1/ U.S. banks have $1.4Tn in *excess reserves, yet liquidity problems that more QE won't fix. After years of Fed-driven speculation & debt issuance, the banking regs (LCR) created to avoid another crises now pinch at the first sign of unusual withdrawals. h/t @wesbury @LukeGromen 
@wesbury @LukeGromen 2/ Total reserves and excess reserves as a fraction of total bank assets. Compare these to pre-crisis history. To imagine that recent repo spikes reflect some inadequacy of reserves or the need for more QE is to misunderstand the nature of the constraint.
@wesbury @LukeGromen 3/ Another driver of repo spikes is lopsided reserve holdings at some banks. "Reserves are concentrated, and the excess reserves relative to the minimum level each bank is demanding is concentrated. The key question is how smoothly they are redistributed." - Lorie Logan, NY Fed
@wesbury @LukeGromen 4/ Remember also that the largest risk in bank-to-bank repos is counterparty risk. If you aren't doing interbank repos regularly, there's also administrative nuisance. Remember that even getting ~3% above IOER on an overnight repo is just 1bp, or $100k to take on $1bn in cp risk.
@wesbury @LukeGromen 5/ … and b/c repos are counterparty transactions & rate is set at the margin, if xs reserve holders aren't willing cp's, it would be futile for @FederalReserve to increase aggregate reserves w/untargeted QE. That's why it's doing a direct backup repo facility for strapped banks.
@wesbury @LukeGromen @federalreserve 6/ … I'd just set the minimum rate on the direct repo facility at something like IOER+2%. Bagehot's Rule: lend freely, but at high rates of interest, and only to institutions that are solvent and able to provide collateral for the loans. @FederalReserve
