Emerging Market Debt Report 2013 — The Dollar Discovers Its Reach

In 2013 an official communication indicated that asset purchases might slow at some future point. No purchase was reduced. No rate was changed. Emerging market currencies, local bond yields and hard-currency spreads moved violently within weeks.

That reaction is the whole finding. If a change in the expected future path of policy — with no change in policy itself — can produce that response, then the position being unwound was built on the expected path rather than on anything happening at the time. The tantrum did not create a vulnerability. It revealed one that had been accumulating since 2009.

What had accumulated was a liability structure. With policy rates at their lower bound across the major economies, capital searched for yield, and a large share of it went to emerging market debt — sovereign and corporate, hard currency and local. From the borrower's side, dollar funding was available at rates that had never previously been offered to them, and they took it.

The resulting exposure is a mismatch, and it is the archive's recurring mechanism in a new form. The Global Investment Outlook 2008 describes institutions holding long assets against short liabilities. Here the mismatch is not maturity but currency: revenue earned in pesos, liras or reais against obligations fixed in dollars. Nothing about it is visible while the exchange rate is stable, and it is the entire exposure when the dollar moves.

Two things about the sorting are worth stating up front, because both were misread at the time.

The market did not discriminate on growth or on credit quality. It discriminated on external funding need — the current-account deficit and the reliance on foreign capital to close it. A fast-growing economy with a large deficit was punished more than a slower one that funded itself. That is a funding judgement, not an economic one.

Local-currency issuance did not remove the problem. It is genuinely safer for the sovereign, which can no longer be forced into default by a currency move. But when a large share of local-currency debt is held by foreign investors who account in dollars, a currency fall is a loss to them regardless of the coupon being paid in full — and their response is to sell, which lowers the currency further.

What four years of low rates did to capital flows

The position that unwound in 2013 was built by policy in the major economies, and its construction is worth setting out because the mechanism is entirely ordinary.

The starting condition. Policy rates at or near zero, with large-scale asset purchases compressing yields further along the curve. For an investor with a return requirement — a pension fund, an insurer, an endowment — the assets that had historically met it no longer did.

The consequence is a search, and it has a direction:

• Duration first. Buy longer-dated versions of the same credit. Available until the curve flattens.
• Credit second. Move down the quality spectrum for spread.
• Geography third. Move into markets where the local policy rate has not been compressed, which meant emerging markets.

From the borrower's side the offer was genuinely attractive, and it would have been irrational to refuse it:

• Dollar funding at rates far below their domestic cost, available in size and at maturities not previously offered.
• A deep and receptive investor base, so issuance could be large without moving the price much.
• A currency that had been stable or appreciating, which made the dollar liability look cheap in local terms — the same appreciation-substituting-for-affordability pattern the US Housing & Mortgage Report 2008 describes in a different market.

Corporate issuance is the part most often understated. Attention focused on sovereign borrowing, but a great deal of the dollar liability was taken on by emerging market corporates — frequently through offshore subsidiaries, which meant it sat outside the external debt statistics that officials and investors were monitoring. The measured exposure was smaller than the actual one, and the gap was structural rather than concealed.

Every participant behaved sensibly. The investor needed yield, the borrower needed capital, the price was fair on the information available. The aggregate position was a single bet on the dollar, and nobody had placed it deliberately.

The mismatch, stated precisely

The mechanism deserves the same precision the archive gives the maturity mismatch, because it behaves the same way.

The structure: earn revenue in a local currency, owe obligations in dollars. Profit from the difference between the low dollar borrowing rate and the higher return on local assets.

Why it is profitable: dollar rates were far below local rates. The carry is real and it accrues steadily, which is what makes the position accumulate.

Why it is fragile: the profit depends on the exchange rate not moving against you. The borrower is not being paid for credit risk alone — they are being paid for taking currency risk, and that risk is invisible for long stretches because exchange rates trend.

