Repurchase agreements and covert nationalization
The Fed announced that it would auction off $100B in loans this month rather than the previously announced $60B via its TAF facility. In the same press release, the FRB announced plans to offer $100B worth of 28 day loans via repurchase agreements against “any of the types of securities — Treasury, agency debt, or agency mortgage-backed securities — that are eligible as collateral in conventional open market operations”.
The second announcement puzzled me. After all, the Fed conducts uses repos routinely in the open market operations by which they try to hold the interbank lending rate to the Federal Funds target. In aggregate, the quantity of funds that the Fed makes available is constrained by the Fed Funds target. So, what do we learn from this? Fortunately, the New York Fed provides more details:
The Federal Reserve has announced that the Open Market Trading Desk will conduct a series of term repurchase (RP) transactions that are expected to cumulate to $100 billion outstanding… These transactions will be conducted as 28-day term RP agreements.. When the Desk arranges its conventional RPs, it accepts propositions from dealers in three collateral “tranches.” In the first tranche, dealers may pledge only Treasury securities. In the second tranche, dealers have the option to pledge federal agency debt in addition to Treasury securities. In the third tranche, dealers have the option to pledge mortgage-backed securities issued or fully guaranteed by federal agencies in addition to federal agency debt or Treasury securities. With the special “single-tranche” RPs announced today, dealers have the option to pledge either mortgage-backed securities issued or fully guaranteed by federal agencies, federal agency debt, or Treasury securities. The Desk has arranged single-tranche transactions from time to time in the past.
There are a couple of differences, then, between this new program and typical repo operations:
- The loans are of a longer-term than usual. Ordinarily, the Fed lends on terms ranging from overnight to two weeks in its “temporary open market operations”. The Fed will now offer substantial funding on a 28 day term.
- The Fed is effectively broadening its collateral requirements by collapsing what are usually 3 distinct levels of collateral which are lent against at different rates to a single category within which no distinctions are made.
The Fed offered the first $15B of repo loans under the program today, so we can see how things are going to work. First, how did the Fed square the circle of ramping up its repos without pushing down the Federal Funds rate? Just as it had done with TAF, the Fed offset the “temporary” injection of funds with a “permanent open market operation”. The Fed sold outright $10B of Treasury securities today at the same time as it offered $15B in exchange for mortgage-backed securities under the new program (at a low interest rate than in traditional repos against MBS collateral). The net cash injection was small, but the composition of securities on bank balance sheets changed markedly, as illiquid securities were exchanged for liquid Treasuries.
In James Hamilton’s wonderful coinage, the Fed is conducting monetary policy on the asset side of the balance sheet. This is an innovation of the Bernanke Fed. Conventionally, monetary policy is about managing the quantity of the central bank’s core liability, currency outstanding. When the Fed wants to loosen, it expands its liabilities by issuing cash in exchange for securities. When it wants to tighten, it redeems cash for securities, reducing Fed liabilities. The asset side is conventionally an afterthought, “government securities”. But the Bernanke Fed has branched out. It has sought to lend against a wide-range of assets, actively seeking to replace securities about which the market seems spooked with safe-haven Treasuries on bank balance sheets without creating new cash. By doing this, the Fed hopes to square the circle of helping banks through their “liquidity crisis” without provoking a broad inflation.
“Monetary policy on the asset side of the balance sheet” is a bit too anodyne a description of what’s going on here though. The Fed has gotten into an entirely new line of business, and on a massive scale. Prior to the introduction of TAF, direct loans from the Fed to banks, including the discount window lending and repos, amounted to less than $40B, the majority of which were repos collateralized by Treasury securities. By the end of this month, the Federal Reserve will have more than $200B of exposure in its new role as Wall Street’s genial pawnbroker. Assuming the liability side of the Fed’s balance sheet is held roughly constant, more than a fifth of the Fed’s balance sheet will be direct loans to banks, almost certainly against collateral not backed by the full faith and credit of the US government (and beyond that we just don’t know). This raises a whole host of issues.
Caroline Baum wrote a column last week poopooing concerns about the Fed taking on credit risk via TAF lending. (Hat tip Mark Thoma.) I usually enjoy Baum’s work, but this column was poorly argued. In it, she points out that the Fed has all the tools it needs to manage credit risk. The Fed offers loans only against collateral, and requires that loans be overcollateralized. If the collateral has no clear market value or if there are questions about an asset’s quality, the Fed has complete discretion to force a “haircut”, writing down the asset (for the purpose of the loan) to whatever value it sees fit. And the Fed can always just say no to any collateral it deems sketchy.
