It is now (almost) official: q1 dollar reserve growth exceeded the US current account deficit …

The BIS reported data on central banks’ accumulation of dollars in the international banking system today (Table 5c).    Central banks added $24.5b to their dollar holdings – but also repaid $53.5b in loans from the world’s banks.   That works out to a $78b increase in central bank’s net dollar position – which rose from $534.5b to $612.5b.   When a central bank pays back a loan from a commercial bank, the commercial bank is able to lend those funds out – increasing the “liquidity” available to private borrowers.    Not all these dollars flowed to the US either.  But the availability of dollar financing in offshore financing centers certainly makes it easier for a range of financial players to buy US securities.  

The same BIS data shows a $50.7b increase in central banks net euro position in q1 ($47b in flow terms) and a $10.7b fall in central banks net pound position (-10.3b in flow terms).   

In ball park terms, net dollar deposits ($78b) grew almost twice as fast as net euro and pound deposits ($40b).  That doesn’t suggest much of a shift away from the dollar.  I wouldn’t put too much faith in this data though.  It covers a small subset of total central bank claims, and I suspect that central banks ran down the pound deposits to buy other pound assets, not to buy dollars or euros.

The BIS data can be combined with the US data to get an estimate of the total increase in central banks dollar holdings.  The US BEA reported $147.8b in net official inflows to the US in q1.  Most of that came from the purchase of US securities, but there was also a $29.4b increase in (“onshore”) bank deposits.  To get an estimate of reported dollar reserve growth that avoid double counting, I need to subtract the increase in onshore deposits from the US total and then add in the overall increase in dollar deposits reported by the BIS.   That produces a $196.4b increase in central banks “known” net dollar holdings. 

The q1 2007 US current account deficit was $192.6b.  

Even if not all central bank offshore dollar deposits flowed directly to the US, it doesn’t take a genius to figure out who is currently financing most of the US deficit.

The US data doesn't pick up all offshore dollar deposits.  And the US data also likely underestimates official purchases of US securities.   The BEA data always gets revised up after the survey of foreign portfolio holdings.    And we know that the increase in the FRBNY’s custodial holdings in q1 ($127.6b) was $21.3b larger than the purchases of Treasuries and Agencies reported in the US data.  Add that $21.3b to $192.6b, and there was a $213.9b increase in (net) official dollar holdings in the first quarter. 

If that isn’t a record, it has to be close –

A backward looking data point is admittedly a strange thing to calculate on a day when the broad market is selling off.  But I suspect it also provides a bit of insight into some of the underlying reasons for recent market moves.

The original idea behind all of my efforts to track reserve growth was to get some advance notice if central banks ever decided to turn away from US assets.    That hasn't happened; all available data suggests record official demand for US assets.  But I don’t think my efforts have been a total waste.  Data on central bank reserves helped, for example, to understand what likely drove the Treasury sell-off in June (that sure seems a long-time ago; Treasuries are back well under 5%).    

But it certainly doesn’t help to explain the current sell-off in risk assets.    The “official” bid for risk assets is rising – though perhaps a few big buyers are having second thoughts about diversifying into “private” mortgage-backed securities.   And the latest FRBNY custodial data shows a $27.8b increase in official holdings of treasuries and agencies over the last four weeks.   The FRBNY data doesn’t capture all central bank related flows – it is a floor, not a ceiling.  For all that attention generated by sovereign wealth funds, a lot more money is still flowing into traditional reserve assets.  

The markets haven’t seized up because central banks stopped buying dollars, or stopped buying dollar-denominated bonds. 

Rather, they have seized up because a lot of private investors who had reached for yield over the past few years, in part because central banks drove down the yields on “safe” assets as well as contributing to a fall in market volatility, seem to have reached a bit too far.   They lent a lot of money to folks trying to afford to buy a house.    And they snapped up CLOs stuffed with LBO related debt, financing the buyout boom that drove up equity markets.  As a result,  until recently, “even the riskiest companies could obtain credit cheaply.”

Mohammed El-Erian puts it well.  For a while, 

“individual investors' performance was essentially a function of the degree of their exposure to the most illiquid and leveraged asset classes.”

That naturally created pressure to take on more leverage to buy more illiquid assets. 

Things now have changed.  Lots of subprime mortgages were bundled into mortgage backed securities and various tranches of different mortgage backed securities in turn were bundled together in often illiquid CDOs.  Some of those CDOs were bought with borrowed money.   That turned into a problem when folks with limited income started defaulting on their mortgage debt as housing prices turned.    

More recently demand for LBO debt disappeared, in part because of concerns that some companies may be taking on more debt than they could service.  The IMF has a nice chart showing the rise in leverage in private equity deals (p.3)  – along with some interesting graphs on subprime default rates in various “vintages” (p.2) in their recent market overview.    Throw in concerns that triple AAA rated structures might have somewhat more risk than the ratings agencies initially thought, and, to quote PIMCO's Bill Gross, the market resembles a constipated owl.

“To be blunt, they seem to be thinking that if Moody’s and Standard & Poor’s have done such a lousy job of rating subprime structures, how can the market have confidence that they’re not repeating the same structural, formulaic, mistake with CLOs and CDOs? That growing lack of confidence – more so than the defaults of two Bear Stearns hedge funds and the threat of more to come – has frozen future lending and backed up the market for high yield new issues such that it resembles a constipated owl: absolutely nothing is moving.”

Felix notes that the fact that a lot of the new instruments that emerged to satisfy investors demand for yield back are untested isn’t evidence that they are unsafe.  All new instruments by definition haven’t been tested.  

But it still seems to me that the market for a lot of instruments that hadn’t been tested by a downturn got really big really fast.   Right now that worries me more than the risk that central banks will suddenly lose their appetite for all dollar assets.