Recent Posts

Wednesday, July 17, 2019

The External Sector And Financial Crises

If we look at the full history of financial crises around the world, one could argue that crises related to external debt and/or fixed exchange rates are dominant. Such crises could represent an entire chapter of this book. However, I will only offer a brief overview of the subject. From the perspective of recession forecasting, the addition of a fixed exchange rate regime adds a new wrinkle to analysis: at what point will the peg fail, causing a crisis? As I will discuss below, this is quite different than an analysis of the domestic economy, which one might hope is amenable to something resembling econometric analysis.

(Note: This is an unedited excerpt from my manuscript on recessions. The section overlaps another section, and I will need rewrite them to eliminate redundant text. As a result, I am making a lot of assertions without much in the way of references or data, I will fill in details later.)

It should be noted that this is a topic that is a source of deep controversy between Modern Monetary Theory and other groups of post-Keynesians. However, my experience is with floating currency developed sovereigns, and as I discuss here, there are few crises that are precedents for such countries. As such, I lack the familiarity with these episodes, and prefer to keep my comments general.

Fixed Exchange Rates: They Eventually Fail

A fixed exchange rate system is where a country declares that its local currency can be converted into another instrument, typically either gold or a foreign currency. If we look at the history from after the end of World War I, being able to convert to gold was largely the same thing as pegging to one of the major currencies (since they were all nominally pegged to gold).

It is very easy for such systems to break down. If a country has a current account deficit, it implies that there must be a corresponding capital inflow (in a broad sense) as a result of accounting identities. More simply, if a local entity buys something from a foreigner, they need to buy the foreign currency to import it – which implies the need for an offsetting purchase of the local currency. That is, one of the following typically occurs.
  • Foreign entities are buying financial instruments in the local economy; quite often fixed income instruments. This is essentially the same thing as “buying the local currency,” since that is typically executed by buying financial instruments (or making deposits in local banks); buying and selling of notes and coins (e.g., dollar bills) is relatively small when compared to cross-border flow.
  • Local entities are borrowing overseas (which can be viewed as functionally equivalent to the previous point).
  • Local entities sell foreign assets and use the proceeds to “buy the local currency.”
  • If the following are not enough to balance the foreign exchange market, the central bank needs to intervene in the market: buying or selling reserve currencies or gold in order to buy back its local currency.
The final possibility – central bank intervention – is what differentiates a peg from a floating currency. During normal conditions, market participants expect the peg to hold. When the currency gets near the limit of its trading band (in practice, currencies are allowed to trade in a range around the peg value), private sector actors will step in to buy/sell in such a way to move the currency back towards the centre. For example, if the price is near the low end, traders will buy. If the peg was credible, this was a low risk trade: the central bank would prevent any further fall, while it could rise towards the middle of the peg band again. As such, the peg could remain in place, even though the central bank does not intervene.

The problem is the credibility of the peg. If a currency is strengthening, there are no problems: the central bank buys foreign currencies (or gold) with domestic currency to keep its value from appreciating. Since the domestic currency is just a liability of the central bank, there is no limit to this process. The problem is currency weakness. The central bank needs to have something to sell to intervene with – either gold or a foreign currency (currency reserves). When they run out, they can no longer intervene. This means that the currency will have to fall to a level where supply and demand in the private sector is matched. (In practice, pegs are typically broken before those reserves run out.)

The previous discussion was quite loose, but neoclassical economists have formalised this discussion with mathematical models, such as the analysis of self-fulfilling currency runs.[i] The process seems similar to other market processes (such as self-fulfilling runs on borrowers), so it seems to be a distraction to chase after the details of the formalisation here.

Since central banks can stop appreciation, if all countries trusted each other completely, they could always intervene to allow the peg values to hold. In practice, such unlimited trust does not exist, and so foreign central banks will allow foreign currencies to suffer runs and have the peg break on the other side.

The problems posed by a peg breaking are quite clear. In addition to the policy reaction (discussed next), domestic entities may be borrowing in a “hard” foreign currency. If the value of the currency collapses versus those hard currencies, the value of the loan rises markedly in the local currency. This will cause financial distress, and thus a financial crisis.

Recessions Ahead of Peg Breaking

However, peg breaking is the end game. Countries typically preserved parities by fiscal austerity policies – cutting spending, raising taxes. This reduces domestic demand, and thus imports. This should improve the trade balance, which allows the country to rebuild foreign currency reserves. The other avenue is to raise interest rates, which will attract foreign capital inflows, and presumably slow the economy (as will be discussed in Volume II). Although it is unclear why raising interest rates will raise a currency value in a floating rate regime, there is a mechanism for it support a currency with a peg. So long as the peg holds, local bonds with a higher yield will outperform lower-yielding foreign bonds. (For a floating currency, currency movements normally swamp the effect of yield differentials on short holding periods.) This creates capital inflows – which balances the currency market without central bank intervention.

