Read Rogoff and Reinhart

by Brad Setser
January 9, 2008

Let me second Martin Wolf. Ken Rogoff and Carmen Reinhart’s recent paper is brilliant. They also have mastered the art of saying a lot in a little space. There is no excuse not to read it in full.

Rogoff and Reinhart compare the crisis in the US banking system — and in the shadow banking system, which turned out to be rather linked to the real banking system — triggered by the subprime crisis to other recent crises in advanced economies. They find "stunning qualitative and quantitative parallels across a number of standard financial crisis indicators."

Their figure 1 — which plots the increase in real US housing prices against the increase in real housing prices that preceded other severe bank-centered financial crises in industrial economies (Spain in the 70s, Japan and the Scadanavian countries in the early 90s) — is stunning. Rogoff and Reinhart note that "The United States looks like the archetypical crisis country, only more so" before concluding, in Rogoff and Reinhart conclude, in fairly Roubini-esque fashion, that:

"The United States should consider itself quite fortunate if its downturn ends up being a relative short and mild one."

In one somewhat surprising way, the US is in a better position than the other advanced economies that got into severe trouble: the recent increase in public debt was actually a bit smaller than the increase in other countries. The US current account deficit, by contrast, is much larger than in the other countries — a source of concern, if the associated risks haven’t really been realized.

I particularly enjoyed the parallel they drew between today’s world and the world during the 1970s oil shock. They write:

"During the 1970s, the U.S. banking system stood as an intermediary between oil exporting surpluses and emerging market borrowers in Latin America and elsewhere. While much praised at the time, the 1970s petro-dollar recycling ultimately led to the 1980s debt crisis, which in turn placed enormous strain on the money center banks. It is true that this time, a large volume of petro-dollar are again flowing into the United States, but many emerging market economies have been running current account surpluses, lending rather than borrowing. Instead, a large chunk of money has effectively been recycled to a developing economy that exists within the United States own borders. Over a trillion dollars was channeled into the sub-prime mortgage market, which is comprised of the poorest and least credit worth[y] borrowers within the United States."

Clever. It also sounds right to me.

Two additional points are worth mentioning.

First, I suspect there is a real risk that the European banking system has been doing much the same thing as the US banking system, though on a smaller scale. An awful lot of money — effectively recycled petroeuros along with Chine’s burgeoning euro reserves — has been channeled through Europe’s banking system to the "poorest" borrowers within the European Union. Eastern Europe in particular. With luck, though, these borrowers will prove to be among the most credit-worthy borrowers in Europe, on the back of a convergence story. Eastern Europe is growing more rapidly than the rest of Europe. Nonetheless, a disruption in the internal recycling of external flows to the European Union remains a potential risk.

Second, emerging market central banks and sovereign funds have effectively "stood as an intermediary" between their own current account surpluses and capital flows to the US and Europe. Almost as importantly, they intermediated between private inflows into their own economy and the United States need for funds. Private capital is flowing into emerging Asian economies — India as well as China. Private capital is also now flowing into the Gulf. The net outflow from these surplus regions comes entirely from the "official" sector. Dollars flowing into China and other emerging economies with surpluses are effectively transformed into demand for US bonds (And now US banks) by emerging market central banks.

The scale of this intermediation is even larger that the $1 trillion that flowed into the US subprime market. That is a cumulative sum. Emerging market central banks are adding over a trillion to their reserves a year.

The largely but not exclusively American "intermediaries" that took the credit risk associated with channeling petro-dollar and sino-dollar inflows into the "developing economy that exists within the United States own borders" are currently taking large losses. The emerging market financial intermediaries that channeled the oil savings surplus, the Chinese savings surplus and large private inflows to the US — and to a lesser degree Europe — are also likely to take losses. Not from credit risk but from currency risk. By and large, though, the currency losses have yet to be realized.

Once those losses are factored in, the total losses associated with the recent mis-allocation of global savings into US real estate are likely to be even bigger — and a bit more dispersed than the credit losses.

Post a Comment29 Comments

  • Posted by Macro Man

    What struck me about the charts was the virtually identical GDP trajectory that the US has taken when compared to the average crisis as opposed to the “big 5.” The paper seemed rather taken with the current account situation in the US as an outlier, but appeared dismissive of the notion that mitigating circumstances might apply. Perhaps BWII isn’t any different from Spain two years after Franco…but common sense suggests that it probably is.

  • Posted by Guest

    The other thing that distinguishes the US from some of these other economies is that it is experiencing population growth both natural and through illegal and legal migration.

