Dollar, funding currency for the global carry trade?
With the Fed now aggressively cutting US interest rates to keep the bursting of a real estate bubble from devastating the US financial system and economy, is the dollar about to turn into the yen- that is, the key funding currency for the global carry trade?
Mark Gongloff of the Wall Street Journal — citing a set of currency traders — posed the question yesterday:
Japan’s yen has been the carry-trade vehicle of choice for years, given the country’s superlow interest rates. The Swiss franc has been another. If the Fed keeps cutting rates, carry-traders might line up to ride the dollar like a birthday pony, with important implications for markets and the economy.
In my view, it will be hard for the dollar to emerge as major funding currency. At least not in the absence of truly unprecedented generosity from the world’s central banks.
Why?
Fundamentally, because the US has a huge external deficit and needs to borrow money from the rest of the world even in a slump, while Japan’s slump added to Japan’s external surplus and thus the supply of funds it could lend to the world.
Funding currencies usually have — in addition to low interest rates - lots of savings, and consequently spare funds to invest globally. That doesn’t sound much like the US.
The United States own need for funding limits the dollar’s ability to serve as a global funding currency.The US saves far less than it invests, and has to make up the difference by borrowing from the rest of the world. That is a constraint on the United States’ capacity to supply dollars to investors who want to sell (borrowed) dollars to buy higher yielding currencies.
Or, to put it a bit differently, the only way the US dollar can be a funding currency in a deep macroeconomic sense is if more funds are coming to the US - even with lower interest rates in the US than in the rest of the world — than are needed to finance the US current account deficit. Those investors borrowing dollars to invest outside the US wouldn’t be borrowing US savings; they would be borrowing the rest of the world’s savings that happens to be held in dollars.
And who might be the source of all those dollars?
Right now, there is only one realistic answer: the world’s central banks. I rather doubt that they are willing to finance both the US current account deficit and a large US dollar carry trade.
I have hardly been a cheerleader for sovereign wealth funds. In my view, they raise at least as many problems as they solve: I would be much more comfortable if sovereign investors stuck to index funds for their equity exposure. But the whole point of sovereign wealth funds is to avoid a world where private investors use dollar supplied by central banks to fund lucrative investments outside the US. Sovereign funds let governments make such investments on their own.
It is hard right now to say anything is impossible. Five years ago I never would have guessed that the emerging world’s governments would add over a trillion dollars to their external assets in a year, or that the US government would tacitly encourage US banks to be recapitalized by the investment funds of cash-rich (and less-than-democratic) emerging market governments.
But a substantial US dollar carry trade would fly in the face of the United States need to attract large sums of savings from the rest of the world to "fund" its current account deficit.
That of course doesn’t mean that some US investors haven’t sold the dollar to buy foreign assets, or that some foreign investors haven’t borrowed in dollars to invest abroad. Recent market moves suggest that "deleveraging" supports the dollar as well as the yen.
But there is an important difference between post-bubble Japan and post-bubble America: Japan always ran a current account surplus and never needed financing from the rest of the world.
As US rates fall, though, the US may find it hard to be attract carry trade flows from places like Japan. The dollar may no longer be a destination currency, even if it isnt quite a funding currency. The absence of a yen/ dollar interest rate differential has in the past been associated with Japanese intervention in the currency market, as the Japanese government picks up the slack from the carry trade.
Something to watch.
You are right about the lack of savings in the USA. However, if there is a global crash, will flight to quality make a lot of people park their capital in dollars? Others will then borrow these dollars to invest outside.
Would the carry trade need to deal directly with US institutions? I mean, there’s tons of USD around and about, In China, mid east, etc…..?
Anonymous,
Flight to quality will make people to park their money in dollars???
I which century are you living?
Brad,
If others don’t want put money into the US, wouldn’t that mean that interest rates would go up? Greenspan claimed he was not able to raise the long term rates, despite he tried (tho I don’t believe him), but wouldn’t that mean, that a reverse process, money fleeing the US would lead to high long term interest rates regardless of what the Fed tries to do?
This reminds one of detailed discussions on the yen carry trade some time ago.
Issues:
a) Definition of the carry trade at the macro level (e.g. leveraged money and/or real money). One definition might include the net foreign asset position (e.g. Japan’s NFA as a natural macro component of the yen carry trade)
b) The degree to which the NFA is a facilitator (Japan) or inhibitor (US) to a macro level carry trade.
c) The most contentious issue - the size of the carry trade (e.g. Japan), often hopelessly confused with intertwining debate about the definition.
I have no answers. But I know the size of the yen carry trade was a most contentious issue when it was a popular topic of debate.
I think Brad has it right about the reason the USD cannot become a weak funding currency. There’s too much demand for it because of the debts we owe in USD.
OT: I’ve figured out what I think about Brad’s view of China’s monetary hoarding, so I’ll say it now. I agree that US’s willingness to borrow and consume was necessary to China’s breakneck economic development. So was China’s willingness to lend and to work.
Now the US may have some problems, and indeed the problems would be smaller or different if China had not been so willing to lend. The blame is properly shared, if blaming is what we want to do. But blaming is unproductive, here we are, and hard work and thrift from the Chinese are showing good results vs. US consumption mania. Power is shifting because of it. Blaming is pointless.
Anyway, if the US can’t find a way to be thrifty in a world of temptation, we should blame ourselves first.
