Courtesy of the recently declassified Fed discount window documents, we now know that the biggest beneficiaries of the Fed's generosity during the peak of the credit crisis were foreign banks, among which Belgium's Dexia was the most troubled, and thus most lent to, bank. Having been thus exposed, many speculated that going forward the US central bank would primarily focus its "rescue" efforts on US banks, not US-based (or local branches) of foreign (read European) banks: after all that's what the ECB is for, while the Fed's role is to stimulate US employment and to keep US inflation modest. And furthermore, should the ECB need to bail out its banks, it could simply do what the Fed does, and monetize debt, thus boosting its assets, while concurrently expanding its excess reserves thus generating fungible capital which would go to European banks. Wrong. Below we present that not only has the Fed's bailout of foreign banks not terminated with the drop in discount window borrowings or the unwind of the Primary Dealer Credit Facility, but that the only beneficiary of the reserves generated were US-based branches of foreign banks (which in turn turned around and funnelled the cash back to their domestic branches), a shocking finding which explains not only why US banks have been unwilling and, far more importantly, unable to lend out these reserves, but that anyone retaining hopes that with the end of QE2 the reserves that hypothetically had been accumulated at US banks would be flipped to purchase Treasurys, has been dead wrong, therefore making the case for QE3 a done deal. In summary, instead of doing everything in its power to stimulate reserve, and thus cash, accumulation at domestic (US) banks which would in turn encourage lending to US borrowers, the Fed has been conducting yet another stealthy foreign bank rescue operation, which rerouted $600 billion in capital from potential borrowers to insolvent foreign financial institutions in the past 7 months. QE2 was nothing more (or less) than another European bank rescue operation!
For those who can't wait for the punchline, here it is. Below we chart the total cash holdings of Foreign-related banks in the US using weekly H.8 data.
Note the $630 billion increase in foreign bank cash balances since November 3, which just so happens is the date when the Fed commenced QE2 operations in the form of adding excess reserves to the liability side of its balance sheet. Here is the change in Fed reserves during QE2 (from the Fed's H.4.1 statement, ending with the week of June 1).
Above, note that Fed reserves increased by $610 billion for the duration of QE2 through the week ending June 1 (and by another $70 billion in the week ending June 8, although since we only have bank cash data through June 1, we use the former number, although we are certain that the bulk of this incremental cash once again went to foreign financial institutions).
So how did cash held by US banks fare during QE2? Well, not good. The chart below demonstrates cash balances at small and large US domestic banks, as well as the cash at foreign banks, all of which is compared to total Fed reserves plotted on the same axis. It pretty much explains it all.
The chart above has tremendous implications for everything from US and European monetary policy, to exhange rate and trade policy, to the current account on both sides of the Atlantic, to US fiscal policy, to borrowing and lending activity in the US, and, lastly, to QE 3.
What is the first notable thing about the above chart is that while cash levels in US and US-based foreign-banks correlate almost perfectly with the Fed's reserve balances, as they should, there is a notable divergence beginning around May of 2010, or the first Greek bailout, when Europe was in a state of turmoil, and when cash assets of foreign banks jumped by $200 billion, independent of the Fed and of cash holdings by US banks. About 6 months later, this jump in foreign bank cash balances had plunged to the lowest in years, due to repatriated fungible cash being used to plug undercapitalized local operations, with total cash just $265 billion as of November 17, just as QE2 was commencing. Incidentally, the last time foreign banks had this little cash was April 2009... Just as QE1 was beginning. As to what happens next, the first chart above says it all: cash held by foreign banks jumps from $308 billion on November 3, or the official start of QE2, to $940 billion as of June 1: an almost dollar for dollar increase with the increase in Fed reserve balances. In other words, while the Fed did nothing to rescue foreign banks in the aftermath of the first Greek crisis, aside from opening up FX swap lines, one can argue that the whole point of QE2 was not so much to spike equity markets, or the proverbial "third mandate" of Ben Bernanke, but solely to rescue European banks!
What this observation also means, is that the bulk of risk asset purchasing by dealer desks (if any), has not been performed by US-based primary dealers, as has been widely speculated, but by foreign dealers, which have the designatin of "Primary" with the Federal Reserve. Below is the list of 20 Primary Dealers currently recognized by the New York Fed. The foreign ones, with US-based operations, are bolded:
- BNP Paribas Securities Corp.
- Barclays Capital Inc.
- Cantor Fitzgerald & Co.
- Citigroup Global Markets Inc.
- Credit Suisse Securities (USA) LLC
- Daiwa Capital Markets America Inc.
- Deutsche Bank Securities Inc.
- Goldman, Sachs & Co.
- HSBC Securities (USA) Inc.
- Jefferies & Company, Inc.
- J.P. Morgan Securities LLC
- MF Global Inc.
- Merrill Lynch, Pierce, Fenner & Smith Incorporated
- Mizuho Securities USA Inc.
- Morgan Stanley & Co. LLC
- Nomura Securities International, Inc.
- RBC Capital Markets, LLC
- RBS Securities Inc.
- SG Americas Securities, LLC
- UBS Securities LLC.
