A Template for Understanding Big Debt Crises

by Ray Dalio

Credit is the giving of buying power. This buying power is granted in exchange for a promise to pay it back, which is debt. Clearly, giving the ability to make purchases by providing credit is, in and of itself, a good thing, and not providing the power to buy and do good things can be a bad thing.

The problem with debt arises when there is an inability to pay it back. Said differently, the question of whether rapid credit/debt growth is a good or bad thing hinges on what that credit produces and how the debt is repaid (i.e., how the debt is serviced).

To give you an idea of what that might mean for an economy as a whole, really bad debt losses have been when roughly 40 percent of a loan’s value couldn’t be paid back. Those bad loans amount to about 20 percent of all the outstanding loans, so the losses are equal to about 8 percent of total debt. That total debt, in turn, is equal to about 200 percent of income (e.g., GDP), so the shortfall is roughly equal to 16 percent of GDP. If that cost is “socialized” (i.e., borne by the society as a whole via fiscal and/or monetary policies) and spread over 15 years, it would amount to about 1 percent per year, which is tolerable. Of course, if not spread out, the costs would be intolerable. For that reason, I am asserting that the downside risks of having a significant amount of debt depends a lot on the willingness and the ability of policy makers to spread out the losses arising from bad debts. I have seen this in all the cases I have lived through and studied. Whether policy makers can do this depends on two factors: 1) whether the debt is denominated in the currency that they control and 2) whether they have influence over how creditors and debtors behave with each other.

Economies whose growth is significantly supported by debt-financed building of fixed investments, real estate, and infrastructure are particularly susceptible to large cyclical swings because the fast rates of building those long-lived assets are not sustainable.

you need better housing and you build it, the incremental need to build more housing naturally declines. As spending on housing slows down, so does housing’s impact on growth. Let’s say you have been spending $10 million a year to build an office building (hiring workers, buying steel and concrete, etc.). When the building is finished, the spending will fall to $0 per year, as will the demand for workers and construction materials. From that point forward, growth, income, and the ability to service debt will depend on other demand. This type of cycle—where a strong growth upswing driven by debt-financed real estate, fixed investment, and infrastructure spending is followed by a downswing driven by a debt-challenged slowdown in demand—is very typical of emerging economies because they have so much building to do.

One classic warning sign that a bubble is coming is when an increasing amount of money is being borrowed to make debt service payments, which of course compounds the borrowers’ indebtedness.

Based on my examinations of them and the ways the levers available to policy makers work, I believe that it is possible for policy makers to manage them well in almost every case that the debts are denominated in a country’s own currency. That is because the flexibility that policy makers have allows them to spread out the harmful consequences in such ways that big debt problems aren’t really big problems. Most of the really terrible economic problems that debt crises have caused occurred before policy makers took steps to spread them

There are four types of levers that policy makers can pull to bring debt and debt service levels down relative to the income and cash flow levels that are required to service them: Austerity (i.e., spending less) Debt defaults/restructurings The central bank “printing money” and making purchases (or providing guarantees) Transfers of money and credit from those who have more than they need to those who have less

Typically debt crises occur because debt and debt service costs rise faster than the incomes that are needed to service them, causing a deleveraging.

But over time, each bottom and top of the cycle finishes with more economic activity than the previous cycle, and with more debt. Why? Because people push it—they have an inclination to borrow and spend more instead of paying back debt. It’s human nature. As a result, over long periods of time, debts rise faster than incomes. This creates the long-term debt cycle.

Remember that money serves two purposes: it is a medium of exchange and a store hold of wealth. And because it has two purposes, it serves two masters: 1) those who want to obtain it for “life’s necessities,” usually by working for it, and 2) those who have stored wealth tied to its value.

Throughout history these two groups have been called different things—e.g., the first group has been called workers, the proletariat, and “the have-nots,” and the second group has been called capitalists, investors, and “the haves.” For simplicity, we will call the first group proletariat-workers and the second group capitalists-investors. Proletariat-workers earn their money by selling their time and capitalists-investors earn their money by “lending” others the use of their money in exchange for either a) a promise to repay an amount of money that is greater than the loan (which is a debt instrument), or b) a piece of ownership in the business (which we call “equity” or “stocks”) or a piece of another asset (e.g., real estate).