Brad, a couple of thoughts:
1) The absence of net private sector inflows into the US over the past year or two suggests that the dollar has already been a funding currency; while BRIC, GCC etc central banks and SWFs have been buying dollars, pvt sector traders and fund managers have been selling dollars for investment into…well..just about everything else.
That the dollar roared higher against most currencies when the poo first hit the fan in August is a testament to this.
2) More tangibly, the dollar has indeed now entered that most generic of carry strategies, the G10 “three up, three down” strategy, wherein practioners buy the three highest yielding G10 currencies and sell the three lowest yielding. Moreover, many risky asset longs have previously hedged their equity/credit/em holdings by selling USD/JPY.
These two facts perhaps explain the relative reluctance of the dollar to decline against the yen (i.e., it didn;t go down as much as one might have thought) during much of the recent market fireworks; the market’s already short (e.g., funding out of) dollars.
FWIW, my personalo bet is that the Japanese cannot really intervene in USD/JPY this side of 100, given a) their reluctance to intervene above 120, b) the still elevated level of EUR/JPY, and c) their signing a G7 communique calling on China to quit taking the piss.
Unlike 2004, there is no daiko henjoinflow to offset this time around, so the MOF seems to be taking a justifiably relaxed attitude towards recent movements.
Macroman — interesting thoughts, as always.
on 1)a lot depends on the scale of BRIC/ GCC/ SWF inflows into dollars. Say total official asset accumulation is @ $1400b in 2007. a 60% flow into $ = $840b — enough to finance some outflows, but not on a very large scale. Say it is more like 70% — ($980b). that would significantly over finance the US current account deficit ($750b, ballpark), and there would be some scope for a central bank dollar funded carry trade, or, to put a less sophisticated spin on it, a net capital outflow from american residents. Tis hard tho to argue this started before global reserve growth gapped up to $300b a quarter (give or take) in q1 07.
if there is a bit shift to SWFs and if SWFs have a lower $ allocation than Cbanks — the same sovereign asset growth would imply smaller dollar inflows, and less capacity to “fund” dollar carry trades by borrowing from the GCC/ BRIC central banks. i.e. you can tell a “big” SWF story supporting risk assets over the next few years or you can tell a dollar carry trade story, but i don’t see how the math works for both barring another uptick in sovereign asset growth.
on 2 — i was wondering about that. should have checked a bloomberg before posting. my sense is that if the scale of this gets too big, then you really do enter into $ meltdown kind of scenarios. A funding currency that already has a large current account deficit is in an interesting position.
I basically agree with your story on positioning and how it has impacted the $/ yen during the market fireworks. and i agree that us investors are now more long foreign assets than they have been, though whether this is best described as a dollar funded leveraged carry strategy (financed by central banks) or an asset swap (inflows into the US from places like europe fund outflows) is another debate — probably one that is similar to the debate over whether real money outflows from Japan should counted as part of the yen carry trade.
“that most generic of carry strategies, the G10 ‘three up, three down’ strategy, wherein practioners buy the three highest yielding G10 currencies and sell the three lowest”
is this done wrt real rates?
Brad, cannot help but join this discussion… You know I have a weak spot for carry trades… I’m afraid I do not agree with your premise that a C/A deficit country is not a good source of funding carry just because it requires a certain amount of funding itself… Japan has been a source of carry funding in spite, not because of its persistent C/A surplus. The latter is a hurdle, a structural headwind, if you will, for funding carry. US, on the other hand, finds itself in a situation where everything points towards carry funding: structural deficits are dollar-negative, private sector demand for dollars is low, CBs and SWFs may be cooperating, but they are far from eager buyers, the economy is slowing, interest rates are coming down and, as Macro Man pointed out, the buck is now in the bottom 3 carry basket… Why wouldn’t you fund in dollars? The question, rather, is to find a relatively safe higher-yielding currency, which is more or less structurally sound…I would be interested in Macro Man’s suggestions on this.
Andrew — I think the real question is not whether you would want to fund in dollars, but whether it is possible to fund in dollars in some deep macro sense. Funding in dollars means moving money out of the US at a time when the US still has a huge funding need for its current account deficit. It is a wrong way flow in that sense. or to put it differently, for every dollar funded carry trade moving money out of the US, in equilibrium there has to be an equal and offsetting inflow — and that inflow is on top of the inflow needed to sustain the us deficit.
the global net dollar long position has to increase by $750b a year to cover the US current account deficit. to cover a US CAD and a carry related outflow, that net dollar long position has to be much bigger — and get bigger at a time when us rates are low so there isn’t much obvious incentive to hold dollars.
the other argument against the $ as a funding currency is selling $ to buy higher yielding GBP or EUR or AUD risks coming to the party a bit late.
I recall asking about this some time ago and got detailed replies from both Brad and Macro Man, yet I still don’t really understand. If I want to borrow dollars in order to fund another leg of a carry trade, why do those dollars have to exist beforehand? Cannot the credit simply be created by a bank? Thanks in advance for your patience (again). I don’t see why there is a requirement for excess savings in order for dollars to be lent.
I think you may underestimate the structural savings in US, i.e. the wealth inherent in the land and other US assets. Once Americans decide they can get better returns overseas, a lot of that illiquid wealth will fund the dollar carry trade. That will simply require expansion of debt against these assets, not new overseas funding. Can’t say I understand the process entirely, but the shadow banking system is more than capable of creating such credit out of thin air.