Fiscal austerity and raising interest rates are conventionally viewed as a policy tightening, and hence can cause a recession. These are policy-induced recessions, and so seem to be somewhat forecastable. Therefore, this perhaps is a strong counterexample to my arguments that recessions are hard to forecast. If we can forecast the need for a policy tightening to preserve a currency peg based on trends in the currency reserve position, we could presumably forecast the recession.

Even if we accept that argument, we are faced with the hard-to-forecast issue of political risk. Rather than absorb a recession – or a depression – policymakers can abandon the currency peg. Judging the likelihood of such an outcome is perhaps a job for analysts of political economy, and not amenable to econometrics.

The previous summary captures the main issues around fixed exchange rates in this context; beyond that, we need to dig into institutional detail. Those details are important for analysing the risk of crisis or recession. In the modern era, the euro area and developing countries give us examples. If we look further back, the relationship between the interwar Gold Standard and the Great Depression gives us a large data set on the interactions. My thinking is largely based on the analysis found in the classic Golden Fetters by Barry Eichengreen.[ii] It should be noted that there are many defenders of the Gold Standard, who argue that the interwar problems were due to other policy errors, and/or problems with fractional reserve lending. These are typically from the Austrian school, as will be discussed in Section 5.6.

Crises in Floating Currencies?

There is not much doubt that fixed exchange rates are associated with marked recession risks that are associated with the regime. The controversy lies more in the area of floating currencies. Can the external sector induce a crisis that leads to recession? For example, if one follows the business press, one argument that pops up periodically is that foreign borrowers own a disproportionate share of U.S. securities, and that there will be a “buyer’s strike,” and an associated financial crisis.

From a more academic perspective, there were considerable worries about financial imbalances and the “unsustainable” nature of the American current account deficit. It has been argued that the problem was not that economists did not predict the 2008 Financial Crisis, rather they predicted the wrong crisis.[iii] Since I do not see much value in taking apart forecasts made a decade earlier, I will not dwell on that point.

In my view, searching for an external sector culprit behind a financial crisis is not needed – since the factors leading to the crisis do not materially depend upon the nationality of the sources of finance. My reasoning is straightforward. Firstly, floating currencies do not fall forever; at some valuation becomes attractive and the market reaches a new flow equilibrium Secondly, currency volatility prevents the buildup of positions by investors who are concerned about currency risk in the first place. I discuss these in turn.

Currencies Eventually Stabilise

An exchange rate is a relative price: one currency unit for another. If we look at the post-1990 period, inflation rates in the developed countries have been quiescent, bouncing around 2% for most countries. As such, each currency has relatively stable purchasing power for domestic goods and services, including the cost of wages.

The stability of wages has one side effect: if a currency falls rapidly versus its developed peers, the cost of wages falls relative to other countries. This drops input costs for production relative to other countries. And even if imported inputs rise in price in domestic terms due to the drop in the exchange rate, those input costs are unaffected when expressed in terms of the foreign currency unit.

The result is that domestic exporters suddenly have greater prospective profit margins versus their international peers. This will have two effects: buoy the attractiveness of the local equity market and attract investment inflows (either reallocations of capital by multinationals, or foreign direct investment).

These capital flows (and the prospect of future flows) help put a floor under the domestic currency. This helps explain why there have been no cases of developed currencies going to zero in the foreign exchange marketplace.

Who Takes Foreign Exchange Risk?

One empirical regularity that is often overlooked: the general absence of defaults caused by exchange rate movements in developed countries. The explanation is a reversal of one of Minsky’s catch phrases: instability is stabilising.

Borrowers are aware of the risks of borrowing in a foreign currency. To the extent that it is done, it is done on a currency-hedged basis. These hedges may be outright hedges, or the implicit hedges created by having foreign operations. For example, Canadian firms operating in the United States, or with significant U.S. dollar denominated revenue may borrow in the U.S. dollars; this is reducing their foreign exchange risk. Obviously, some firms can default, but defaults happen for any number of reasons. (The situation in Iceland in 2008 was a fiasco, but it is impossible to characterise the Icelandic firms as being run by responsible grownups.)