Three properties make it acute, and they compound:

• The debt burden rises exactly when capacity to service it falls. A currency depreciation increases the local-currency cost of the dollar debt at the same moment it usually signals domestic difficulty.
• Hedging is available and is not free. Hedging the exposure costs approximately the interest differential — which is the entire reason the borrowing was attractive. A fully hedged position removes the advantage, so in practice much of it was unhedged, and the decision not to hedge was a rational response to the economics.
• The response is collective. When the currency moves, every holder of the mismatch wants dollars simultaneously, and buying dollars is what moves the currency further.

This is a run in a different variable. In 2008 the withdrawal was of short-term funding. Here it is of foreign capital, and the price that adjusts is the exchange rate rather than an asset price — but the self-fulfilling structure is identical, and so is the fact that solvency is not what determines the outcome.

Why the announcement moved more than the policy

The most instructive feature of 2013 is the asymmetry between how the position was built and how it was unwound.

How it was built: gradually, and against resistance. Asset purchases were introduced incrementally, debated at each stage, and absorbed by markets over years. The flow into emerging markets accumulated slowly as investors reallocated.

How it was unwound: immediately, and on words. A statement about the possible future path of purchases repriced the asset class within weeks, before any purchase was reduced.

Three reasons this asymmetry is structural rather than a quirk of that episode:

• Prices reflect expected paths, not current levels. The value of a long-dated asset depends on the entire expected future discount rate. Changing the expected path changes the price today, whether or not anything has yet happened.
• Positioning is one-directional and crowded. A trade that has accumulated over four years is held by a large number of participants with similar exposure and similar triggers. The exit is narrower than the entrance, which is the positioning-event framework the Global Investment Outlook 2018 develops.
• Uncertainty itself is repriced. The announcement did not merely shift the expected path; it reintroduced the possibility that the path could change at all. After four years of one-directional policy, that possibility had been priced out.

Entry into unconventional policy is a series of actions absorbed over years. Exit is a change in belief, priced in an afternoon. That asymmetry is not a communication failure; it is what prices are.

The lasting policy consequence was the recognition that the exit path itself required managing — which is why subsequent normalisations were signalled far in advance and executed slowly, and why markets in later cycles reacted less. The Global Investment Outlook 2021 examines what happened when that carefully managed approach met an inflation shock.

What the market actually sorted on

The differentiation across countries in 2013 is the clearest natural experiment in the episode, and it did not follow the variables most analysts were watching.

What did not predict the move:

• Growth rates. Fast-growing economies were among the worst affected.
• Sovereign credit ratings, which barely moved and did not order the reaction.
• Fiscal position. Several of the most affected had unremarkable public finances.

What did predict it:

• The current-account deficit. The larger the external deficit, the more foreign capital required to fund it, and the more exposed to that capital stopping.
• Reliance on portfolio inflows rather than on foreign direct investment. Portfolio capital can leave in days; direct investment cannot.
• Short-term external debt relative to reserves — the coverage ratio determining whether a country can meet near-term obligations without market access.
• The share of local-currency debt held by foreigners, for the reason set out below.

The countries grouped as most vulnerable shared external funding characteristics and little else. They differed in growth, politics, sector composition and institutional quality. The common factor was the need for continuous foreign capital, which is a funding characteristic, and it is the same finding as the Global Investment Outlook 2008: institutions and countries alike are sorted by liability structure rather than by asset quality when funding conditions change.

Reserves worked, and the way they worked is worth noting. Countries with substantial reserves relative to short-term obligations were less affected — not primarily because reserves were spent, but because their existence made a funding failure implausible. The same unexercised-backstop mechanism the Spain & Banking Union Report 2012 describes: capacity that removes a scenario need not be used.

Local currency debt and the mismatch that moved

A widely drawn lesson from earlier emerging market crises was that countries should borrow in their own currency. Many had done exactly that. 2013 showed what it does and does not solve.

What it genuinely solves. A sovereign borrowing in its own currency cannot be forced into default by a currency move, because it can always create the currency of repayment. That removes the mechanism behind several earlier crises, and it is a real and durable improvement.