All of that is quite true, and (as Baum snarkily points out) not hard to find on FRB websites. But it fails to address the core issue. Sure the Fed has all the tools it needs to manage credit risk. But does it have the will to use those tools? In word and deed, the Fed’s primary concern since August has been to “restore normal functioning” to financial markets. The Fed has chosen to accept some inflation risk in its fight against macroeconomic meltdown. Why wouldn’t it knowingly accept some credit risk as well? No one has suggested that the Fed is being “snookered”. Skeptics think the Fed is intentionally taking on bank credit risk while still lending at very low rates. Some of us find that troubling.
Which brings us to the more postmodern issue of what credit risk even means to a lender with unlimited cash and an overt unwillingness to let those it lends to default. In a way, I agree with Baum. Until the current crisis is long past, I think it unlikely that any large bank will default and stiff the Fed with toxic collateral. Why not? Because for that to happen, the Fed would have to pull the trigger itself, by demanding payment on loans rather than offering to roll them over. Since TAF started last fall, on net, the Fed has not only rolled over its loans to the banking system, but has periodically increased banks’ line of credit as well. In an echo of the housing bubble, there’s no such thing as a bad loan as long as borrowers can always refinance to cover the last one.
The distinction between debt and equity is much murkier than many people like to believe. Arguably, debt whose timely repayment cannot be enforced should be viewed as equity. (Financial statement analysts perform this sort of reclassification all the time in order to try to tease the true condition of firms out of accounting statements.) If you think, as I do, that the Fed would not force repayment as long as doing so would create hardship for important borrowers, then perhaps these “term loans” are best viewed not as debt, but as very cheap preferred equity.
Let’s go with that for a minute, and think about the implications. One much discussed story of the current crisis is the role of sovereign wealth funds in helping to capitalize struggling banks. Will they, won’t they, should we worry? Sovereign wealth funds have invested about $24B in struggling US financials. Meanwhile, the Fed is quietly providing eight times that on much easier terms.
If we view TAF and the new 28-day, broad-collateral repos as equity, what fraction of bank capitalization would they represent? I haven’t been able to find current numbers on aggregate bank capitalization in the US. In June of 2006, the accounting net worth of U.S. Commercial Banks, Thrift Institutions and Credit Unions was 1.25 trillion dollars. Putting together remarks by Fed Vice Chairman Donald Kohn and data on bank equity to total assets from the St. Louis Fed yields a more recent estimate of about 1.6 trillion. The average price to book among the top ten US banks is about 1.3. So, a reasonable estimate for the current market value of bank equity is 2 trillion dollars. The $200B in “equity” the Fed will have supplied by the end of March will leave the Federal Reserve owning roughly 9.1% of the total bank equity. Obviously, the Fed isn’t investing in the entire bank sector uniformly. Some banks will be very substantially “owned” by the central bank, whereas others will remain entirely private sector entities. As Dean Baker points out, the Fed is giving us no information by which to tell which is which.
What we are witnessing is an incremental, partial nationalization of the US banking system. Northern Rock in the UK is peanuts compared to what the New York Fed is up to.
You may object, and I’m sure many of you will, that our little thought experiment is bunk, debt is debt and equity is equity, these are 28-day loans, and that’s that. But notionally collateralized “term” loans that won’t ever be redeemed unless and until it is convenient for borrowers are an odd sort of liability. Central banks are very familiar with the ruse of disguising equity as liability. Currency itself is formally a liability of the central bank, but in every meaningful sense fiat money is closer to equity.
I do not, by the way, object to nationalizing failing banks. There are (unfortunately) banks that are “too big to fail”, whose abrupt disappearance could cause widespread disruption and harm. These should be nationalized when they fall to the brink. But they should be nationalized overtly, their equity written to zero, and their executives shamed. That sounds harsh. It is harsh. One hates to see bad things happen to nice people, and these are mostly nice people. But running institutions with trillion-dollar balance sheets is a serious business. Accountability matters. These people were not stupid. They knew, in Chuck Prince’s now infamous words, that “when the music stops… things will be complicated.“, and they kept dancing anyway.
But accountability has gone out of style. The Federal Reserve is injecting equity into failing banks while calling it debt. Citibank is paying 11% to Abu Dhabi for ADIA’s small preferred equity stake, while the US Fed gets under 3% now for the “collateralized 28-day loans” it makes to Citi. Pace Accrued Interest (whom I much admire), I still think this all amounts to a gigantic bail-out. And that it is a brilliantly bad idea from which financial capitalism may have a hard time recovering. Like a well-meaning surgeon slicing up arteries to salvage the appendix, the Federal Reserve is only trying to help.
Update: Many thanks to commenter Fullcarry, who noticed that I flipped the sign on Friday’s permanent market op. Changed “purchased outright” to “sold outright”, since that’s what the Fed actually did.
- 8-Mar-2008, 1:00 p.m. EST: Added update re changing “purchased outright” to “sold outright”, since that’s what the Fed actually did.
- 8-Mar-2008, 2:30 p.m. EST: Changed “trillions dollar” to “trillion-dollar”.