I highly recommend Anton Brender and Florence Pisani’s recent monograph, "Globalized finance and its collapse." In a lot of ways, it is something that I wish I could have written. I don’t agree with every detail, but in my view they get the broad story right.
Brender and Pisani both teach at Paris-Dauphine. But Anglo-Saxon chauvinism shouldn’t get in the way of appreciating quality work. And in this case, there is no excuse: the translation (by Francis Wells) is superb.
In some deep sense, Brender and Pisani have updated the core arguments of Martin Wolf’s Fixing Global Finance (also highly recommended) to reflect many of the things what we all have learned in the last nine months of crisis.
Like Martin Wolf, Brender and Pisani recognize that globalization took an unusual turn over the past several years: the globalization of finance resulted in a world where the poor financed the rich, not one where the rich financed the poor. And what’s more, this “uphill” flow was essentially a government flow. Despite the talk of the triumph of private markets over the state a few years back, the capital flow that defined the world’s true financial architecture over the past several years was the result of the enormous accumulation of foreign exchange reserves in the hands of the central banks of key Asian and oil-exporting economies.
Dooley and Garber recognize this. They don’t pretend that private investors in the emerging world drove the uphill flow of capital. But Dooley and Garber also assert that there is no connection between this uphill flow and the current crisis. In a March Vox EU piece they wrote:
"We have argued that the decisions of governments of emerging markets to place an unusually large share of domestic savings in US assets depressed real interest rates in the US and elsewhere in financial markets closely integrated with the US ... Low risk-free real interest rates that were expected to persist for a long time, in the absence of a downturn, generated equilibrium asset prices that appeared high by historical standards. These equilibrium prices looked like bubbles to those who expected real interest rates and asset prices to return to historical norms in the near future ... Along with our critics, we recognised that if we were wrong about the durability of the Bretton Woods II system and the associated durability of low real interest rates, the decline in asset prices would be spectacular and very negative for financial stability and economic activity ... This is not the crisis that actually hit the global system. But the idea that an excessive compression of spreads and increased leverage were directly caused by low real interest rates seems to us entirely without foundation." Emphasis added.
It actually isn’t that hard to find examples of how low returns on "safe" investments induced more risk-taking throughout the system, especially private intermediaries started to believe in the essential stability of Bretton Woods 2. The Wall Street Journal recently reported that many bond funds underperformed their index in 2008 because their managers had been taking on more risk to juice returns in the good years, and thus went into the crisis underweight the safe assets that central banks typically hold. US money market funds that lend ever-growing sums to European commercial banks were making a similar bet. As were the European banks that relied on wholesale funding to cover their growing portfolios of risky dollar debt. The inverted yield curve forced vehicles that borrow short and lend long to either go out of business or take ever more credit risk -- and as volatility fell and spreads compressed, there was a constant temptation to take on more leverage to keep profits up. The pressures to take more risk were there.
Brender and Pisani document clearly how the process ended up working.
They recognize that the current crisis could not have happened in the absence of an accumulation of credit risk by private financial intermediaries (big banks, broker-dealers and the “vehicles” that operated in the shadows) in the US and Europe. But they also recognize that the accumulation of credit risk by private financial intermediaries would not have been possible if emerging market governments hadn’t been so willing to accumulate exchange rate risk.
They consequently highlight the impossibility of assigning blame for the current crisis solely to the financial sector (and their regulators) in the West or the central banks of the East. Both ultimately were responsible, just in slightly different ways.
Brender and Pisani write:
“The emerging region’s savings surplus could not have been built up if there had not been a counterpart in the form of an increased financing requirement in the developed countries. However, it was not sufficient for the latter to import these savings. Since the emerging region’s surplus was for the most part invested risk-free, it was necessary that the developed regions take on the risks that the emerging regions did not. In order that savings invested risk free should finance investments that are risky by nature, someone somewhere had to take on the associated risks. That has probably been the most original, and the least emphasized, contribution of globalization. By considerably facilitating the circulation of financial risks, it enabled the developed world to relieve the emerging countries of a significant part of those related to the investing of their savings surplus – at the cost, obviously, of the accumulation of risks in the globalised financial system.” (Brender and Pisani, p. 59 – by risk free, they mean credit risk free)
I am -- no surprise -- sympathetic to this argument. I (independently) made a somewhat similar argument back in November:*
"China wasn’t directly making loans to the most risky borrowers in the US – or actually lending to those who were buying Chinese goods. But the inflow from China was still central to the process that allowed the extension of credit in an economy that itself wasn’t saving, and thus wasn’t generating new funds to lend. Think of it this way: when China bought a Treasury bond from an American insurance company or bank, if provided the pension fund or bank with funds to invest in riskier assets that offered a higher yield than Treasury bonds. Wall Street proved more than capable of churning out ever more complex kinds of mortgage backed securities – and securities composed of parts of other mortgage backed securities – to meet this demand."