“the other argument against the $ as a funding currency is selling $ to buy higher yielding GBP or EUR or AUD risks coming to the party a bit late. ”
I don’t buy this either. With negative real rates, it would be used to fund commodities trades as inflation hedges (gold, energy, food etc). I believe this has already started.
Brad, agreed on the high-yielding target currencies. I also understand your logic with regard to macro flows, but I’m wondering if it’s not missing the point in the same way as in our previous discussions some commentators focused on physical cross-border lending, without taking into account massive carry trading in the forward and swap markets… Granted, net foreign flows are vastly important, but their effect is arguably felt much more in long duration assets, sometimes even overriding the Fed (hence the infamous “conundrum”). But with regard to carry trading, it is the nominal short-term interest rates that matter, and here the Fed is much more in the driving seat. I suspect that as long as the Fed is happy to keep interest rates low and continue pumping liquidity into the banking system, there will be plenty of short-term dollars available for funding of speculative leveraged trades - in other currencies, but also, lest we forget, in dollar-denominated longer duration assets! After all, that’s very much part of the Fed’s strategy of helping the banks recapitalize themselves via steeper yield curves… In other words, the Fed is intentionally supporting duration-based carry trades in dollars…Dollar-funded carry trades in the FX market might just be a natural, if not necessarily a hugely welcome, by-product…
bsetser on 2008-01-29 17:49:41
Every carry trade requires an inverse trade, ex the current account imbalance.
This is the case for both deficit and surplus nations.
The FED pegs interest rates and not the supply of dollars. So the idea that the supply of dollars is limited is false by definition. If there is demand for dollars (from qualified borrowers) the FED stands ready to lend at the Funds rate. There is no limitation of dollars for the carry trade.
anonymous — don’t think so.
take a surplus country, and take an on balance sheet simple carry trade.
the country has an excess of savings over investment.
a non-resident borrows $1 billion of yen — that shows up as a capital outflow in the bop (bank loan to non-resident = outflow in other investment). the nonresident then sells the yen for euros, pounds or dollars, setting up the carry trade — the nonresident owes the bank yen, but has say dollars.
the overall result is a capital outflow — a rise in japanese lending to the world.
i don’t see the inflow.
or take the mof. it borrows yen domestically, raising yen cash (no bop impact). it sells yen for dollars, increasing its foreign assets. that is a bop outflow.
what am i missing?
moving money out of a country (funding currency) generates a capital outflow
moving money into a currency (destination currency) produces an inflow.
it gets more complicated if you do things off balance sheet, but that seems to me to be the basics.
Brad, your argument (about low savings rate in the US constraining the ability of the dollar to serve as a carry-trade currency) does not make sense. You seem to have forgotten the “printing press”.
The constraining factor will only be that the resulting depreciation of the dollar, though splendid for those we will be early in the dollar-carry-trade game, will ultimately become unbearable to Bernanke or Triche or some other central banker who will change interest rates to stop the carry trade. Whethen we Americans save or not is irrelevant.
The inflow is from the Fed, which can mint unlimited dollars. Ie, it can use any asset to balance dollars on its books, accept it as TAF collateral, etc. The Fed can issue dollars against stale burritos it finds under its car seat, dreams, rights to the next Kevin Federline album, etc. Since there is not and cannot be any shortage of assets of this class (for one thing, K-Fed can agree to record an infinite number of albums), there can be no shortage of dollars. I believe this argument was first made by a Professor Bernanke, though his formulation was perhaps more memorable.
The same tools you use to model aggregate international flows work for flows between any set of distinct entities. If you treat the Fed as a country of its own, the math will come out right.
Of course, the Fed can also refuse to accept K-Fed as collateral. This will create a shortage of dollars (at least with respect to a given interest rate) and short rates will rise, terminating the carry trade.
The general calculation is that a currency A is a funding currency for a carry trade with currency B if (a) B offers a risk-free return above the funding rate of A, and (b) exchange rate futures across the term of the loan do not wipe out the gain. If both these criteria are met, it’s free money, assuming there is no counterparty risk in the loan or your forex hedges. Which might be a big assumption these days.
But this is a subjective calculation. It does not depend on balance-of-payment aggregates.
I think your balance-of-payment calculations are getting mixed up with a larger political point that you want to make, which is that if the dollar becomes perceived as a funding currency, yen style, it might make the BWII accumulators think twice about the craziness of the whole system. This point is certainly inarguable. Although as you point out, if global irrationality were some great obstacle, we would not have BWII in the first place…
Dang, moldbug, you write well. Almost well enough to pledge to back my currency with mold. I love the image of the fed accepting k-fed as collateral. Big B discounts K-fed, and all is well. The bigger point I wanted to make is indeed as you describe: the dollar can only be a “funding” currency if folks can borrow excess central bank dollars, which would indeed highlight the craziness of the system.
The debate here touches on two complicated points — the role of the central bank, namely is “savings” a constraint on any country than can print money, and derivatives, which allow “positions” to be established without any formal flows.