  • Posted by Twofish

    Question: Can the rerun the graphs to show how the crisises look in relation to each other? The comparison is between the US and the “average crisis” but this doesn’t mean much if the crises end up markedly different from each other.

    One reason I’m wondering is that the shape of average for all banking crises is very different from the average for the big 5 and there is the implication that the higher the peak the sharper the fall and I am interested if that is the case if you plot all data.

    Also Figure 2 shows that the US is already following a different trajectory, and that might suggest that the fed is going a good thing by stimulus.

  • Posted by Dave Chiang

    Federal Reserve Chairman Bernanke advocates making unlimited credit available to solve “liquidity” issues when it is really the intrinsic value of financial assets and the economic unsustainability of a “credit bubble” that is at issue. The fiat US currency is becoming increasingly worthless and this loss of trust in the currency will result in the real economy ceasing to properly and normally function. The massive Fed injections of liquidity will destroy the US currency in addition to not addressing the underlying economic issues.

  • Posted by bsetser

    macroman — true. the US “boom” was also weaker though. the risk is that the US, with a lag, experiences the same kind of nasty recession some of the others experienced. i.e. given the nature of the boom in housing prices (which didn’t translate into a super-strong boom, perhaps b/c it was working against the tailwind from the equity bubble), there is a risk of steep fall in output. hence the current emphasis in macro circles on taking proactive steps to insure against this outcome.

    others of course take the view that trying to smooth out the path of output and asset prices is a mistake, and it would be better not to stimulate to purge the system of its excesses (and in the process produce a quicker return to a lower equilibrium trade deficit by pushing up national savings/ slowing consumption growth).

    guest — rogoff and reinhart plotted GDP growth per capita, presumably precisely for this reason.

  • Posted by Anonymous

    I don’t recall you associating US bank credit excesses so very closely with offical inflows into US assets before. Of course, the official inflows aren’t going directly into bank liabilities. Is it really evident that official flows into treasuries etc. is the true cause of excess credit creation in the US?

  • Posted by Anonymous

    There are at least two characteristics of this crisis that are different at least in degree from its predecessors:

    a) The banking system, including the shadow system, is generally much more distintegrated. I suspect most of the other banking crises have been focused on “universal banks”. There is probably less concentration of the total risk in those type of banks today – Citi and UBS being good examples of such “universal banks”. With the increased complexity of financial product origination and distribution, the total risk is spread over many more specialized banks (including the major New York nonintegrated dealers) than in previous crises. While not diminishing the enormity of the “lesser developed housing” country risk within the US, this complexity, albeit with the price discovery challenges associated with its opacity, is still a risk diffusion outlet.

    b) The same argument holds for the globalization of risk absorption, given US subprime risk that ends up in a Tokyo portfolio, for example.

  • Posted by Stormy

    Anonymous 2008-01-09 13:16 asks a good question, at least for me. I know that the equation has always been implicit, i.e., official inflows, along with other factors, feed the credit excesses. You have made the connection quite often, I think, irrespective of Anonymous’ assertion.

    Yet the connection has always puzzled me. I think I can easily see how official inflows sustained the account deficit: A treasury bought goes into the official coffers. How that purchase makes its way to a lender….I just get lost.

  • Posted by Anonymous

    Stormy, I said ‘so very closely’ – today’s piece seems to be pushing the envelope more explicitly on the general theme which I agree has often been stated

  • Posted by mheck82

    Brad,
    I’m just wondering, if there is a special reason for that you are so interested in eastern Europe.
    The paper from Reinhart and Rogoff in end mentioned as well UK, Spain and Ireland as the European countries where such a crisis might occur (as I suggested on your outlook piece). Spain already has shown first signs of stress.
    And as they are more developed already they have less capability to overcome this simply by high convergence growth rates as eastern Europe countries may.

  • Posted by Emmanuel

    (1) You seem to have gotten into an EU-bashing mode as of late ;-) Why will EM investments in Europe eventually lead to FX losses? It seems to me that those wise enough to avoid Sammy’s little green depreciation machines have done well enough. EUR/CNY went up in 2007, more so the euro against the oilers’ currencies which is the same banana as EUR/USD. Among the major reserve currencies–USD, JPY, GBP, EUR–the latter has help up the best. Lumping the Euro in the same class with the ghastly dollar is very unfair. Yes, Sammy is a ripoff artist, but the Euro would have to take a pretty big dive for its holders to take large losses.

