Understanding balance of payments crises in a fiat currency system
It's weird. Whenever I say that floating exchange rates can't absorb all shocks and that balance of payments crises can happen even in fiat currency systems, I am accused of gold standard thinking. Gold standard? Me? Perish the thought. I am the world's biggest fan of fiat currencies. And of floating exchange rates, too. But that doesn't mean I regard them as a panacea.
Firstly, about gold standards. Under a strict gold standard, the quantity of money circulating in the economy is effectively set externally. The domestic money supply can only grow through foreign earnings, which bring gold into the country. I have said a "gold standard", but the same is true of any FX reserve-backed fixed exchange rate system such as a currency board: gold is really only a universal FX reserve. The domestic money supply grows as the supply of FX reserves rises, and falls as the FX reserve supply falls. It isn't strictly true to say that a trade surplus is necessary for the economy to grow, but if the economy grows without a corresponding increase in net exports, the result is deflation.
This is evident from the quantity theory of money equation MV = PQ, which is fundamentally flawed in a fiat currency fractional reserve system but works admirably under a strict gold standard or equivalent. When the supply of goods & services (Q) rises but the quantity of money in circulation (M) remains fixed (and assuming that V also remains constant), consumer prices (P) fall. Some people regard this type of deflation as benign, though it is worth remembering that even in a naturally deflationary system, expectations of future price falls can dampen aggregate demand. A little bit of inflation can be a good thing. Nonetheless, this sort of deflation is far removed from the vicious deflationary spiral described by Irving Fisher in his Theory of Debt Deflation.
So under a gold standard or equivalent, it is much easier for countries to grow if they run trade surpluses. This is not because exports create growth, but because inflows of FX reserves act as a monetary stimulus: the risk, of course, is inflation. Conversely, trade deficits involve a net outflow of FX reserves, which is monetary tightening. There is thus a direct relationship between domestic growth, domestic inflation and the external balance.
More seriously, FX reserve outflows in a fixed exchange rate system can result in a balance of payments crisis. Because central banks can't print foreign currencies or gold, a country running a persistent trade deficit* can literally run out of money. The IMF's original role was to provide emergency funding to countries facing such a disaster and help them implement policies to restore the trade balance. Those policies typically involved depressing domestic demand to curb imports and reduce inflation, putting downwards pressure on business costs in order to improve external competitiveness, reducing external debt, and devaluing the currency relative to its anchor. This last was key. In a fixed exchange rate system, devaluation is domestic monetary stimulus, because it allows more money to be created relative to the anchor. Thus the IMF's standard solution to a balance of payments crisis was austerity offset by monetary stimulus in order to generate export-led growth. Does this sound familiar?
The era of gold standards ended in 1971 when President Nixon suspended the convertibility of the US dollar to gold. We now supposedly have a universal system of fiat currencies managed by inflation-targeting central banks.
In a fiat currency system with floating exchange rates, the quantity of money in circulation is determined by the central bank**. The country is not reliant on FX or gold inflows for domestic money growth, and FX or gold outflows do not threaten the country's solvency. All the central bank needs to do is create (or destroy) the amount of money needed to maintain a target inflation level and allow the external value of the currency to adjust. The external balance is not directly linked to growth or inflation, and trade deficits do not cause crises.
In such a system, a trade deficit is a monetary drain which must be offset by money creation. Central banks can create unlimited quantities of their own currency: if the currency is flowing out of the country via a trade deficit or capital flight, the central bank can simply create more of it. Note that in a floating-rate fiat currency system - unlike the gold standard and its relatives - it is not necessary to devalue the currency deliberately to enable more money to be created. Creating money for domestic monetary stimulus may devalue the currency, of course, but that is a side effect.
Equally, the inflationary effect of a trade surplus can be offset by a monetary drain (raising interest rates). This tends to raise the exchange rate, discouraging exports and encouraging imports. In a floating exchange rate system where central banks are targeting inflation, trade imbalances should in theory be benign and short-lived.
However, this is not what we have. True, we have fiat currencies and inflation-targeting central banks. But we don't have universal floating exchange rates. The ghost of the gold standard still haunts the world monetary system, living on in the proliferation of currency boards, crawling pegs and other, more subtle versions of currency fixing. And there remains an enduring belief that depressing domestic demand and improving external competitiveness is necessary to restore growth. This is gold standard thinking, and it is not appropriate in fiat currency systems.
But even if every country had a floating exchange rate, balance of payments crisis would remain a risk for many countries.
The idea that currency depreciation will absorb all shocks rests on the assumption that there is infinite demand for all currencies. The balance of trade therefore simply reflects the movement of the domestic currency in and out of the country. Since there will always be some price at which the currency is exchangeable, the country can always pay for everything in its own currency and has no need for FX reserves. If an exporter doesn't want to accept the currency of a particular country, the country can refuse to trade with him, to his detriment.
Sadly, it is not that simple. All currencies are not equal. Just as the price of government debt depends on market views of future economic performance and government trustworthiness, so too does the exchange rate of a sovereign currency. There is a hierarchy of currencies. Currencies at the top of the hierarchy - the "reserve currencies" and a few other advanced-economy currencies - are highly liquid: they are readily exchangeable for almost any other currency and welcomed in most countries. The introduction of central bank swap lines also means that these countries can effectively treat certain other "premier" currencies as their own, further reducing their need for FX reserves. The chances of a genuine balance of payments crisis in these countries are very low.
