As a consequence of the Federal Reserve’s large-scale asset purchases from 2008-14, banks’ reserve balances at the Fed have increased dramatically, rising from $10 billion in March 2008 to more than $2 trillion currently. In that new environment of abundant reserves, the FOMC put in place a framework for controlling the fed funds rate, using the interest rate that it offered to banks and a different, lower interest rate that it offered to non-banks (and banks). Now that the Fed has begun to gradually reduce its asset holdings, aggregate reserves are shrinking as well, and an important question becomes: How does a change in the level of aggregate reserves affect trading in the fed funds market? In our recent paper, we show that the answer depends not just on the aggregate size of reserve balances, as is sometimes assumed, but also on how reserves are distributed among banks. In particular, we show that a measure of the typical trade in the market known as the effective fed funds rate (EFFR) could rise above the rate paid on banks’ reserve balances if reserves remain heavily concentrated at just a few banks.