Historians say that the problems that arose from credit creation were why usury (lending money for interest) was considered a sin in both Catholicism and Islam.2

During such times, most people think the assets are a fabulous treasure to own—and consider anyone who doesn’t own them to be missing out. As a result of this dynamic, all sorts of entities build up long positions. Large asset-liability mismatches increase in the forms of a) borrowing short-term to lend long-term, b) taking on liquid liabilities to invest in illiquid assets, and c) investing in riskier debt or other risky assets with money borrowed from others, and/or d) borrowing in one currency and lending in another, all to pick up a perceived spread. All the while, debts rise fast and debt service costs rise even faster. The charts below paint the picture.

In markets, when there’s a consensus, it gets priced in. This consensus is also typically believed to be a good rough picture of what’s to come, even though history has shown that the future is likely to turn out differently than expected. In other words, humans by nature (like most species) tend to move in crowds and weigh recent experience more heavily than is appropriate. In these ways, and because the consensus view is reflected in the price, extrapolation tends to occur.

The typical bubble sees leveraging up at an average rate of 20 to 25 percent of GDP over three years or so. The blue line depicts the arc of the long-term debt cycle in the form of the total debt of the economy divided by the total income of the economy as it passes through its various phases; the red line charts the total amount of debt service payments relative to the total amount of income.

GDP. To convey a few rough average numbers, below we show some key indications of what the archetypal bubble looks like:

pop. While I won’t go into exactly how it works here, the most defining characteristics of bubbles that can be measured are: Prices are high relative to traditional measures Prices are discounting future rapid price appreciation from these high levels There is broad bullish sentiment Purchases are being financed by high leverage Buyers have made exceptionally extended forward purchases (e.g., built inventory, contracted for supplies, etc.) to speculate or to protect themselves against future price gains New buyers (i.e., those who weren’t previously in the market) have entered the market Stimulative monetary policy threatens to inflate the bubble even more (and tight policy to cause its popping) As you can see in the

To anticipate a debt crisis well, one has to look at the specific debt-service abilities of the individual entities, which are lost in these averages. More specifically, a high level of debt or debt service to income is less problematic if the average is well distributed across the economy than if it is concentrated—especially if it is concentrated in key entities.

People tend to underestimate the size of this effect. In the early stages of a bubble bursting, when stock prices fall and earnings have not yet declined, people mistakenly judge the decline to be a buying opportunity and find stocks cheap in relation to both past earnings and expected earnings, failing to account for the amount of decline in earnings that is likely to result from what’s to come. But the reversal is self-reinforcing. As wealth falls first and incomes fall later, creditworthiness worsens, which constricts lending activity, which hurts spending and lowers investment rates while also making it less appealing to borrow to buy financial assets. This in turn worsens the fundamentals of the asset (e.g., the weaker economic activity leads corporate earnings to chronically disappoint), leading people to sell and driving down prices further. This has an accelerating downward impact on asset prices, income, and wealth.

As an illustration, when you own a $100,000 debt instrument, you presume that you will be able to exchange it for $100,000 in cash and, in turn, exchange the cash for $100,000 worth of goods and services. However since the ratio of financial assets to money is high, when a large number of people rush to convert their financial assets into money and buy goods and services in bad times, the central bank either has to provide the liquidity that’s needed by printing more money or allow a lot of defaults.

Some people mistakenly think that depressions are psychological: that investors move their money from riskier investments to safer ones (e.g., from stocks and high-yield lending to government bonds and cash) because they’re scared, and that the economy will be restored if they can only be coaxed into moving their money back into riskier investments. This is wrong for two reasons: First, contrary to popular belief, the deleveraging dynamic is not primarily psychological. It is mostly driven by the supply and demand of, and the relationships between, credit, money, and goods and services—though psychology of course also does have an effect, especially in regard to the various players’ liquidity positions. Still, if everyone went to sleep and woke up with no memory of what had happened, we would be in the same position, because debtors’ obligations to deliver money would be too large relative to the money they are taking in. The government would still be faced with the same choices that would have the same consequences, and so on.

For a write-down to be effective, it must be large enough to allow the debtor to service the restructured loan. If the write-down is 30 percent, then the creditor’s assets are reduced by that much. If that sounds like a lot, it’s actually much more. Since most lenders are leveraged (e.g., they borrow to buy assets), the impact of a 30 percent write-down on their net worth can be much greater. For example, the creditor who is leveraged 2:1 would experience a 60 percent decline in his net worth (i.e., their assets are twice their net worth, so the decline in asset value has twice the impact).7 Since banks are typically leveraged about 12:1 or 15:1, that picture is obviously devastating for them and for the economy as a whole.

