The Bank of International Settlements (BIS) this week published a notable report on LIBOR replacement. No one can be blamed at this stage for fatigue on the topic, but the BIS report is nonetheless worth a read.
Here’s why: The authors show that while a secured transactions-based replacement rate should provide a robust and accurate representation of overnight interest rates, such a rate would largely be untethered from bank funding costs. This rate would be tied instead, at least in part, to the demand-and-supply dynamics of collateral in the repo market.
These points may already be understood, at least in principle. What’s helpful in the BIS report are the illustrations of what these traits mean in practice. The authors approximate an estimated SOFR pricing series for the financial crisis era and find that the scarcity of repo collateral during this flight-to-safety period drove overnight repo rates down and, therefore, resulted in a SOFR of at or close to zero precisely during periods when bank funding costs and LIBOR spiked.
More recently, repo market dynamics again drove SOFR pricing when the General Collateral Finance repo rate spiked on December 31, 2018, and SOFR followed higher by more than 50 bps. Quarter-end positioning at banks and a heavy Treasury issuance calendar explained the move up.
Both episodes provide thought-provoking points comparing SOFR and LIBOR. Compared to LIBOR, SOFR removes the credit-pricing component linked to bank wholesale funding costs, and, as a trade-off for referencing a broad set of actual transactions, follows the demand-and-supply dynamics in the repo market.