Owners of financial assets have a mixed picture.
  • Banks balance sheets are almost entirely hedged. They may run foreign exchange risks that appear large to individuals, but those risks are very small versus the size of their balance sheets. If we look at the Financial Crisis, the problems faced by banks was that many foreign banks were holding U.S. dollar-denominated assets, but their deposit base was in foreign currencies. The willingness of counterparties to offer them U.S. dollar wholesale funding or via hedges (cross-currency basis swaps) waned, and they were forced to turn to their central banks’ swap lines with the U.S. Federal Reserve. That is, the current account deficit country had to extend backdoor bailouts to banks in other countries – including the trade juggernaut, Germany.
  • Bond funds are slaves to their investment mandates. At present, the bulk of bond funds are being held for liability management purposes, particularly by pension funds and insurance funds. These funds are required to manage their portfolios against their projected future cash flows – which are in the domestic currency. Multi-currency funds exist but are a hard sell in a world of low yields. The funds with domestic benchmarks may hold foreign currency bonds, but they do so in a hedged fashion. This need for hedging is the counterpart to the desire of issuers to hedge their borrowings. Offsetting flows allow borrowers and lenders to diversify their sources/sinks for funding – although these do not provide net currency flows. The international bond market is mainly for allocating credit and duration risk; the effect on currency flows is smaller.
  • Individuals may hold foreign currency bonds; Japan is well-known for having retail investors that gamble on foreign currency debt. Although these exposures are unhedged, they are quite small relative to institutional flows.
  • Foreign exchange funds, and currency overlays on portfolios do take currency positions, but their balance sheet capacity is still relatively small when compared to gross capital flows.
  • Foreign official currency reserve managers bear currency risk on their portfolios. However, as governmental institutions, they are normally spared the worries about mark-to-market. Instead, their objective is to manage their exchange rate, typically trying to keep it at a level that facilitates exports. The other constraint on their actions is that the largest reserve managers are trapped in the markets with the deepest bond markets (notably the United States). Even if they really liked New Zealand bonds, they cannot do too much about that view.
  • The residual class of investors are the ones who absorb most of the currency risk: equity funds. International equity funds are popular, and they are rarely currency hedged. Courtesy of buoyant equity valuations, the equity market has the heft to absorb cross-currency flows from trade.
The reality that a falling currency helps local equity valuations is the secret sauce behind the resilience of floating currencies. To use an expression from Canadian/American football, floating currencies bend, but do not break.[iv]

No Firm Dividing Line Between Domestic and External Crises

The nature of modern financial market behaviour makes it hard to draw a line between an “external” and “domestic” private sector financial crisis: foreign and domestic investors are commingling across the globe already. If a sector of the economy launches an ill-fated investment scheme that draws in credulous investors, the odds are that those investors are from all over the globe. Since those investors have currency hedges in place, their concern is not the value of the currency, rather it is the prospect of credit losses. The tendency to panic at the prospect of credit losses is one of those wonderful properties that is shared by all human beings, regardless of their nationality.

For this reason, searching for an “external” cause of a financial crisis is largely a waste of time: the crisis will happen because some class of borrowers incurred debts that cannot be repaid. The reasons why they cannot be repaid will depend upon domestic conditions; the sources of their lending is only a detail of interest to purveyors of financial horror stories.

Events in the Financial Crisis do point to a counterexample to this logic: domestic financial institutions can grow out of control in foreign markets, such as the Icelandic banks. They fail because of events outside the domestic economy. However, such an event is largely driven by regulatory failures, and not conditions in the domestic economy. The only way to diagnose such problems is understanding the risks of the financial institutions – but those risks are almost certainly being obscured by said institutions.

Policymaker Panic

We now turn to the possibility of a governmental financial (fiscal) crisis. From a practical perspective, the main risk associated with a floating currency is the probability of policymakers panicking because of a falling currency value. For example, Canadian policymakers were worried about the falling Canadian dollar in the early 1990s. (One could point to the United Kingdom going to the IMF in the 1970s, but that was clouded by fixed exchange rate concerns.)

The usual panic responses are “emergency rate hikes” and/or cutting spending. Any resulting recession is policy-induced. Although the business press might write about an external “fiscal crisis,” the reality is that foreigners have no more power to force an involuntary bankruptcy than domestic government bond buyers.

Nevertheless, we need to be realistic about the politics. There is a large contingent of politicians/economists that want to see government spending cut, and they will seize any excuse to justify those cuts. We need look no further than the voluntary fiscal austerity policies pursued in the United Kingdom after the Financial Crisis as an example of such political opportunism.

Therefore, predicting such a self-imposed recession is a question of political economy. Certain governments might panic in response to a weaker currency, causing a deep recession. Other governments will just sail through the currency weakness, waiting for valuations to revert. As such, predicting such recessions is not going to be easy based on econometric analysis.