Typically, increased taxation takes the form of greater income, property, and consumption taxes because these forms of taxation are the most effective at raising revenues. Wealth and inheritance taxes are sometimes also increased,9 though these typically raise very little money because so much wealth is illiquid that it is practically difficult to collect on, and forcing the taxpayer to sell liquid assets to make their tax payments undermines capital formation. Regardless, transfers rarely occur in amounts that contribute meaningfully to the deleveraging (unless there are “revolutions” and huge amounts of property are nationalized).

People ask if printing money will raise inflation. It won’t if it offsets falling credit and the deflationary forces are balanced with this reflationary force. That’s not a theory—it’s been repeatedly proven out in history. Remember, spending is what matters. A dollar of spending paid for with money has the same effect on prices as a dollar of spending paid for with credit. By “printing money,” the central bank can make up for the disappearance of credit with an increase in the amount of money. This “printing” takes the form of central bank purchases of government securities and nongovernment assets such as corporate securities, equities, and other assets, which is reflected in money growing at an extremely fast rate at the same time as credit and real economic activity are contracting. Traditional economists see that as the velocity of money declining, but it’s nothing of the sort. What is happening at such times is that credit destruction is being offset by money creation. If the balance between replacing credit and actively stimulating the economy is right, this isn’t inflationary.

If a currency falls in relation to another currency at a rate that is greater than the currency’s interest rate, the holder of the debt in the weakening currency will lose money. If investors expect that weakness to continue without being compensated with higher interest rates, a dangerous currency dynamic will develop. That last dynamic, i.e., the currency dynamic, is what produces inflationary depressions. Holders of debt denominated in the poorly returning currency are motivated to sell it and move their assets into another currency or a non-currency store hold of wealth like gold. When there is a debt crisis and economic weakness in a country, it is typically impossible for the central bank to raise interest rates enough to compensate for the currency weakness, so the money leaves that country and currency for safer countries. When so much money leaves the country that lending dries up, the central bank is faced with the choice of letting the credit markets tighten or printing money, which produces a lot of

But a lot depends on politics. If the markets are allowed to run their courses, the adjustments eventually take place and the problems are resolved, but if the politics get so bad that productivity is thrown into a self-reinforcing downward spiral, that spiral can go on for a long time.

While inflationary depressions are possible in all countries/currencies, they are far more likely in countries that: Don’t have a reserve currency (so there is not a global bias to hold their currency/debt as a store hold of wealth) Have low foreign-exchange reserves (the cushion to protect against capital outflows is small) Have a large foreign debt (so there is a vulnerability to the cost of the debt rising via increases in either interest rates or the value of the currency the debtor has to deliver, or a shortage of the availability of dollar denominated credit) Have a large and increasing budget and/or current account deficit (causing the need to borrow or print money to fund the deficits) Have negative real interest rates (i.e., interest rates that are significantly less than inflation rates), therefore inadequately compensating lenders for holding the currency/debt Have a history of high inflation and negative total returns in the currency (increasing lack of trust in the value of the currency/debt) Generally speaking, the greater

The positive, self-reinforcing cycle is enhanced when the demand for the currency is improving. If the currency is cheap enough to offer attractive opportunities to foreign investors (who will typically lend to or invest in entities that can produce inexpensively in that country and sell into export markets to earn the foreign currency to provide them with a good return), and/or the country sells more to foreigners than it buys from them, a country’s balance of payments will become favorable—i.e., the demand for its currency will be greater than its supply. This makes the central bank’s job easier—i.e., it can get more growth per unit of inflation—because the positive inflows can be used to appreciate the currency, to lower interest rates, and/or to increase reserves, depending on how the central bank chooses to handle

2) The Bubble

Foreign capital flows are high (on average around 10 percent of GDP) The central bank is accumulating foreign-exchange reserves The real FX is bid up and becomes overvalued on a purchasing power parity (PPP) basis by around 15 percent Stocks rally (on average by over 20 percent for several years into their peak)

Debt burdens rise fast. Debt to GDP rises at an annual rate of about 10 percent over three years. Foreign-currency debt rises (on average to around 35 percent of total debt and to around 45 percent of GDP). Typically, the level of economic activity (i.e., the GDP gap) is very strong and growth is well above potential, leading to tight capacity (as reflected in a GDP gap of around +4 percent). The charts below convey what

During the bubble, the gap between the country’s income and its spending widens. The country requires an increasing inflow of capital to drive continued growth in spending. But levels of economic activity can remain strong at the top of the cycle only as long as continued inflows, motivated by expectations of continued high growth, drive up asset prices and cause the currency to strengthen further. At this point, the country is increasingly fragile and even a minor event can trigger a reversal.

