This past month or so, the Fed has created facilities capable of pumping about $400bn worth of liquidity into the US financial system. I expect that the Fed will have to do a lot more by way of liquidity injections. Over the remainder of the year, even an additional $1 trillion of collateralised lending by the Fed (over and above the normal levels seen before the second half of 2007) is likely to be inadequate to get key financial markets functioning effectively again and to keep them that way. It would barely be more than $100bn extra per month – too little to make a substantial difference. My guess is that somewhere between $1.5trillion and $2.0 trillion of further above-normal liquidity provision will be required in the US before this crisis is over.
This guestimate is based on the size of the balance sheet of the leveraged sector, the maturity distribution of its liabilities, the likely additional demand for liquidity arising from the need to transfer the assets of off-balance-sheet vehicles onto the balance sheet of the banking sector, and the probable continued disruption of private wholesale financial markets.
I expect that while most of this will be through repo-type operations (using vehicles like the Term Auction Facility’ (TAF) for banks and the ‘Term Securities Lending Facility’ TSFL) for primary dealers), some of it will ultimately be through outright purchases of non-US agency (GSE)–guaranteed private debt instruments like RMBS and corporate bonds. I believe the Federal Reserve Act does not rule this out (although it does not explicitly allow it either). I am assuming that, in a crisis, anything that is not explicitly forbidden is allowed; even some things that are explicitly forbidden will be condoned.
The Federal Reserve Act allows the Fed to invoke ‘unusual and exigent circumstances’ under which it would be allowed to conduct outright open-market purchases of ‘bills of exchange’ and ‘bankers’ acceptances’. The question then becomes how elastic the interpretation of ‘bills of exchange’ and ‘bankers’ acceptances’ would be. It is my belief that in the midst of a financial crisis, it would be hard for any agency or court to stop the Fed from declaring ‘bills of exchange’ and ‘bankers’ acceptances’ to be virtually any kind of private security, including corporate bonds. Perhaps they would draw the line at pure equity.
When it has declared circumstances to be ‘unusual and exigent’, the Fed can also lend, against any collateral it deems appropriate (including second-hand cars and dead dogs), to individuals, partnerships and corporations. Extending the range of eligible counterparties is one way of bypassing the bottlenecks preventing Fed liquidity from flowing freely through the depositary institutions and the primary dealers that hoard the liquidity to other financial institutions and the rest of the economy that need it.
When the Fed, in an outright purchase, takes mortgage-backed securities risk directly on its balance sheet, (rather than the risk it runs in a collateralized loan, that both the borrower and the issuer of the collateral will default), it is even more important that the collateral or the securities that are purchased outright be valued aggressively and subject to a severe haircut or discount on this aggressive valuation. Without that, moral hazard could become rampant.
There is nothing wrong with the Fed or any other central bank taking credit risk onto its balance sheet, as long as that risk is appropriately priced. An appropriately higher expected rate of return compensates for greater risk, including default risk. In the end, the Fed is backed by the US Treasury, so it is always possible to recapitalize the central bank without recourse to the inflation tax.
As long as there is a non-zero probability of default on the securities the Fed acquires as collateral or outright as a result of its liquidity management for the economy, it is possible that the tax payer will have to come to the rescue of the central bank. The appropriate risk premium does not cover the biggest loss that could conceivably occur. But the securities acquired by the Fed should be priced low enough to ensure that their expected rate of return at least compensates (ex-ante) the Fed for the risk of capital losses on these securities.