I have a bit more confidence in my views on the role of the role of the central bank than my views on derivatives, so let me start there. Now it is true that a modern central bank can issue cash v collateral, and pretty much any collateral it so chooses (tho it generally protects its balance sheet by discounting high quality stuff). lots of emerging markets issue local currency against us treasuries for example.
but that doesn’t mean that things like savings do not matter. just because the fed can discount anything doesn’t mean it will. just today a prominent economist noted that the fed’s actual balance sheet hasn’t expanded all that much recently — unlike say in 02/03. no free source of money there.
nor are “funding” currencies necessarily countries where the central bank just prints currency; the swiss have perhaps lost a bit of the reputation for financial severity over the years, but the swiss national bank hasn’t been printing swiss franc like crazy over the years either.
so let’s assume that the fed can create money — and that through the money multiplier, any supply of bank reserves leads to a larger supply of credit in the economy. but let’s also assume that the fed’s decisions about the right amount of high powered money to inject are not going to be determined by the carry trade. in other words, if more funds move outside the US, the fed won’t put more funds on deposit in us banks to allow them to lend more. decisions about money supply are going to be a function of other variables, including domestic conditions in the us.
moreover, if the fund prints money in unlimited quantities and allows unlimited credit expansion, that will eventually have an impact on prices. credit extended to sell dollars through the carry trade rather than to buy us goods will put pressure on the dollar, push up import prices and ultimately push up domestic prices. that creates some limits.
ultimately, then, if lending to the rest of the world requires saving — or really saving more than you invest, not a printing press. that is what the us lacks, and that is what japan and switzerland have.
the absence of a limit on the carry trade comes from the fact that if say a us investor sells dollar for real, brazil’s central bank buys dollar and sells real and effectively takes the opposite side. the inflow from brazil’s central bank back to the us as it builds up its reserves in turn keeps us rates low and prevents the emergence of a big net outflow from the us. us investors build up claims on brazil, brazil’s central bank builds up claims on the uS but there is no net outflow.
Bottom line — I don’t find this argument persuasive.
Unless you think printing money creates savings (with savings defined as the difference between what you make and what you spend), the lack of us savings is a constraint on the amount of “real” outflows from the us — at least in the absence of “real” inflows of savings from the rest of the world that can be intermediated in the us and lent out …
more in the next comment
the use of derivatives to set up a “carry” position takes me a bit out of my comfort zone; the real pros should feel free to correct me.
it is certainly true that it is possible to place big bets on currencies that have carry through various derivatives markets. and it is possible to take positions with a lot of embedded leverage — i.e. a small amount of committed funds allows the player to benefit in a big way from market moves. the constraint here is the willingness of someone to take the other side of the position.
i.e. if i want to sell dollars to buy euros forward, some else has to be willing to sell euros and buy dollars forward. and the “notional” is only a way of keeping track of the size of the bet — what really matters is my ability to settle up at the end.
so how does this all relate the carry trade and bop flows?
this is how i understand things, and again, it could be off. If i want to go long (bet on a rise) dollars v yen, someone else has to be willing to go short dollars (i.e. ebet on a fall) against the yen. or in the case of a high carry currency in the emerging world like brazil, my long real/ short dollars position (a bet that the real will raise, despite the depreciated priced into the forward market b.c of brazil’s higher carry) has to be offset by a short real/ long dollar position (i.e. a bet the real will fall ..).
in practice, those selling real to buy dollars often want to hedge their “off balance” sheet risk — and the easiest way to do so is to hold real on balance sheet. a bank for example might offset its forward sale of real for dollars by moving some dollars into real — and building up its real deposits. that limits its risk — it is short off balance sheet and long on balance sheet.
in this example, the need to hedge off balance sheet exposure creates the real flows — the actual movement of dollars into brazil (something that would be offset by central bank intervention). absent those flows, there would be less willingness to take the offsetting position.
you see this with china, where capital controls limit movement into the rmb. the forward market pricing therefore (NDFs) is not a reflection of interest rate differntials, unlike for the $/ real or $/ yen. the imf also did some nice work on this for korea — published i think in one of the regional outlooks for asia.
to make a long story shorter — yes, it is possible to go out in the market and take a big position without moving a lot of money around. you just agree to settle up at the end. that has no impact on the balance of payments, apart from any flows associated with settlement. but those taking the opposite side of the bet — the “short carry” position so to speak — often hedge their exposure through real money flows that do show up in the balance of payments.
you saw this with japan — big outflows from other investment. and i think you see it with brazil as well.
here though i am less than 100% sure i have all the details right.
i need an editor; both of my previous comments are too long.
Basically, there is nothing that Brazil’s CB can do that the Fed can’t do. Brazil’s CB supports the longer end of the curve by buying Treasury and agency bonds of multiyear maturity. It does this by (a) issuing reais in exchange for dollars and (b) exchanging the dollars for bonds. (a) is a no-op for the Fed, (b) is the same.
The sight of a government buying its own bonds is always unseemly. BWII puts a figleaf over this practice by threading it across borders and around pegs. The circuit is basically the same, however. I am not at all arguing that this is not (a) a pernicious practice, (b) both politically and economically dangerous and unstable, etc, etc.
There is a considerable tendency in 19C and 20C economics to consider money as a sort of symbolic proxy for “real” goods and services. Even the Austrians often fall into this pattern of thinking. By just considering money as a good in its own right and abandoning the superstructure of “purchasing power”, one can often shed a lot of mental baggage.
So either “savings” are dollars or they are something else. If they are dollars, the Fed can create them. If they are not dollars, what are they? A variable in a model, I suspect. Unless everyone on Wall Street is running this exact same model, I have trouble seeing how its variables can be assigned any causal significance. (For example, people often talk about an “inflation premium” on interest rates without any particular theory of how the CPI, a classic model variable, is supposed to affect the decision process of either borrowers or lenders of dollars.)