    (2) This is my pet peeve, but let’s not be so quick to conclude that the AAA debt rating for Treasuries is warranted, let alone a sign that Sammy’s IOUs are credit “risk-free”. This is a cheap shot at Sammy’s dinky wares, but wasn’t a lot of subprime also certified AAA? Treasuries = public sector subprime may not be such a far-fetched comparison. We’ll see.

    (3) I’m not so sure that Eastern Europe 2008 = East Asia 1997. Yes, many Eastern European countries are running fairly hefty external deficits. But, a lot of these guys are in the EMU or are soon to be in the EMU. So, currency risk is less of an issue. Also, the kind of investments made by Western Europe in Eastern Europe are tangible, long-term FDI such as with auto manufacturing moving eastwards instead of flighty hot money.

    And, of course, many of these Eastern European states have been running CADs for over a decade, so why become more concerned now? Don’t cry for me Herzegovina…

  • Posted by gillies

    not difficult to sum up the current irish property investor -
    wylie o’coyote.

    not difficult to sum up the u s election. the candidates have raised millions of dollars and promised the voters ‘change’ -
    the loose change ?

    not difficult to quantify risk associated with debt – it is an unknown known. (the debt is the known – and even after you compute a mandlebrot set of calculations on the known, the unknown factor in the equation remains.) the repayment is the unknown.

    buying in to lending is like bungee jumping off a wall street skyscraper of unknown height.

    meanwhile behind all of the complexities there is still only one way to make money out of sheep – pull the wool over their eyes.
    .

  • Posted by Stormy

    sorry, anonymous. I did not mean to mistate you.

  • Posted by Dave Chiang

    “According to statistics released by the IMF, the euro’s share of known foreign exchange holdings rose to 26.4 percent in the third quarter of 2007, up from 25.5 percent in the previous three months and 24.4 percent in the third quarter of 2006. Meanwhile, the US dollar’s share of known foreign exchange reserves for the third quarter of 2007 was 63.8 percent, down 2.7 percent from the third quarter in 2006.”

  • Posted by Gabor

    Emmanuel, Eastern European (ex Soviet block) countries are in the EMU – but their currency is in a 3o% wide band and usually on the strong side, so those currencies can depreciate 3o%, or more, if central banks cannot defend them at the weak side of the band.

    There are two main ways how money is flowing into these economies:
    1. portfolio investment into T bonds denominated in local currencies – in this case the investor holds the FX risk
    2. FX loans denominated in EUR, SWF and JPY – in this case the borrower holds the FX risk
    CAB is negative mainly because interest and dividend payments from earlier investments.
    Public and private consumption financed from inflows is only one leg of the growth story. Export led industrial growth is pretty robust here (almost as good as China). In the last months however industrial growth is losing steam.
    Gabor from Hungary

  • Posted by mheck82

    Dresdener Bank in a research paper from last year has looked as well on the eastern Europe economies and said, much of the deficit is due to FDI, not consumption. Highest risk is for Latvia, a roughly 2 million people country.
    In general they are rather conservative, e.g. had no off-sheet investement as so many other banks.

  • Posted by bsetser

    i don’t think I am negative on europe. I think Americans haven’t given europe enough credit for euroep’s strong recent growth during an appreciation, something that has done more than anything else to bring the us current account deficit down. i also generally applaud the fact that capital flows from rich to poor inside europe’s institutional structure.

    I am though starting to get a bit worried about the scale of the current account deficits in eastern europe — perhaps because i read the IMF’s paper on southeastern europe and some of danske banks’ material on the baltics. plus, i generally worry about current account deficits above 10% of GDP. I don’t think it is all investment, unless investment in housing stock counts as investment like investment in export-industries — and in any case, it was all “investment” in emerging asia before 97. but yes, I may be a bit too alarmed.

    i do tho worry that the deficits — while generally good — haven’t gotten too big, and that some eastern european currencies are getting strong v the eur at a time when the euro is strong globally.

  • Posted by bsetser

    on the connection between Asia and the oil exporters savings surplus and subprime –

    Here is how i would explain it, and how i think rogoff would explain it. He is a very classically trained, balance of payments focused international economists.