Finally, one can imagine a country being cut off from foreign trade as a result of some geopolitical crisis and forced to defend the currency in order to be able to pay for strategic imports. Although possible, this is yet another question of politics, and not econometric analysis.

Ultimately, Political Risk

In summary, the risks around financial crises (including fiscal crises) that are due to the external sector end up being political: will the authorities induce a recession to defend a peg, or let it break? Will a floating currency sovereign government panic and raise rates/cut spending in order to defend its currency value? Economic analysis can extrapolate trends and tell us what the economic pressure are on the government, but how policymakers react are not amenable to solution in a mathematical model.


[i] One early paper is “Rational and Self-Fulfilling Balance-of-Payments Crises” by Maurice Obstfeld, NBER Working Paper 1486, November 1984. URL:

[ii] Golden Fetters: The Gold Standard and the Great Depression 1919-1939, by Barry Eichengreen, Oxford University Press, 1992. ISBN: 0-19-510113-8.

[iii] For example, see “Macroeconomics Predicted the Wrong Crisis”, by Adam Tooze, September 10, 2018. URL:

[iv] A football defence that “bends, but does not break” is one that is willing to give up short gains, but prevent big plays. It forces the opposing offence to slowly grind down the field, waiting for a miscue that forces a punt.

(c) Brian Romanchuk 2019


  1. I don't think you are looking at the root of the import-export issue here. I think the root is found by looking at whether exporter or importer owns currency at the end of a currency-goods exchange. Then figure out how any long term imbalance is funded on a financial level. Exports can be a national policy.

    Let's assume that a foreign producer sends semi-durable goods to an importing country as a part of national policy and retains ownership until final sale to an importing entity. The critical junction here is which party holds who's currency at exchange completion.

    Having made the sale, the exporting producer owns some of the importers currency. At this point, someone has a problem. The semi-durable goods have been made using one currency and sold in another currency. Does the goods maker (the exporter) have a problem or does the financier funding the goods maker have a problem? The financier is the one with a problem.

    Now let's assume that a central bank is the base financier, such as in China. The Chinese CB can accept foreign collateral (such as US Treasury bonds) as repayment for the cost of semi-durable goods construction. This does not need to be done directly, but can be done indirectly by using private (but state supervised) banks or other mechanisms.

    Within this framework, floating currencies become a secondary balancing mechanism. Exactly what is to be balanced when governments and central banks use money creation (by increasing debt) to fund job creation, regardless that the fruits of the jobs are to be exported.

    How this framework would fit into recessions is unclear. The fundamental effects should be slow to discern but once connections are made, political reactions tend to be swift. Witness tariffs.

  2. From the viewpoint of an engineer learning economics, this is a brilliantly clear exposition. Please be careful not to muddy it when you rewrite (to eliminate redundant text).

    1. Thanks - from an ex-engineer who’s still learning economics.

  3. Brian,

    It seems to me you are firmly making the case for foreign currency being a constraint, a constraint that MMT vehemently denies.

    Henry Rech

    1. I don't think that MMT denies that a lack of foreign currency could be a constraint on those that need that particular currency to import goods that can't be produced domestically.

      Not that I totally agree with what MMT says about imports/exports either...

    2. Henry,

      I don’t see how you draw that conclusion. I just did my best to present the case.

    3. Brian,

      Firstly, you point out that crises in fixed exchange rate systems can cause catastrophic circumstances and eventual collapse of the currency. (A good example is the international capital inflow/outflow boom/bust cycle that precipitated the Asia Crisis in the late 1990s.)

      Secondly, really what is the difference between a collapse in a fixed currency and a floating currency? - nothing other then the rate of adjustment - the apparent shock effect. (It's only a shock to those that weren't expecting it. Generally, capital markets are the bellwether - they know and they move, ultimately causing the pressure for a step devaluation in the case of a fixed exchange rate system.)

      So you have made the point that currency collapse has severe consequences and the only difference between the collapse of fixed currency and a floating currency is the rate of collapse. A collapse is a collapse.

    4. The Canadian dollar has made very large movements versus the USD over the past decades. Nothing bad has really happened. You can call it a “collapse” and write op-eds about how terrible it is, but you cannot point to anything that’s particularly negative. Very different from the Great Depression, where policymakers here had an insane desire to defend the gold parity at any cost.

    5. Brian,

      Same with the Aussie dollar - it has gyrated between US$0.50 and US$1.10 since floating in the mid 1980s.

      What is difficult to discern is what pressures it placed on fiscal and monetary policy. Perhaps nothing bad happened because the policy responses were adequate. But there were po0licy consequences.


    6. It’s very hard to see any actual consequences. The pass-through to inflation has been largely trivial in any study I’ve seen.


Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.