But this is not true of other countries. Currencies at the bottom of the heap are not welcomed externally: there may be no market for them, and all trade with the country may be conducted in foreign currencies. The worst currencies may not even be particularly welcomed domestically: widespread use of "hard currency" is a feature of basket case economies. Most emerging market currencies are welcomed domestically but have only a qualified welcome externally.
Clearly, a country whose currency is not widely accepted externally must trade in foreign currencies. It will have to pay for imports in foreign currencies, and since it cannot print foreign currencies and there is a limit to FX borrowing, it will look to balance this with foreign currency income from exports. If the country's FX income from exports exceeds its FX obligations, it can meet those obligations from current FX income regardless of the external value of its own currency. In fact, when the currency depreciates, improved terms of trade would mean an increased inflow of FX reserves. This looks very much like the gold standard, doesn't it?
Of course, if the country is dependent on imports for essential goods - foodstuffs, medical supplies, energy - then a rapidly depreciating currency would be likely to result in rising demand for FX as exporters to the country demand payment in other currencies. But terms of trade improvement could offset rising FX demand. Alternatively, the country can run a persistent trade surplus in order to build up FX reserves, creating a buffer to protect itself from short-term trade fluctuations. This is what many emerging market countries learned to do after the 1997 Asian crisis. Better beggar-my-neighbour competition and depressed global trade than a repeat of 1997 - or 1931.
When a fiat currency is not welcomed externally, its exchange rate falls regardless of the country's external balance. We see this at present with Venezuela: the real exchange rate (not the government fix) of the bolivar is falling catastrophically despite Venezuela's trade surplus. Venezuela is an oil producer, so the (real) exchange rate of its currency to the US dollar naturally tracks the price of oil. But the bolivar's value is falling far more than that. Such rapid falls set up a feedback loop, whereby those holding bolivars exchange them for hard currency, gold and other non-perishables as fast as they can, causing the exchange rate to fall even faster. This is how hyperinflations develop. Venezuela's inflation rate is currently thought to be over 750% and rising fast. Hyperinflationary capital flight causes FX and gold reserves to leave the country at a rate of knots. No improvement in terms of trade can possibly keep up with this. Severe balance of payments crisis is characteristic of hyperinflations.
Even without hyperinflation, balance of payments crisis in an FX-dependent economy with a local floating-rate fiat currency has widespread and damaging effects on private and public sector balance sheets. Most countries that run persistent trade deficits have to import FX. The central bank does this by selling its own currency, causing it to devalue, which in normal times helps to rebalance the trade position: the central bank then provides FX liquidity to banks for them to support FX demand from households and corporations. It is normal for banks to have currency mismatches on their balance sheets, and it is also normal for many exporters and importers: these mismatches may take the form of FX debt and local currency assets, or vice versa. Households, too, may have currency mismatches if they take on foreign currency debt, for example FX mortgages (Poland please note).
These currency mismatches are toxic. A balance of payments crisis is fundamentally a solvency crisis. A sovereign currency issuer can't run out of its own money, but it can run out of other people's. And if it does, then neither the private sector nor the sovereign can service foreign currency debts. If the currency is already depreciating rapidly, the central bank selling the currency to buy FX simply worsens its collapse: monetizing sovereign deficits has a similar effect. Since the private and public sector cannot service their existing debts, they cannot borrow to meet their obligations. The result is sudden, widespread and very damaging debt default among banks, corporations and households, along with severe cutbacks in consumption. Even the sovereign can be involved, if it has high levels of FX debt. Currency mismatches on corporate, sovereign, household and bank balance sheets are the blue touchpaper that is lit when a balance of payments imbalance becomes a crisis.
It is thus quite wrong to suggest that in our fiat currency system, balance of payments crises cannot happen. They can, and they do. And this, I think, partly explains the prevalence of export-led growth models, particularly in emerging market economies. The protection offered by a sustained trade surplus from the long-lasting and horrible effects of balance of payments (FX) crisis more than compensates for reduced living standards arising from repressed domestic demand.
* For the purposes of this post, "trade" includes services.
** As I don't wish to get caught up in arguments about whether governments do or don't create money when they spend, I am preserving the fiction of central bank and government separation. This means that the language in this post is that of monetarism, rather than MMT. I do not apologise for this: it is my firm belief that MMT and market monetarism are brothers under the skin, and the differences between them are largely semantic. Though there might be a difference in political ideology too.
UPDATE TO FOOTNOTE: This post is NOT about "who creates money". I'm well aware of the limitations of the monetarist framing I have chosen, and I did say that the quantity theory of money is fundamentally flawed in a fiat currency system. However, I've used a monetarist framing to enable a direct comparison of gold standard and fiat currency systems from a trade perspective. For the purposes of this post it does not matter whether the CB creates money, commercial banks do so when they lend or the government does when it spends. I'm not going to host a debate here about the effectiveness of monetary policy versus fiscal policy. Comments about the creation of money are therefore off topic for the purposes of this post.
Rethinking government debt
Competitive devaluation is not a free lunch
Competitive devaluation plus monetary expansion does create a free lunch - David Glasner
The limits of understanding of MMT - Neil Wilson
How to think about the balance of payments - JW Mason
Image from The Sunday Times.