    A rise in savings v investment in one part of the world has to be offset by a fall in savings v investment somewhere else.

    we know that savings rose faster than investment in china and the oil exporters, creating surplus savings that was exported. the offsetting fall in savings v investment (a rise in the current account deficit) has large been found in the US, Australia, britain and parts of eastern europe (i am leaving the eurozone out b/c it is in overall balance and has stayed there). and the fall in savings v investment in the US has come from the household sector. savings fell, allowing more consumption for a given level of income. and residential investment boomed. in that sense, the savings surplus in asia and the oil states financed the “subprime boom’ — whether the rise in home equity withdrawal that facilitated the fall in savings or the surge in investment in real estate.

    this didn’t happen directly — the oil states and the PBoC didn’t directly buy a lot of repackaged subprime debt. But in aggregate, that is what happened. and the mechanism was probably something like this –

    the surge official outflow from asia and the oil exporters led to a surge in demand for US debt and specifically “safe” treasuries and agencies. that pushed yields on these instruments down. US investors didn’t like the yields and sold, often to foreign central banks. they then invested their funds in higher yielding instruments — often MBS or CDOs that included repackaged subprime debt. the flat yield curve also meant that banks could not make money simply by playing the yield curve. they had to take on credit risk. and they did, in a big way, through their SIVS and conduits. and the dollars on deposit in london and switzerland from the conservative oil states were also lent out to folks who used the leverage to buy higher yielding assets. the net result was an increase in demand for subprime mortgages.

    and since rising house prices — fueled by lower long-term rates as well as finacial innovation that made credit available to more people (including to people whose income wasn’t sufficient to afford as big a home as they wanted according to traditional metrics) — spawned expectations that home prices would continue rise, the process developed a bit of momentum.

    those at least are the connections that i see.

  • Posted by mheck82

    brad
    Have you a suggestion what would have been an appropriate politics answer to that, which would have not lead into such mess?

    I would have suggested for the US, to invest massively. If neccessary the gov could have done that in railroad, renewable energies….
    To prevent people not to take too high loans on housing and consumption, high taxes could have been used. Regarding housing as well very high standards regarding insulation + use of thermic solar…
    If this would have been done, the Gov could cut taxes in case of a crisis as now and people would not have to pay so much for heating/warm water despite higher oil costs with a falling dollar. Both would mean, that those who could have afforded the loan in times of high house price increases and high taxes, could afford it now as well.

    By the way, I don’t think at all that you are negative on Europe.

  • Posted by 50 Cent

    bsetset: “US investors didn’t like the yields and sold, often to foreign central banks. they then invested their funds in higher yielding instruments — often MBS or CDOs that included repackaged subprime debt.”

    No doubt this mechanism is mostly correct. But we do know that it was not only US investors who bought the toxic MBS/CDO securities. Do we actually know for a fact that China and the oil states didn’t buy this stuff? It may be true that they “mostly” bought treasuries and agencies, but even a small fraction of a $1.2T portfolio is tens of billions…

  • Posted by Anonymous

    re – ‘classical’ approach to explanation of connection between Asia/oil surplus and subprime – is there such a thing as a monetary or ‘monetarist’ approach to balance of payments that would describe what happens in a different way?

  • Posted by adiemuso

    If I might add on to Brad’s point.

    Asia has a painful memory of the 1997 crisis, it seems that many western analysts overlook this, thus it prefers building a safe portfolio on its reserves. Thus that might explains why “safe” T-Bonds/Bills/Agencies were bought despite the “unappealing” yields.

    Since 1990s, we have seen an awful lot of “new” financial engineered products being “invented” and sold as “safe” investment vehicles to those traditionally long “safe” T-Bons/Bills/Agencies funds (especially in the West). I cannot argue against the fact that Asian/Oil Exporters might have bought some, but the ratio is relatively small in comparison to the West. (Not tryin to start a East/West divide here)

    And now, I believe we are seeing an unwinding of such risk appetites in the East/West. East selling their “safe” T-bonds/bills/agencies and buying the “riskier” assets from the West. And those holders of now “toxic” structured/engineered products reverting back to “safe” traditionals.

    Perhaps someone can come up with a quantitative study to show a clearer picture?

  • Posted by Judy Yeo

    Brad; connections comment

    In a big (ok, very big) picture kind of way, sure on the equations part, a rise in savings and investment in Asia “had to be” balanced by a decline in savings and investment in USA, UK, Australia etc,but that equation works as a backward check test, not as a cause effect flow. You cannot force people to increase consumption or deplete savings. Japan in the late 90s and early 2000s is a prime example. For lack of a better word, it is greed that causes such phenomena.

    The same for investment, risk preferences and the chase for higher returns are never enforced, they are a product of human greed. Not going to go further on this ‘cos don’t want to repeat my blog post. Just to conclude, one has free will, face the consequences of your choices when you have the freedom to make those choices.