What it relocates. If a large share of that local-currency debt is held by foreign investors who measure returns in dollars:

• A depreciation is a loss to the holder even when every coupon is paid in full. The credit performed; the investment did not.
• The rational response is to sell, which requires converting proceeds to dollars, which pushes the currency down further.
• Domestic yields rise as foreigners exit, tightening domestic financial conditions at the worst moment, without any change in the sovereign's creditworthiness.

So the mismatch did not disappear. It moved from the borrower's balance sheet to the holder's, and became an exit risk rather than a default risk. The sovereign is safer; the market is not more stable.

The practical implication is about the composition of the holder base, not the currency of issuance. A local-currency bond market funded domestically — by resident pension funds and insurers with local-currency obligations — is genuinely resilient, because its holders have no reason to sell on a currency move. The same instrument held predominantly by foreign investors is a hot-money position wearing local-currency clothing. Two markets that look identical on issuance statistics behave completely differently under stress.

The dollar as the global cycle

The general conclusion 2013 supports is broader than emerging markets, and it reframes what a domestic monetary policy is.

The observation. Financial conditions across a large part of the world move together, and the variable they move with is the dollar — its interest rate, its exchange rate, and the risk appetite of institutions funding themselves in it.

Why the dollar specifically:

• A large share of cross-border credit is denominated in dollars, including between parties with no US connection.
• Trade is invoiced in dollars well beyond trade with the United States, so the dollar price level affects margins globally.
• Global banks fund in dollars, so their capacity to lend anywhere depends on dollar funding conditions.

The consequence for policy independence is uncomfortable. A floating exchange rate is supposed to grant monetary independence — a country can set its own rate and let the currency absorb external shocks. The dollar cycle limits that: a depreciation that should be expansionary is contractionary when domestic borrowers hold dollar liabilities, because their balance sheets deteriorate as the currency falls.

That inverts the textbook adjustment mechanism, and it is the most consequential single point in this report. A country expecting depreciation to cushion a shock instead finds it amplifies one, and the amplification is proportional to how much dollar debt its private sector has accumulated — which is precisely what the preceding low-rate years encouraged. The Global Investment Outlook 2015 develops this into the dollar-debt burden framework the later archive relies on.

What an allocator could act on

Ask what currency the liability is in and what currency the revenue is in. A currency mismatch is not visible in a credit rating, a leverage ratio or a growth forecast. It is a separate question and it has to be asked separately.

Look for the exposure outside the official statistics. Emerging market corporate borrowing through offshore subsidiaries sat outside external debt measures. Where an exposure can be booked in a jurisdiction the statistics do not capture, the measured number is a floor.

Sort countries by external funding need, not by growth. The 2013 differentiation followed current-account deficits, portfolio-flow reliance and reserve coverage. Those are liability-structure characteristics, and they are what determines behaviour when funding conditions change.

Examine who holds the local-currency debt. Domestic issuance held by domestic institutions is resilient. The same market held by foreign investors is a hot-money position, and the two are indistinguishable on issuance data alone.

Expect exit to be repriced faster than entry was. Positions built gradually over years unwind on a change in expectation. Any position whose thesis depends on a policy path is exposed to the path being revised, not to it being executed.

Treat reserves as a scenario-removal device. Their value lies in making a funding failure implausible, not in being spent. Coverage relative to short-term obligations is the measure that matters, and the effect operates before any intervention occurs.

What 2013 established

• The currency mismatch as a general mechanism — the maturity mismatch of 2008 expressed across borders, invisible while the exchange rate is stable and total when it moves.
• Exit from unconventional policy is priced on announcement, while entry is absorbed over years — a structural asymmetry, not a communication failure.
• Markets sort on external funding need, so liability structure rather than economic performance determines who is punished.
• Local-currency issuance relocates the mismatch to the holder and converts default risk into exit risk; the composition of the holder base is what determines resilience.
• The dollar sets a global financial cycle that limits monetary independence, and can invert the sign of exchange-rate adjustment for economies with dollar liabilities.