That's right, out of 20 Primary Dealers, 12 are.... foreign. And incidentally, the reason why we added the (if any) above, is that since this cash is fungible between on and off-shore operations, what happened is that the $600 billion in cash was promptly repatriated and used by domestic branches of foreign banks to fill undercapitalization voids left by exposure to insolvent European PIIGS and for all other bankruptcy-related capital needs. And one wonders why suddenly German banks are so willing to take haircuts on Greek bonds: it is simply because courtesy of their US based branches which have been getting the bulk of the Fed's dollars in 1 and 0 format, they suddenly find themselves willing and ready to face the mark to market on Greek debt from par to 50 cents on the dollar. And not only Greek, but all other PIIGS, which will inevitably happen once Greece goes bankrupt, either volutnarily or otherwise. In fact, the $600 billion in cash that was repatriated to Europe will mean that European banks likely are fully covered to face the capitalization shortfall that will occur once Portugal, Ireland, Greece, Spain and possibly Italy are forced to face the inevitable Event of Default that will see their bonds marked down anywhere between 20% and 60%. Of course, this will also expose the ECB as an insolvent central bank, but that largely explains why Germany has been so willing to allow Mario Draghi to take the helm at an institution that will soon be left insolvent, and also explains the recent shocking animosity between Angela Merkel and Jean Claude Trichet: the German are preparing for the end of the ECB, and thanks to Ben Bernanke they are certainly capitalized well enough to handle the end of Europe's lender of first and last resort. But don't take our word for this: here is Stone McCarthy's explanation of what massive reserve sequestering by foreign banks means: "Foreign banks operating in the US often lend reserves to home offices or other banks operating outside the US. These loans do not change the volume of excess reserves in the system, but do support the funding of dollar denominated assets outside the US....Foreign banks operating in the US do not present a large source of C&I, Consumer, or Real Estate Loans. These banks represent about 16% of commercial bank assets, but only about 9% of bank credit. Thus, the concern that excess reserves will quickly fuel lending activities and money growth is probably diminished by the skewing of excess reserve balances towards foreign banks."
Which brings us to point #2: prepare for the Bernanke hearings and possible impeachment. For if it becomes popular knowledge that the Chairman of the Fed, despite explicit instructions to enforce the trickle down of "printed" dollars to US banks, was only concerned about rescuing foreign banks with the $600 billion in excess cash created out of QE2, then all political hell is about to break loose, and not even Democrats will be able to defend Bernanke's actions to a public furious with the complete inability to procure a loan. Any loan. Furthermore the data above proves beyond a reasonable doubt why there has been no excess lending by US banks to US borrowers: none of the cash ever even made it to US banks! This also resolves the mystery of the broken money multiplier and why the velocity of money has imploded.
Implication #3 explains why the US dollar has been as week as it has since the start of QE 2. Instead of repricing the EUR to a fair value, somewhere around parity with the USD, this stealthy fund flow from the US to Europe to the tune of $600 billion has likely resulted in an artificial boost in the european currency to the tune of 2000-3000 pips, keeping it far from its fair value of about 1.1 EURUSD. If this data does not send European (read German) exporters into a blind rage, after the realization that the Fed (most certainly with the complicity of the G7) was willing to sacrifice European economic output in order to plug European bank undercapitalization, then nothing will.
But implication #4 is by far the most important. Recall that Bill Gross has long been asking where the cash to purchase bonds come the end of QE 2 would come from. Well, the punditry, in its parroting groupthink stupidity (validated by precisely zero actual research), immediately set forth the thesis that there is no problem: after all banks would simply reverse the process of reserve expansion and use the $750 billion in Cash that will be accumulated by the end of QE 2 on June 30 to purchase US Treasurys.
The above data destroys this thesis completely: since the bulk of the reserve induced bank cash has long since departed US shores and is now being used to ratably fill European bank balance sheet voids, and since US banks have benefited precisely not at all from any of the reserves generated by QE 2, there is exactly zero dry powder for the US Primary Dealers to purchase Treasurys starting July 1.
This observation may well be the missing link that justifies the Gross argument, as it puts to rest any speculation that there is any buyer remaining for Treasurys. Alas: the digital cash generated by the Fed's computers has long since been spent... a few thousand miles east of the US.
Which leads us to implication #5. QE 3 is a certainty. The one thing people focus on during every episode of monetary easing is the change in Fed assets, which courtesy of LSAP means a jump in Treasurys, MBS, Agency paper, or (for the tin foil brigade) ES: the truth is all these are a distraction. The one thing people always forget is the change in Fed liabilities, all of them: currency in circulation, which has barely budged in the past 3 years, and far more importantly- excess reserves, which as this article demonstrates, is the electronic "cash" that goes to needy banks the world over in order to fund this need or that. In fact, it is the need to expand the Fed's liabilities that is and has always been a driver of monetary stimulus, not the need to boost Fed assets. The latter is, counterintuitively, merely a mathematical aftereffect of matching an asset-for-liability expansion. This means that as banks are about to face yet another risk flaring episode in the next several months, the Fed will need to release another $500-$1000 billion in excess reserves. As to what asset will be used to match this balance sheet expansion, why take your picK; the Fed could buy MBS, Muni bonds, Treasurys, or go Japanese, and purchase ETFs, REITs, or just go ahead and outright buy up every underwater mortgage in the US. This side of the ledger is largely irrelevant, and will serve only two functions: to send the S&P surging, and to send the precious metal complex surging2 as it becomes clear that the dollar is now entirely worthless.
That said, of all of the above, the one we are most looking forward to is the impeachment of Ben Bernanke: because if there is one definitive proof of the Fed abdicating any and all of its mandates, and merely playing the role of globofunder explicitly at the expense of US consumers and borrowers, not to mention lackey for the banking syndicate, this is it.