  • Posted by bsetser

    Judy — true enough. Americans seem to be a bit more willing to borrow in response to lower equilibrium interest rates than others, so the us ended up absorbing the bulk (from 03-05) of the increase global savings. that is a bit less true for 06 and 07. a full explanation would need to explain why the US ended up borrowing so much. a preference for $ on the part of US creditors is part of it. a greater responsiveness to lower rates is part of it too.

    anonymous — do we know for a fact that China/ the oil states didn’t buy this stuff. the answer is no. China and the oil states don’t disclose enough to be able to tell, and the us data has enough gaps that it doesn’t rule out much of anything. especially with respect to CDOs — a CDO might be legally based in the caribbean or europe, so the us data provides no information on who buys its various tranches.

    that said, I am fairly confident that China’s exposure is in the tens of billions (mostly with the banks) not the hundreds of billions. the fact that so much of china’s holdings are known to be in treasuries and agencies sets a limit on their exposure to the more toxic stuff.

    as for the gulf, all the evidence is anecdotal, since next to zero gulf flows appear in the us data and the UK and the swiss have not opted to provide detailed data on flows in and out of their respective economies. the absence of better data out of the uk — given all the talk in the UK about transparency — ought to be a scandal. (and yes, i understand fully why there is a data hole; i am naive, but not that naive … )

    but the anecdotes suggest that the gulf likes property (of the big office building kind), that the gulf likes equity and private equity, and the gulf likes fairly safe and secure treasuries. conversely, the gulf supposedly hasn’t been a big player in the more exotic bits of the debt market.

    if someone has good evidence to contrary, i am all ears. i am uncomfortable with the absence of hard date to back my inferences about GCC demand.

    We know that Russia has been a big buyer of agencies, and the safe agencies — there isn’t much space left over for more risky assets, and the bank of russia’s guidelines don’t really allow it to be too aggressive. some other oil central banks seemingly still have large deposits — not securities of any kind.

    again — if someone has data or anecdotes to the contrary, do talk!

  • Posted by Stormy

    Anonymous,

    At the risk of being foolish, uninformed and certainly untechnical, I see the connection in this way:

    Four things initially stimulated the housing bubble:

    1. Lower interest rates
    2. “Innovative” packaging with little regulatory oversight
    3. Lower taxation of the wealthy.
    4. Enormous profits within the multinationals and the financial community as they moved to and invested in developing countries. Some of those profits went to repurchase stock—but in the top tiers, salaries escalated rapidly. A great deal of that money went into bigger houses, plush mansions, etc.

    In real estate, everything is location. If a McMansion goes up near you, then your property value rises. A great deal of money was being made—and spent. These purchases were like stones thrown into a pond. Property values escalated. With low interest rates, those who shouldn’t have sold and resold. (I, for example, was able to sell my home at twice the value it had a year earlier. I lived, at the time, on the Eastern Shore. People from Washington were moving in, flush with cash. Bidding wars ensued…and this was at the very beginning of the bubble!) Smartly enough, I took the money and moved to rural Canada.

    Official inflows greatly exacerbated the bubble, fueling it indirectly.

    Official inflows of cash from developing countries—purchase of Treasuries, etc.—allowed the U.S. to avoid meeting its debt obligations. Instead of using taxes to pay our way, we used credit, credit in the form of those Treasury purchases. If we had kept taxes higher, then the bubble would not have been as dangerous. Taxes take money from the taxpayer so that the government can pay its bills. With less money in circulation, the bubble would have been more subdued. (In one way, you could say that China financed the Iraq war.)

    GDP of course rose. But GDP is only a number, hiding, sometimes, a lot of rather nasty surprises.

    I tend to be a bit bizarre, I suspect, in the way I put stuff together.

  • Posted by Guest

    http://www.guardian.co.uk/business/2008/jan/10/northernrock.goldmansachs

    The stupidity of some western corporations has put them at the “disposal” of wiser people than they. They richly deserve this fate.

  • Posted by Anonymous

    Judy Yeo – you’ve made a subtle point, which may be the most important point of all

    Stormy – interesting; also bizarre – I found the Rogoff paper somewhat underwhelming. The idea of a historical correlation between asset prices and banking problems is not explosive. The how is more interesting.

  • Posted by Anonymous

    Your pop-up is very annoying. You would think that once is enough, but instead it seems to pop up every 30 seconds.

  • Posted by TRM

    Demographics demographics demo you get it, or not. One needs to fill space for $1.61 inc FL. tax, but I don’t.

Bad Behavior has blocked 6578 access attempts in the last 7 days.