Or, for that matter, Icelandic krona and Turkish lira …
Borrowing at 1%, maybe less, in a depreciating currency, and buying an appreciating currency that pays 12.4% or so is pretty good business. The Turkish lira pays more 20% , as does the Icelandic krona (14.25% -- Kaupthing Bank has a nice description of the yen funded krona carry trade). But they are a bit more volatile – for a comparison of BRL and YTL volatility last year, see the last two charts of this presentation. And since last July, it is has been really good business.
The 2 and 20 crowd doesn’t even need to use that much leverage on yen funded carry trades in the Brazilian real, Turkish lira or Icelandic krona to maintain their position at the top of the income distribution. That kind of spread works with real money. Or if you finance it with low-yielding Latin currencies rather than yen ...
My colleague Victoria Saddi has noted that Brazil seems to have attracted $10b in portfolio inflows in January alone, judging from the balance of payments data. It attracted $14b in all of 2006 (There was a little sell off in May and June that slowed things down). Brazil’s reserves increased by over $5b in January. That is a $60b annual pace. $60b is huge. It is about twice what the IMF provided back in 2002 - funds that in all probability kept Brazil from a messy default.
Moreover, the pace of reserve growth seems to be accelerating. There are press reports (in Portugese) that Brazil bought $2.5b in three days at the end of last week. That is real money – almost China style money. China has to buy about $1b a business day to keep the RMB down. In the past couple of days, Bloomberg reports that Brazil's central bank bought around $0.5b a day. That isn't quite China's pace, but relative to Brazil's economy, it is a lot. The central bank's data, which lags just a bit, shows that Brazil's reserves are up $2b in the first few days of February.
No wonder State Street’s positioning data shows that an exceptionally large number of investors are betting on the Brazilian real.
Brazil’s government is understandably concerned that this inflow will push the real up too much, undermining Brazil’s competitiveness. Sure, Brazil sells a lot of soy and a lot of iron ore. But Brazil also makes goods that compete with Chinese goods – and, for that matter, Argentine goods. And while high inflation is pushing up the real value of the Argentine peso (despite nominal exchange rate stability), the same cannot really be said of China. The real has been appreciating v. the yuan.
But with market pressure for the real to move up, that means that the central bank is constantly buying dollars to keep the real from rising more. In the process, it removes a lot of volatility from the market. And then I suspect of a bit of reflexivity (to use a Soros term) kicks in. Lower volatility makes carry trades more attractive. Existing carry traders can gear up more without getting into trouble with their risk manager. Others get pulled in. That only adds to the pressure on the real. And that only forces the central bank to intervene even more.
I did some work on Turkey a while back. This is exactly what happened with the Turkish lira in 2005 and early 2006. There was an extended period in 2005 and early 2006 when there was less volatility in the Turkish lira/dollar than in the euro/ dollar. That was the period when Turkish reserves were rising strongly. That, of course, changed in May 2006 ….
Brazil looks to be in a similar boat. So long as the money is flowing in and the central bank is resisting appreciation, there won’t be much volatility. And without much volatility, there is a big incentive to bring money in …
The Brazilian carry trade was popular a year ago (Felix Salmon has the details). it seems to be even more popular now. No wonder there is now talk of doing a reverse currency swaps action to try to stem some of the pressure [Edit: the previous sentence should read "stepping up its use of reverse currency swaps," as Brazil has intervened in the future market for some time -- thanks to OC in the comments. It seems like the Bloomberg story was a reference to an ongoing debate in Brazil over whether to do reverse currency swap auction on Tuesday to roll over a maturing swap.]. Brazil’s central bank sold a fair amount of insurance against the depreciation of the real back when that was a risk (in 1998, in 2002), helping to support the real in the spot market. It now seems to be considering doing more or less the opposite …
But capital inflows of 10% of GDP are hard for most emerging economies to handle. And $10b a month works out to a pace well above 10% of GDP for an economy of Brazil’s size. Outflows of 10% of GDP caused the Argentine crisis of 2001 and the Turkish crisis of 2001. Inflows of that side are a bit easier to manage, but only a bit.
The US has a current account deficit of around 7% of GDP, so it needs that kind of inflows (though it would rather not pay Brazilian interest rates to get the money). But for now, Brazil has a small current account surplus – so it doesn’t need the money.
Remember, while investors are borrowing in low-yielding yen to buy high-yielding real, the Brazilians – whether the central bank or the finance ministry – are selling high yielding real debt to buy low-yielding dollars. If every yen that comes into Brazil is bought by the central bank and used to buy dollars, well, the net result is a yen funded purchase of US treasuries. That is the net global flow. The 2 and 20 crowd gets the interest rate on the real and the yen/ real risk. The Brazilian central bank pays the interest rate in real and gets the interest rate on dollars and the dollar/ real risk.
And whenever someone feels forced to take a losing hand, they just might consider changing the rules of the game. I don’t quite see how Brazil can sustain this level of intervention.
The PBoC has positive carry and massive exchange rate risk. The BCB has negative carry and still quite a bit of exchange rate risk.