Constraints on Lending: Reserves vs Capital
I had an interesting conversation with a former executive at a commercial bank (let's call him D), who confirmed for me some things related to our monetary system which I had been reading about
here and
here (particularly the comments).
In a nutshell, these sources assert that bank lending is not constrained by
reserves. D confirmed for me that during his tenure as a commercial banker, he never had to worry about his bank's reserve ratio -- loans were made without regard to it. The biggest constraint on lending (on the supply side) was bank
capital, or the
capital ratio. D told me that if you asked any bank president what his bank's reserve ratio was, most wouldn't have any idea. Capital ratios are a whole other ball of wax though, and careful attention is paid to them by management.
If loans come first, how do banks keep their reserve ratios in line? First of all, reserve requirements have been steadily eroded in the US banking system, with sweep accounts and other "innovations." Second of all, if the banking system is short reserves, the Fed accommodates it by injecting more. But note
: the Fed follows loan issuance,
it does not cause it.
Observe that this completely invalidates the traditional money multiplier model of the US monetary system as taught in econ 101, wherein the Fed injects new reserves which then get multiplied into the money supply via fractional reserve lending. If banks lend regardless of reserves, it follows that the money supply is not determined by the level of reserves in the system.
Applications to Current Situation
Applied to today's situation in the banking system, this insight leads to some interesting conclusions. It's been well-publicized that excess reserves have reached a level around $1T. This has caused many to ask why banks aren't lending out this huge horde of reserves, and others to predict massive inflation around the corner when banks do start to lend it out. Armed with our insight, we can refute the first group: since lending isn't reserves-constrained, the $1T number is irrelevant. The second group may or may not be right, but since reserves don't matter, their reasoning is certainly wrong.
If reserves don't matter, capital matters a great deal. And here is where the US banking system is desperately in need -- banks have taken huge losses which have decimated their capital base. D observed that many banks today are open with capital ratios which would've been unheard of in normal conditions (as low as 2%), and the only thing saving them from being shut down is that the FDIC doesn't have enough staff to handle all the closures which need to occur. Additional hits to capital are still in the pipeline in the form of future CRE losses, for example.
Given that we know bank lending is capital-constrained and banks are short on capital, we can predict that lending isn't about to take off anytime soon -- quite the opposite, in fact.
The Fed's Role: Rhetoric vs Reality
Ben Bernanke and other Fed officials have repeatedly stated that their goal is to restore the flow of credit, i.e. bank lending. As examples of steps they have taken to achieve this goal, they have pointed to actions such as lowering the Fed Funds rate, MBS purchases, and TARP. Knowing what we now know about reserves and capital, we can say that the first two actions are non-events, except for one possible exception which I will take up in the next section. Lowering the Fed Funds rate and MBS buys have made bank reserves more plentiful than ever, but bank lending is capital-constrained not reserves-constrained. Of the three, TARP has certainly had the biggest impact by injecting capital into the banking system. However, it appears that the hits to capital are of a magnitude that exceeds even the $700B TARP capital injections. Given that securitization and SIVs and the like are now history, and access to capital is difficult, capital constraints appear to be severe. Even after TARP, many banks are teetering on the edge of insolvency.
There appears to be one further step which the Fed could take to get banks lending, which I would now like to discuss.
The Fed's Role: Are They Really Trying to Get Banks to Lend?
Banks' capital ratios are impaired, and any new loan they make without increasing capital drags their capital ratio down even further. Any new loan, that is, except for one kind -- to the US Government. D confirmed that banks do not have to include Treasuries in the denominator of the capital ratio -- lending to the USG is always possible, even for banks whose capital ratios are on the brink. Of the many risks of making bank loans -- default risk, capital risk, liquidity risk, market risk, etc. -- Treasuries suffer from just two: interest rate risk and duration risk. If I buy a 30-year Treasury today, rates could spike next week, leaving me with a loss. Likewise, a bank holding a 30-year Treasury as an asset would need to have a matching duration liability.
An answer to both of these risks is for a bank to lend to the US government very short term by borrowing from the Fed. If the Fed wished, it could offer the banks a spread in a no-lose arbitrage strategy: borrow from the Fed at the Fed Funds rate for 90 days (the borrowing term is typically overnight, but the Fed can offer to roll these loans for any length of time), and buy a Treasury of similar or shorter duration, i.e. 3 months or less. All the Fed would have to do is to lower the Fed Funds rate below the 3 month Treasury bill rate. Since the Fed can lower the Fed Funds rate to any amount they like, it would appear that they have a sure-fire way of giving the banks a profitable spread, which would certainly spur bank lending as Ben Bernanke & Co. have said they wish to do -- they would be literally giving the banks a guaranteed profit with no downside, in the process allowing the banks to recapitalize themselves. When I asked D about this as a possible Fed strategy, he said he would have to think about it, but couldn't see anything wrong with it off the top of his head.
There is only one kink in this argument -- the Fed Funds rate has stayed consistently
above the 3 month Treasury bill:
(See also
Are Fed interest rates high or low?)
It would appear that this guaranteed way to spur bank lending isn't something that interests the Federal Reserve. I think this indicates that their stance is currently very deflationary despite all their protestations to the contrary. My conspiracy theory is that this is because the Fed is a bank, and bankers generally prefer deflation as long as they don't go bust -- and the Fed can't go bust, what with the taxpayer backing them. But that's a subject for a different post.