Imagine a hard gold standard in a world in which a single, ineffably gigantic, golden asteroid had crashed into the earth, creating one and only one effectively inexhaustible gold mine. The mine is in the Fed’s backyard. As a result, the Fed can lend any amount of gold, provide a risk-free guarantee to any gold security, etc, etc. As it can today. The analogy is useful because somehow our brains are more used to thinking of something as “real” if it hurts when you drop it on your toe. But dollars, even virtual electronic dollars, are every bit as much real goods as Krugerrands.
One problem with the world in which the Fed has an infinite gold mine in its backyard is that it might well end up on a silver standard. Monetary competition is only one of the many nasty consequences of indefinite currency expansion. (I really think John Stuart Mill’s Currency Juggle ought to be required reading for any economist today. Even better, when they fill out their union cards, they should have to indicate whether they agree with Mill or the Messrs. Attwood, or neither, and why.)
Also: for the reasons I discuss here, I think looking at the Fed’s balance sheet is not a very useful way to calculate the number of dollars in the world. For example, the Fed’s printing press confers a effectively risk-free status on FDIC-insured deposits and a nearly risk-free status on agency bonds. But there is no formal relationship between the Fed and the FDIC or the Fed and the GSEs, and certainly neither is anywhere near the Fed’s balance sheet. If a dollar is anything it is a Fed liability, and these off-balance-sheet liabilities, though informal - representing implicit obligations that are deduced as a consequence of political inference - must surely dwarf the formal balance sheet.
Response to bsetser on 2008-01-29 20:12:07
I said:
“Every carry trade requires an inverse trade, ex the current account imbalance.
This is the case for both deficit and surplus nations.”
Let me offer a proof of sorts that responds directly to your specific examples.
Example 1
Use your example. Obviously the surplus country is Japan.
Non-resident borrows $ 1 billion in yen from a Japanese bank, and sells yen for dollars, setting up a yen carry trade.
The non-resident yen liability is a capital outflow from Japan.
The counterparty in the FX trade has sold dollars for yen.
Either:
a) He borrows $ 1 billion in dollars and sells them for yen, or
b) If he doesn’t borrow the dollars, he already has the dollars as a balance sheet asset. His resulting balance sheet change at the margin is still short dollars, long yen (i.e. the marginal effect on his balance sheet position will be long yen funded by dollars.)
So the FX counterparty has the inverse yen carry.
He has a yen asset and dollar funding.
Moreover, the yen asset results in a capital inflow into Japan. This results either directly from the transaction, or indirectly through a Japanese bank, because all yen transactions must ultimately settle through a domestic Japanese bank. (A foreign bank in order to settle yen transactions must itself have its own bank account (i.e. a yen nostro account) with a Japanese bank in order to settle yen transactions. Otherwise, there can be no ultimate settlement of yen through the banking system.)
The sum of the primary and counterparty transactions is a 0 yen carry position globally.
Example 2
Use your second example. The domestic Japanese MOF transaction works the same way.
The MOF sells yen for dollars.
The FX counterparty is a non-resident in order for that to be a capital outflow.
The counterparty has sold dollars for yen.
This produces the same balance sheet effect on the counterparty as in the first example, with a resulting yen inflow into Japan.
The sum of the primary and counterparty transactions is a 0 yen carry position globally.
Example 3
This is the type that produces a true net yen carry trade on a global basis:
A Japanese exporter earns dollar profits, which he retains.
He keeps that amount of dollar cash in a Japanese bank account.
The Japanese bank keeps the equivalent dollar cash with its own domestic US dollar clearing bank (say Citi New York).
That is a dollar capital outflow from Japan.
(Or alternatively, e.g. the Japanese bank deploys that dollar cash in a dollar interbank deposit with a non-resident bank (which must also have a clearing account with a domestic US bank). That is also a dollar capital outflow from Japan. Dollars must ultimately clear to the domestic US.)
This is part of the true macro net yen carry trade.
The asset is a dollar.
There is no net liability. The true macro yen carry trade is supported by equity on the Japanese national balance sheet. The equity is represented here as equity on the balance sheet of the exporter, which is part of the national wealth. Japanese balance sheet equity is denominated in yen.
This is the only net yen carry trade globally. It is supported by the Japanese equity that is generated by the current account surplus and that is outstanding as reflected in and supporting the NFA position.
(The Japanese current account surplus, no matter how you slice it, is reflected in transactions that start with those such as this exporter example (or a similar investment income example). It is ultimately impossible to determine unambiguously where all the money ends up in terms of translating the current account surplus to Japanese assets (i.e. mapping net export revenue to a specific set of new foreign assets), but in this case, for convenience, we assume that the exporter has kept his profits in the bank. This is a fairly clear example of a specific foreign asset generated by the Japanese CA surplus, since the exporter’s retained earnings (which are national accounts savings) are unambiguously generated by that CA surplus.)
Other Comments
The Fed has nothing to do with this (directly). There are some aspects of banking that are not unique to central banks. This is all a function of global banking for any specific currency. The global banking system for any currency is closed. There is no leakage in term of accounting for assets, liabilities, and equity. The Fed or the Bank of Japan might be involved in some transaction at the margin, but this has no effect on the logic of what is described above.
Everything described above is in cash (as opposed to derivatives), and affects the balance of payments when it is a cross border transaction.
The use of derivatives cannot change the logic of what is described above. The creation of a derivative cannot cause the “disappearance” of cash as described above. Any cross border derivative has no effect on the global net carry trade at the margin - the purchase of a dollar/yen forward by one party is the inverse for the counterparty - consistent with the more general and relaxed observation that derivatives are a zero sum game per se. Portfolios everywhere may mix and match such cash and forward positions, but the net global carry trade is not affected by this.