Brender and Pisani highlight all the steps in a "global chain of risk taking" that allowed the savings of a Chinese household to be used to finance a US subprime mortgage. But the chain was such that the Chinese household actually never took on all that much risk. The household accepted a low yielding RMB deposit in one of China’s state banks. But the state banks – broadly speaking – didn’t buy many risky US mortgages. Taking exchange rate risk wasn’t their core business.
They did put a lot of RMB on deposit at China’s central bank. And they bought a lot of the “sterilization bills” that China’s central bank issued. No risk there: the state banks were acting like narrow banks, taking in deposits and investing in safe government paper. There is a macroeconomic story here too. The rise in business profits – and business savings – inside China meant that China’s firms weren’t borrowing as much from China’s state banks, freeing up the savings of China’s households to be lent to China’s central bank. The fact that China’s central bank imposed pretty severe limits on bank lending in 2003 no doubt played a role as well; so long as tight lending curbs were in place, the state banks didn’t have much of an incentive to induce Chinese borrowers to borrow more.
China’s central bank played a key role in the chain of risk-taking. Brender and Pisani:
“The Chinese central bank is an essential link in the chain. It buys dollars in exchange for yuans … its role then is that of the exchange rate risk-taker …. It is its policy of not allowing the exchange rate to appreciate too much under the impact of China’s current account surplus that prompts it to play this role. Its intervention binds together the various links in the chain. The particular way in which the dollars purchases on this occasion are invested risk-free is of little importance. If, instead of a deposit with an American commercial bank, the Chinese central bank … acquires a Treasury bill, the seller of the bill will hold a deposit in its place.”
The process wasn’t all that different in the oil exporters. In fact it was a lot simpler: the Treasury of the oil-exporters government just put the dollars it received from the national oil company’s oil sales on deposit at the central bank.
But the global system – especially after 2004 – couldn’t have functioned as it did without private intermediaries willing to take the risk that China’s central banks (and other central banks) wasn’t willing to take. There were lots of such risks.
To simplify a bit, central banks wanted fairly short-term, liquid claims that didn’t pose much (or any) credit risk. They were willing to take exchange rate risk, but not credit risk – and not too much liquidity risk.**
They wanted fairly short-term Treasury notes, or – if they wanted a bit more yield and were willing to give up a bit of liquidity, a fixed rate, fixed maturity bond that one of the US housing agencies issued to finance its own “retained” portfolio of mortgages.***
That worked when the US was running a big fiscal deficit – say in 2003 and 2004. Or if most US mortgages were “prime” mortgages that the Agencies could easily repackage into the kind of bonds that they long had been willing to guarantee. It didn’t work quite as well when a lot of mortgages were being issued to American households that didn’t quality for a prime mortgage (whether because the mortgage was too big, or because they couldn’t come up with the down payment and documented income needed to qualify for a conforming mortgage).
Households wanted to borrow for long-terms, with funky payment terms -- and in a lot of cases they wanted to borrow without putting much money down and even without fully documenting their income. Those little details didn’t matter if home prices only rose; any home could eventually be flipped at a higher price. But such mortgages were intrinsically illiquid. Securitizing their payment streams didn’t really change that fact; securitization ended up turning complex mortgages into illiquid securities. And those securities carried a lot of credit risk if home prices ever stopped appreciating, as we all discovered.
The emerging world’s savings surplus, in theory, could have been used to finance borrowing by US household quite directly. But emerging markets governments generally were not willing to take on the credit risk associated with financing households taking on excessive levels of debt. The system only worked if a private intermediaries – as part of the chain of risk-takers – took on the risks that emerging market central banks didn’t want.