3) The Top and Currency Defense The top-reversal/currency-defense occurs when the bubble bursts—i.e., when the flows that caused the bubble and the high prices of the currency level, the high asset prices and the high debt growth rates finally become unsustainable. This sets in motion a mirror-opposite cycle from what we saw in the upswing, in which weakening capital inflows and weakening asset prices cause deteriorating economic conditions, which in turn cause capital flows and asset prices to weaken further. This spiral sends the country into a balance of payments crisis and an inflationary depression. Because at the top people are so invested in the optimistic scenario, and because that optimism is reflected in the prices, even a minor event can trigger a slowing of foreign capital inflows and an increase in domestic capital outflows. Though worsening trade balances typically play a role (usually because of the high currency level and excessive domestic consumption that led to high imports), adverse shifts in capital flows are usually more important.

In the typical cycle, the crisis arises because the unsustainable pace of capital that drove the bubble slows, but in many cases, there is some sort of a shock (like a decline in oil prices for an oil producer). Generally the causes of the top-reversal fall into a few categories: The income from selling goods and services to foreigners drops (e.g., the currency has risen to a point where it’s made the country’s exports expensive; commodity-exporting countries may suffer from a fall in commodity prices). The costs of items bought from abroad or the cost of borrowing rises. Declines in capital flows coming into the country (e.g., foreign investors reduce their net lending or net investment into the country). This occurs because: The unsustainable pace naturally slows, Something leads to greater worries about economic or political conditions, or A tightening of monetary policy in the local currency and/or in the currency those debts are denominated in (or in some cases, tightening abroad creates pressure for foreign capital to pull out of the country). A country’s own citizens or companies want to get their money out of their country/currency.

Weakening capital flows are often the first shoe to drop in a balance of payments crisis. They directly cause growth to weaken because the investment and consumption they had been financing is reduced. This makes domestic borrowers seem less creditworthy, which makes foreigners less willing to lend and provide capital. So, the weakening is self-reinforcing. Growth slows relative to potential as the pace of capital inflows slows. Domestic capital outflows pick up a bit. Export earnings fall, due to falling prices or falling quantities sold. Typically exports are flat, no longer rising.

At this stage, central banks typically try to defend their currencies by a) filling the balance of payments deficit by spending down reserves and/or b) raising rates. These currency defenses and managed currency declines rarely work because the selling of reserves and/or the raising of interest rates creates more of an opportunity for sellers, while it doesn’t move the currencies and interest rates to the levels that they need to be to bring about sustainable economic conditions. Let’s look at this typical defense and why it fails. There is a critical relationship between a) the interest rate difference and b) the spot/forward currency relationship. The amount the currency is expected to decline is priced into how much less the forward price is below the spot price. For example, if the market expects the currency to fall by 5 percent over a year, it will need that currency to yield a 5 percent higher interest rate. The math is even starker when depreciation is expected over short periods of time. If the market expects a 5 percent depreciation over a month, than it will need that currency to yield a 5 percent higher interest rate over that month—and a 5 percent monthly interest is equivalent to an annual interest rate of about 80 percent10 —a level that’s likely to produce a very severe economic contraction in an already weak economy. Because a small expected currency depreciation (say 5 to 10 percent in a year) would equal a large interest rate premium (5 to 10 percent per...

Central banks should not defend their currencies to the point of letting their reserves get too low or their interest rates too high relative to what is good for the economy because the dangers those conditions pose are greater than the dangers of devaluation. In fact, devaluations are stimulative for the economy and markets, which is helpful during the economic contraction. The currency decline tends to cause assets to rise in value measured in that weakened currency, stimulate export sales, and help the balance of payments adjustment by bringing spending back in line with income. It also lowers imports (by making them more expensive), which favors domestic producers, makes assets in that currency more competitively priced and attractive, creates better profit margins for exported goods, and sets the stage for the country to earn more income from abroad (through cheaper and more competitive exports).