Even a commercial bank that instantaneously creates its own desired dollar/yen FX exposure by simply ‘printing’ out the effect of such an exchange from ‘thin air’ (without any additional funding or conversion) has no effect on the global carry position. If the bank creates its own yen carry position by such a ‘transaction’, the depositing dollar customer who accepted the rate for his new yen deposit now has the inverse yen carry position (i.e. a dollar carry trade against yen).
Let’s undertake a small thought experiment. Imagine you are a currency trader, putting on a carry trade. The expected return and risk of this trade will be determined by the nominal short-term interest rate differential between the two currencies, which you lock in at the outset, and the expected direction and volatility of the currency pair during the life of the trade. Let us now consider how real-money macro flow-based liquidity and central bank generated liquidity affect our expectations with regard to this trade.
If you fund your carry trade in a country with a current account surplus (e.g. Japan), then unless there is an offsetting flow on the capital account (either by Japanese private investors, Japanese exporters choosing to keep their FX offshore or the MOF), you are facing a structural trend for JPY appreciation, which is a headwind for your yen-funded carry trade. But if the real-money outflows on the capital account are so large as to overwhelm the current account surplus (e.g. secular diversification by Japanese institutional investors and Mrs. Watanabe into foreign assets), then the net macro flow-based liquidity becomes a tailwind for your yen-funded carry trade position. In the US, we have the opposite: the current account deficit is a structural tailwind for USD-funded carry trades, whereas the real money flows on the capital account are the headwind, although not to the same degree as they used to be in the past…
Now let’s consider how you would set up a USD-funded carry trade. As discussed at length during the various postings on this topic in the past, the most common way to set up a leveraged, speculative carry trade is in the FX forward market. You go short the dollar, say, in 3-month forwards against a high-yielding currency. If there are too many sellers and not enough buyers of dollars in 3 months, the forward price of the dollar will go down, way out of whack with what it should be based on the interest rate differetials, thus creating an arbitrage opportunity for other market participants, who would put on a riskless trade, effectively borrowing and selling dollars spot and buying them back cheaply in 3 month forwards, bringing the relationship between spot and forward FX rates back in line. Put differently, this is how the selling pressure is transmitted from the 3 month forward market to the spot market. But more to our point, this is where derivative-based carry trades are linked with bank borrowing.
The dollar funded carry trade will be constrained by how easily and at what terms arbitrageurs will be able to access 3-month dollar borrowing in the interbank market. If this borrowing pushes 3-month interest rates too high, then this will be the end of the carry trade opportunity. But what determines these 3-month interest rates? Aren’t they anchored by the Fed’s overnight rate, which itself is fixed by the Fed at a certain level? In the end, who supplies this liquidity? If the Fed continues to peg the overnight interest rate at a relatively low level, and if banks are both able and willing to lend to each other in the 3-month interbank market, then this should offer plenty of dollar liquidity to fund carry trades, regardless of what is happening with the real money macro flows…
Am I missing something here? Please point out any holes in the above logic.
Anonymous — you are right, there has to be a net supply of dollars from exporters in my japan example for my story to work. I left out that flow — otherwise the dollar seller is taking a long yen position the financial market. And in the case of japan, that is what you see in the BoP — a current account surplus, offset in recent years not by a portfolio outlfow but by outflows through the banking system.
The argument though basically highlights the issue of who would serve as the counterparty to a dollar carry trade, that is take a long-dollar position. It cannot be us exporters in a macro sense, because the us doesn’t have a trade surplus. And remember that foreigners have to already build up their long-dollar position for the US to cover its current account deficit.
Andrew — I need to spend a bit more time working through your example, but the arbitrage opportunity that keeps the forward price at the interest rate differential requires capital flows, so it would be associated with a net inflow into the US (to fund the carry related outflow). I need to think of it more, but i am intuitively suspicious that it that easy to basically expand the assets and liabilities side of the US external balance sheet (with all short-term assets and short-term liabilities) symmetrically without in some way making it harder to sustain the net inflows needed for current account funding.
basically, your example assumes that instead of a $800b net inflow to cover a $800b current account deficit, the US gets a $1200b net inflow almost automatically to cover a $800b current account deficit and a $400b carry related outflow (long high carry currencies, but short-term)
Failed Monetary POlicy of the US Federal Reserve
http://www.atimes.com/atimes/Global_Economy/JA31Dj01.html
At the whim of trouble in the markets, Fed chairman Ben Bernanke has made it clear that he is inclined to flood the markets with liquidity at any cost. He said: “We stand ready to take substantive additional action as needed to support growth and to provide adequate insurance against downside risks.”
Contrast that with John-Claude Trichet’s comments: the head of the European Central Bank (ECB) recently said that during times of financial turmoil, it is imperative that inflationary expectations remain firmly anchored. The Fed’s increasing isolation is also apparent from recent comments by Bank of England governor Mervyn King, who said investors had been mispricing risk for far too long and that “the repricing of that risk … is not a process that we should try to reverse”.
Take the widely cited focus on “core” inflation that excludes food and energy. Such exclusions may have originally been justified: food and energy prices were notoriously volatile and may make monetary decision making more difficult. But when you have food and energy prices elevated for years, when you burn food to create energy, that justification no longer exists, and neglecting it in monetary decision making is, to put it mildly, inappropriate.
The Fed has wobbled into emergency mode, claiming to be vigilant on inflation while debasing the dollar in the process smells of hypocrisy. And because the US (consumers and the government) has been the world leader in borrowing, the traditional safe haven, the US dollar, is also under tremendous pressure.
Calvo sees stagflation in the future:
http://blogs.ft.com/wolfforum/2008/01/beyond-fiscal-s.html#comment-98954534
bsetser - the US counterpart to my examples on Japan is that there is no net dollar carry trade globally, because the US runs a current account deficit and has a net foreign liability position. The dollar therefore is a global target currency by default.
NFA countries like Japan by default are the hosts for carry currencies; NFL countries like the US are by default target currencies. All other transactions in either case net out globally, as I described. It doesn’t mean there aren’t all sorts of carry trades going on; it just means that somebody must be on the other side, unless the trade is backstopped by a supporting NFA position in a trade that is funded directly from the carry currency host country.
Andrew –
In your example, the interest rate/ forward arb hinges on access to the 3m interbank rate. so let’s assume that the desire to arbritrage a difference between the int. rate implied by the 3m forward and the interbank rate leads to more pressure on the 3m rate.
in the end, equilibrium would be restored by a higher 3m borrowing rate in USD, or less carry.
the pressure here reflects the absence of exporters with fx to sell for $. all those dollars are going to cover imports. instead, there are importers out in the market looking to buy fx to cover their costs .. (or exporters in the rest of the world looking to sell dollars to pay local expenses). the counterparty has to be financial.
now suppose that the fed doesn’t want higher 3m rates. It would have to ease — keeping its 3m target but likely putting additional downward pressure on the dollar. equilibrium in a broad BoP sense would be restored when the dollar falls to the point where:
a) the trade account adjusts
b) the dollar is cheap enough to induce capital inflows sufficient to cover both the remaining deficit and “Carry” related outflows
c) the weak dollar induces the fed to tighten, eliminating the carry …
Andrew R./bsetser -
Andrew’s explanation makes sense to me.
Two points:
a) As Brad points out, apart from the current account component, the “tailwinds” of gross capital outflows from a carry currency host country such as Japan must be offset by the ‘headwinds’ of required offsetting gross capital outflows. The net current account component itself is more or less neutral in its pricing effect, since inflows on current account match net outflows on capital account.
b) Not every US dollar carry trade would require an expansion of the US international balance sheet:
E.g. on a cash basis, borrowing and selling US could occur between European banks, and the net regional clearing effect in New York would be 0. No permanent inter-regional asset or liability is required. The balance sheet effect would occur within Europe.
Still, the net global effect on dollar carry trades would be 0, because somebody is taking the other side within the European region.
The same holds for forward trades and for cash hedges against forward trades.
So the asset-liability effect of gross carry trades is spread out globally, in any of the major currencies.
And not all carry trades require cross border assets or liabilities with the carry currency host country.
Type above; I meant:
a) As Brad points out, apart from the current account component, the “tailwinds” of gross capital outflows from a carry currency host country such as Japan must be offset by the ‘headwinds’ of required offsetting gross capital inflows.
Why ever would you buy the US Dollar?
http://www.prudentbear.com/index.php/BearsLairHome
The Federal Reserve’s unexpected inter-meeting cut of 0.75% in the Federal Funds rate to 3.5% was accompanied by a sharp rally in the dollar bond market, as the 10-year Treasury bond yield dropped to 3.4%. With inflation well above 4% and rising, one can only ask: why? Why would anyone buy the obligations of a shaky deficit-ridden political system in a currency that appears fundamentally unsound?
In the short term, the specter of inflation is looming ever nearer. Sharp reductions in interest rates and insertions of liquidity into the system have increased the supply of money chasing goods, at a time when commodity markets are already stretched. Hence, not only is the dollar likely to decline owing to the extra liquidity, but commodity prices are likely to rise further. The result cannot fail to be accelerating inflation; as I wrote last week, I expect even reported inflation to hit an annual rate of around 10% by the end of 2008.
In summary, like General Motors in 1970, the United States does not deserve its AAA rating and its obligations, particularly those denominated in the local “Bernanke pesos” should be avoided.
anonymous — fair point re: Europe.
presumably there are two main sources of $ liquidity in the eurozone: central bank $ deposits in european banks (eurodollars, in the original sense) and the accumulated profits of US MNCs that keep profits offshore for tax reasons, but generally prefer to hold $ rather than euros for balance sheet reasons.
good discussion; it forced me to sharpen my thinking.
“Dang, moldbug, you write good.”
seconded - the scales have fallen from my eyes; I can see clearly now… how arbitrary The System has become, which I think at least one Fed Prez gets:
“…the dynamics of production and demand among the new participants in the global economy nonetheless impact us in different ways at different times. As these new participants joined the global economy, they provided significant tailwinds, helping us grow by providing cost savings, new sources of productivity enhancement and new sources of demand, helping fatten both the top line and bottom line of our businesses while also holding down inflation… I think it is now clear that the winds have shifted. The growing appetite for raw inputs from the new participants in the global economy represents an inflationary headwind that is unlikely to soon abate… If I am correct, then the situation today is the flip side of the 1990s and early 2000s: In delivering on our mandate to be monetary policy ‘owls,’ we will have to err on the side of running tighter policy than would otherwise be justified if we wish to limit upward inflation pressures…”
“To be owlish, and to avoid the imbalance of emphasis that gave rise to needed harsh discipline imposed by the Volcker FOMC, one has to bear in mind that the seeds of inflation, once planted, can lie fallow for some time, then suddenly burst through the economic topsoil like kudzu, requiring a near-toxic dose of countermeasures to overcome.”
