Finally, the manager reported on potential arrangements that would allow depository institutions to pledge funds held in a segregated account at the Federal Reserve as collateral in borrowing transactions with private creditors and would provide an additional supplementary tool during policy normalization; the manager noted possible next steps that the staff could potentially undertake to investigate the issues related to such arrangements.A slide presentation by the New York Fed's Jamie McAndrews explains it.
The simple version, as I understand it, seems like great news. Basically, a company can deposit money at a bank, and the bank turns around and invests that money in interest-paying reserves at the Fed. Unlike regular deposits, which you lose if the bank goes under, (these deposits are much bigger than the insured limit) the depositor has a collateral claim to the reserves at the Fed.
This is then exactly 100% reserve, bankruptcy-remote, "narrow banking" deposits. I argued for these in "toward a run-free financial system" as a substitute for all the run-prone shadow-banking that fell apart in the financial crisis. (No, this isn't going to siphon money away from bank lending, as the Fed buys Treasuries to issue reserves. The volume of bank lending stays the same.)
A second function of such deposits is that, like the new repo facility, it's going to help the Fed to raise rates. When the Fed wants to raise rates it will pay more interest on reserves. The question is, will banks pass that interest on to depositors? If they were competitive they would, but that's not so obvious. If large depostitors can access interest-bearing reserves through the repo program, or now through this narrow-banking program, it's likely to more quickly transmit the interest on reserves to the wider economy.