As you can see in the chart below, returns to holding the currency for foreigners start out negative, but then rally about a year after the devaluation.

Here is what we typically see when the country reaches the bottom: The collapse in imports improves the current account a lot (on average by about 8 percent of GDP). Capital inflows stop declining and stabilize. Capital flight abates. Frequently, the country turns to the IMF or other international entities for support and a stable source of capital, especially when its reserves are limited. Short rates start to come down after about a year, but long rates continue to stay relatively elevated. After peaking, short rates fall back to their pre-crisis levels in around two years. The decline in short rates is stimulative. As interest rates come down, the forward currency price rallies relative to the spot. As the currency stabilizes, inflation comes down. Usually it takes nearly two years after the bottom for inflation to reach pre-crisis levels.

Classically, the collapse in imports brings the current account into a surplus of 2 percent of GDP, rising from a deficit of -6 percent of GDP about 18 months into the crisis. In the earlier stages of the crisis exports play a smaller role; they actually tend to contract during the worst of the crisis (as other countries are sometimes seeing economic slowdowns too). They rebound in the subsequent years.

When 1) within countries there are economic conflicts between the rich/capitalist/political right and the poor/proletariat/political left that lead to conflicts that result in populist, autocratic, nationalistic, and militaristic leaders coming to power, while at the same time, 2) between countries there are conflicts arising among comparably strong economic and military powers, the relationships between economics and politics become especially intertwined—and the probabilities of disruptive conflicts (e.g., wars) become much higher than normal.

While these big debt crises can be devastating to some people and countries over the short- to medium-term (meaning three to ten years), in the long run they fade in importance relative to productivity, which is more forceful (though less apparent because it is less volatile). The political consequences (e.g., increases in populism) that result from these crises can be much more consequential than the debt crises themselves.

public. As debt monetization is inflationary (there is more money in the economy chasing the same quantity of goods and services), a self-reinforcing spiral ensued—i.e., debt monetization increased inflation, which reduced real interest rates, which discouraged lending to the government, which encouraged additional debt monetization.

while inflationary depressions are possible in all countries/currencies, they are most common in countries that: Don’t have a reserve currency: So there is not a global bias to hold their currency/debt as a store hold of wealth Have low foreign exchange reserves: So there is not much of a cushion to protect against capital outflows Have a large stock of foreign debt: So there is a vulnerability to the cost of debt rising via increases in either interest rates or the value of the currency the debtor has to deliver, or a shortage of available credit denominated in that currency Have a large and increasing budget and/or current account deficit: So there is a need to borrow or print money to fund the deficits Have negative real interest rates: So lenders are not adequately compensated for holding the currency/debt Have a history of high inflation and negative total returns in the currency: So there is a lack of trust in the value of the currency/debt By the end of the war, the German

To be clear: at this point money printing was not the source of the currency weakness so much as currency weakness was the cause of money printing. In other words, capital flight from the currency and the country was driving the currency down, which in turn helped drive higher inflation. That’s classically how inflationary depressions happen.

“The fall of the mark...has brought a real anxiety among the propertied classes. Everyone seeks to do something with their money. Everything is bought that can be bought, not only for present need, not only for future use, but in order to get rid of the paper and have objects to exchange when the time comes that it is worth absolutely nothing.”68

In inflationary depressions, it is classic that with each round of printing, more money leaves the currency instead of going into economic activity. As domestic currency holders see that investors that short

“The prices in the shops change every hour. No one knows what this week’s wages will buy at the end of the week. The mark is at the same time valueless and scarce. On the one hand, the shops do not want to receive marks, and some of them are unwilling to sell at any price at all. On the other hand…the banks were so short of ready cash that the Reichsbank advised them to cash no checks for more than 10,000 marks…and some of the biggest institutions were unable to cash their customers’ checks for payment of weekly wages.”95

People tend to think that hyperinflations are caused by central banks recklessly printing too much money, and all they need to do to stop it is to turn off the printing press. If it were that easy, hyperinflations would almost never occur! Instead, inflation spirals push policy makers into circumstances where printing is the least bad of several terrible options. In the case of Weimar Germany, the cost of not printing was not only potential economic collapse, but political fragmentation. France’s repeated threats to occupy German territory if reparations were not paid made halting the printing press an invitation to a foreign invasion. It also lowered hopes for productive reparations negotiations. As one prominent industrialist put it at the time: