I'm no monetary economist, so I really would like to know. This is my stab at a rudimentary understanding. I'd love to hear from commenters if I'm totally off on any of this.
First, my understanding is that in modern banks, reserves don't really serve as a direct constraint on lending. Clearly, they aren't the constraint now for bank lending. The constraint now must be related to bank capital and regulatory oversight.
Capital conditions among banks continue to recover - credit along with them. Here is a graph of bank credit and the Fed Funds Rate in the Great Moderation era. When investment demand dries up in the early part of cyclical downturns, we see bank credit level out and interest rates decline. Then, bank credit starts growing again, at a remarkably regular rate, appearing to top out at about 10% growth (the left scale is a log scale). Since 2008, bank credit has been stagnant. QE filled some of that gap.
Since the beginning of 2014, loans and leases in bank credit have begun growing again at nearly that 10% rate (the little kink up in the blue line in the top right corner). While real estate loan growth has been part of this return to growth, it has been a very small part of it. So, I expect bank credit to continue to accelerate as real estate loans recover, although, I am a little disappointed with this credit category. It has been growing at an annual rate of about 3% since the beginning of 2014, with no real acceleration. In fact, real estate is still declining as a proportion of total bank assets (minus cash). But, we are just now approaching the pre-recession level of home equity leverage, so there is still a lingering headwind of household deleveraging. I still expect to see real estate credit accelerate soon, as we leave that constraint behind. But, if home values stop moving up and there isn't enough momentum to develop real estate credit growth, we could be in some trouble.
(Calculated Risk notes that some are seeing price concessions among home sellers in California. I still suspect that this is a kind of negative head fake as the buyer's market switches from QE funded cash buyers to bank funded mortgage buyers. Partly what makes me optimistic about intrinsic home values is that I think it is quite amazing that it is considered notable that some sellers are capitulating on price when there has been no support for the real estate market from the banks for more than 5 years.)
In any case, it appears that normally within a few months of bank credit growth reaching that terminal growth level, credit demand begins to push interest rate levels upward. This is the "We are the 100%" economy, where emergent growth leads to surplus. Wages, interest rates, and profits all tend to move up in this context as equilibria prices among all of an economy's participants float upward so that all share some of this surplus as a result of countless marginal adjustments.
The natural interest rate has been negative, and QE has replaced the commercial banks as a source of liquidity in the economy. But, as QE3 unwinds, it appears that banks are emerging as credit creators, and so I think we should expect the typical patterns to follow.
But, with excess reserves at the banks, I think there should be a limit to interest rate pressures. I think the net result of increasing credit demand will be to draw credit outside the banks, as banks hit their capital constraints. Reserves will still end up at the banks, but I think this will lead to a residual increase in currency in circulation and maybe inflationary pressures. So, I wonder if an increase in currency in circulation will be a sign that there are natural upward pressures on short term interest rates.
The Fed intends to lift both the Fed Funds Rate and Interest on Reserves in order to move rates up before they pull all of the excess reserves out of the system. But, with $3 trillion in cash at the commercial banks, a 1% rate increase on reserves means $30 billion in additional annual capital going to banks or depositors. Banks could parlay that $30 billion into an additional $150 billion in loans and leases. I suppose the Fed will be selling treasuries at the same time, which should have a slight disinflationary effect.
So, I'm in over my head on this one. I'd love to be corrected in the comments on my mistakes. Don't fail me, readers...