Tim Geithner has warned, consistently, against betting too heavily that a world with massive imbalances will be a very financially stable world. Those betting on continued stability are betting either that imbalances won't unwind, or that they will unwind in ways that don't lead to major turbulence in financial markets.
Once certain kinds of bets start to go bad, it can go really bad really fast - everyone wants to get out (sell) before the other guy. Particularly those who are playing with borrowed money. Read the counterparty risk management group's report. Or just remember 1998. The sudden unwinding of yen carry trades back in 1998 led the yen to move by 20% against the dollar in a very short time (see the chart here).
Last week's reverberations from krona move suggest that financial contagion hasn't entirely disappeared. The fact that Iceland set off tremors suggests some people are nervous. No one wants to be the last guy out.
But the world has changed since 1998, at least in some ways.
The unwinding of some carry trades will not necessarily have a big impact on all emerging economies. Some emerging economies have "banked" carry-trade related inflows, and used the inflows to build up their reserves. A fall-off in inflows means a welcome fall-off in reserve accumulation, not a crisis. Financing the purchase of Treasuries with the sale of higher yielding debt (another way of describing the process of sterilizing carry-trade related inflows) is a good way for an emerging market central bank to lose money.
Last week I estimated that a flurry of beginning of the year flows led emerging economies to increase their reserves by $60b (including all Saudi foreign assets) in January. It looks like that estimate was too low. My estimate assumed China added $20b to its reserves (not counting valuation gains); unverified leaks suggest China bought almost $35b in January. That would push emerging market reserve accumulation in January up toward $75 b a month - or a stupefying $900 b a year.
Smaller inflows to emerging economies - whether from investors plying the carry trade or investment managers looking for another year of big gains in emerging market equity markets -- would not be a great tragedy for many emerging economies.
The yen carry trade, though, is an altogether different beast. As William Pesek has noted, no one has a good sense of the scale of the yen/dollar carry trade. But no doubt this trade is one vector - along with Japanese real money investors looking for yield -- that has helped to bring Japan's current account surplus to the US after the MoF and BoJ stopped intervening in the spring of 2004. And if investors lost interest in that carry trade in a big way, watch out.
Fitch's downgrade of Iceland clearly didn't trigger a stampede out of the biggest carry trade of them all.
But I still have a sneaky suspicion that the flat yield curve is prompting someone somewhere to take too much risk. And a suspicion that an extended period of stability is prompting folks to sell to much insurance against a return of (financial) instability.
And a suspicion that the next crisis may not play out quite in the same way as past crises. A flight to quality historically has meant a flight into the US dollar and US treasuries. But can the currency of a country with a $1 trillion external deficit really offer a safe haven in an uncertain world?