&
“Inflation is like absinthe. The narcotic allure of inflation is a dangerous thing. It might seem like the remedy to bail out a government or a bad book of business and forget your troubles. Yet our experience in the past has taught us only too well that inflation is a dangerous elixir that ultimately proves debilitating for businesses, consumers, investors—including those foreign investors who have lately come to the aid of some large balance sheets here—and especially for the poor, the elderly and people on fixed incomes. It even inculcates bad financial behavioral patterns in the young by encouraging spending rather than investment and saving. Inflation is bad for Main Street and Wall Street and even for Sesame Street.”
&
“As a voter on the FOMC this year, I stand ready to take substantive action to support growth and provide insurance against downside risk, as long as inflation expectations remain contained. You will note the operative qualifying words there were ‘as long as inflation expectations remain contained’… I have no intention of being party to any action that might shake faith in the dollar.”
&
“The late Dame Mary Douglas was no Ben Franklin. Nor was she a Philadelphian. She was a brilliant British economic anthropologist who wrote a pathbreaking book titled Purity and Danger. In it, she wrote something that Franklin or Stephen Girard or any good central banker since the onset of time has understood implicitly: ‘Money can only perform its role of intensifying economic interaction if the public has faith in it. If that faith is shaken, the currency is useless.’ ”
&c.
there is still a translation* that can be made from the arbitrary nominal world — where K-Fed albums have currency — to the ‘real’ world (thru the inflation rate - however measured and/or believed). if I may presume, moldbug would like to do away with the monetary abstractions of the nominal world (or at least make them explicit - it’s mold, not gold, of course!) and would instead have us operate or ‘program’ on bare metal, so to speak. for me, I think abstractions are convenient yet prone to misinterpretation if not abuse — there are good and bad translations (even languages, if you subscribe to Sapir-Whorf) — so in the end how these abstractions (or lack thereof) affects the real world** should provide the basis for any analysis; real rates are what matter, imo…
—
*moldbug’s apparent adeptitude at software programming on top (knowing the pros & cons) of ISA level hardware abstractions, I think makes him particularly suited to the task
**altho our understanding of what constitutes the real world should also be open to interpretation
Hello there, imho dollar is already a funding currency. It is based on observation of spreads for corporate and goverment and price action in bonds in Russian debt and equity market. Russian CB regulates the rate agains the basket of EUR/USD, but mostly on the bid side. When ruble depriciates there are no offers from CB. So then spreads widen or stocks get hummered USD tends to overshoot it’s basket implied target greatly.
Sorry chaps, been tied up the last couple of days. I concur with guest immediately above and the stuff that Andrew posted earlier- derivatives, et al make it eminently possible for the dollar to be a funding currency. That is perhaps the biggest difference between BWI and BWII- there are plenty of ways of doing an “end around” conventional BoP flows. Moreover, a stock adjustment in dolalr holdings (by either US or foreign holders of $$$) , if not a “carry” trade in the classic sense, will still resemble one in how it appears to the observer.
Andrew, to answer your query…I really don’t know what to buy against the dollar as a “carry” story. My preference is to bet on BWII springing significant leaks- buy GCC, JPY, hell, even CHF. In a higher volatility world, carry is less attractive…so it seems to me that the optimal strategy is to buy currncies where there is either an embedded or implicit short position, or a realy policy need to alter the current regime. It sounds as if the GCC, for example, is “smoking” again. Another option might be RUB, where inflation is clearly running out of control and they have yet to deliver the usual early-year reval in the basket; quite the contrary, as the RUB has weakened against the basket so far this year!
Macro Man - the problem with Rub at the moment is that non residents are pulling money out of Russia. And our CB is not willing to drag reserves to
revalue, so the rouble is really “floating” above the basket. If the market rout ends and foreign money comes back, rouble will appreciate. Januray inflation was 24% annulized.
interesting points re: Russia.
Macroman — so who takes the $ long in the derivative position? if it is someone who then has to hedge via a real money outflow to take an offsetting $ short, it ends up adding to the United States’ funding problem.
If by contrast it is someone who thinks the $ has fallen far enough and wants to be long the $, it doesn’t produce an offsetting outflow — but the fact that the $ long offsetting a $ short means that someone else needs to take a $ long to fund the US current account in dollars.
why do posts on the c. trade produce better discussions than posts on China?
I made a scheme with carry traders, importers and exporters and their hedging needs.
1) Japan -> Australia
carry trader: danger: jpyaud up, hedge: jpyaud long
contract prices in aud
jpy exporter: danger: audjpy down, hedge: audjpy short = jpyaud long
aud importer: safe
contract prices in jpy
jpy exporter: safe
aud importer: danger: jpyaud up, hedge: jpyaud long
2) USA -> Norway
carry trader: danger: usdnok up, hedge: usdnok long
contract prices in nok
usd importer: danger: nokusd up, hedge: nokusd long = usdnok short
nok exporter: safe
contract prices in usd
usd importer: safe
nok exporter: danger: usdnok down, hedge: usdnok short
Seems like there are more natural counterparties for the carry trader in case 2.
Btw. very interesting discussions about carry trade on this blog. Thanks to all participants.
indian rs to us dollar will in august 2008
38-39 RATE DIFF