Brender and Pisnani (p. 66):
"In fact, there is nothing to prevent the commercial bank mentioned earlier that receives a deposit from the Chinese central bank from taking on all the other risks related to the mortgage -- liquidity risk, interest-rate risk and credit risk. .... It can however arrive at a similar result -- taking on, against renumeration, the risks mentioned -- without having to have them on its balance sheet. For this purpose, all it has to do is create an ad hoc financial vehicle -- a so-called conduit -- in which it is the main shareholder. This vehicle will buy mortgage loans, financing itself by borrowing short-term and merely benefiting from a line of credit from the bank. As credit-risk and interest-rate risk taker, the vehicle will generate margin but without being subjected to the same prudential constraints as the bank."
Brender and Pisani’s example focuses on a conduit, but the basic logic applies to any institution -- be it a bank with wholesale funding or a hedge fund -- that was willing to borrow short and lend long and take on the credit risk associated with lending to US households. As time went on, the cumulative risk that the private financial intermediaries needed to absorb rose steadily, as Americans could only consume in excess of their income by taking on ever more debt.
Daniel Gros of the Center for European Policy Studies summarized the Brender Pisani argument well in a recent Vox EU column:
"As is well known, the current account deficit of the US arose from an unsustainable increase in consumption (and residential construction). This excess of domestic spending was financed mainly through an increase in the mortgage debt of US households. One key characteristic of mortgages is that they are long-term (often for 30 years). The consumption spree of US households thus led to a large additional supply of long-term (private) assets. However, this supply of longer-term assets was not matched by a corresponding demand for this type of assets. The excess savings from China (and other emerging economies and oil producers) were mostly intermediated by their central bank, which accumulated huge foreign exchange reserves. These reserves were (and still are) almost exclusively invested in short- to medium-term, safe (i.e. government) and liquid securities (mostly in the US). There was thus a need for maturity (and risk) transformation on a very large scale to meet a persistent excess demand for safe and liquid assets."
As the deficits that offset that the surplus of the emerging world shifted toward the US household sector, the ability of the private financial system to assume a set of risks that central banks (as proxies for the savers of the emerging world) didn’t want to bear became central to the sustainability of the system. This wasn’t widely recognized at the time. Most analysts who worried about the sustainability of the "imbalances" -- a group that includes Eichengreen, Rogoff, Wolf and Setser and Roubini -- focused on the risk that emerging market central banks might lose their appetite for dollar risk. That fear wasn’t entirely off: just look at the headlines that popped up this spring. China has suddenly woken up to the exchange rate risks associated with the accumulation of unneeded dollar reserves.
But as Brender and Pisani demonstrate, the stability of the system that financed the US household deficit -- and a slew of deficits in various European countries with housing booms -- hinged both on the willingness of emerging market central banks to take exchange rate risk AND on the willingness of private intermediaries to take the credit and liquidity risk associated with lending at long-terms to ever more-indebted households. And the weak link in the system -- as Nouriel realized before most -- was the ability of private financial intermediaries to keep on taking credit and liquidity risk.
Bretton Woods 2, it turns out, relied both on the willingness of emerging market central banks to take on exchange rate risk and the willingness of private financial intermediaries to finance the exuberance (or excesses) of American households. Both were integral to the system once the US fiscal deficit started to fall -- especially in a world where the Agencies faced limits on their balance sheet growth and fewer and fewer Americans were taking out prime mortgages.
The collapse of private sector intermediation -- particularly in the shadow financial system -- led to a collapse in lending to households, a collapse in consumer spending, a sharp fall in Asian exports and a global contraction. Rising US household debts no longer support rising Chinese exports (or growing investment in China’s export sector). The system has changed.
Rather than financing growth, China is financing (unhappily) adjustment. And that is never that much fun.
And here I am using China as shorthand for a host of emerging market central banks that are adding to their reserves even as their exports shrink.
* That piece was initially written at the prodding of a US newspaper, which – alas -- subsequently decided not to publish it. Ouch!
** Sovereign funds were willing to take on a broader range of risks. But the flows through sovereign funds were comparatively small (despite the attention they received) And – as Brender and Pisani accurately note – sovereign funds were more willing to invest in real estate and equity markets than to take credit risk.
*** Over time, central banks took a broader set of risks. From mid 2007 to mid 2008, many central banks in Asia were buying “Agency MBS” for their own portfolio. Agency MBS have a rather complicated payments structure, and aren’t necessarily the most liquid of bonds. But they also don’t have much credit risk -- at least not so long as the US government stands behind the Agencies. Central banks were never big buyers of CDOs or CDOs squared. Agency MBS was about as racy as most got. And Agency MBS are ultimately backed by a mix of conforming mortgages (no subprime), the Agencies capital (now gone) and the US government.