Two Views of the Repo Crunch
Monday, September 16th, saw a large spike in overnight repo rates, with disruptions in the market still ongoing. What has caused this reaction? First, an explanation of the overnight repo market is necessary. This market gives banks and companies access to short-term dollar funding through the use of repurchase agreements, where borrowers sell collateral in the form of Treasuries and pledge to buy it back at a slightly higher price that determines the interest rate. This rate should be kept in line with the interest on reserves (IOR) that the Federal Reserve pays banks through simple arbitrage. However, the repo rate blew past the IOR rate the previous Monday, requiring Fed intervention in the form of open-market operations that injected additional reserves into the market. The Fed has been continuing this ever since to calm the market and bring the repo rate within the current target range of 1.75-2.00%.
The repo rate is determined through the supply and demand for reserves. Due to three previous quantitative easing programs that vastly expanded the Fed's balance sheet, the market was operating in a regime of reserve abundance. In this regime, the rate for reserves is set by the Fed using the IOR, as demand for reserves is completely satiated. However, the program of quantitative tightening reduced the supply of reserves held by banks by letting Treasuries held by the Fed slowly expire. The minimum level of reserves necessary for the Fed to operate in the regime of reserve abundance is unknown, but the recent instability indicates that reserves are becoming scarce. In addition, a supply shock caused by a settlement date for corporate taxes with the Treasury reduced the quantity of reserves in the market.
The increasing scarcity of reserves set the conditions for the crunch, but what was its root cause? Why were banks unable expand their lending in the face of higher repo rates, causing the rates to spiral upward? One view is that an increase in demand for reserves has led banks to eschew lending in the repo market. As detailed in this blog post, banks are being constrained by post-crisis regulations to prefer reserves over Treasuries on their balance sheet. Although one measure, the liquidity coverage ratio, counts both reserves and Treasuries as high-quality liquid assets, other requirements cause banks to prefer holding reserves over Treasuries. This inability to substitute reserves for Treasuries prevented banks from arbitraging higher repo rates using their reserve holdings.
Another view, expressed in this post, is that the uncapped foreign RRP facility offered by the Fed is sucking in reserves and reducing lending in the repo market. The reverse-repurchase market allows banks to swap reserves for safe collateral such as Treasuries. As interest rates in the United States are above levels elsewhere in the world, foreign banks are increasingly using the RRP facility to gain higher returns. In addition, the inverted yield curve has sent banks looking for greater yields by lending on the short end of the curve. The RRP market creates a positive feedback loop, as the subtraction of reserves for collateral increases repo rates, which sends more reserves streaming into the facility. The overall effect is to reduce the supply of reserves on the market, which leads eventually to a situation of reserve scarcity and the current repo crunch.
What can the Fed do in the short and long term to ease this situation? Its current policy of OMOs which inject reserves into the market has helped send rates back down to the Fed's target range. The two views of the causes of the crisis do not preclude one another. However, the correct long-term solution requires an understanding of which factors are most affecting repo rates. If the problem is an increased structural demand for reserves, one solution is for the Fed to institute a standing repo facility, detailed here. This facility will allow banks to automatically borrow reserves in the Fed using a repurchase agreement when in need. This will prevent shocks in reserve supply from transmitting to repo rates as the Fed will be on standby to inject reserves. However, if the problem is the uncapped RRP facility further reducing the supply of reserves, instituting a standing repo facility might not be wise, as it will further encourage banks to use the RRP facility by providing an unlimited source of reserves. Instead, one solution might be to cut rates to revert the curve to an upward slope, which would reduce inflows to the foreign RRP facility and result in a greater supply of reserves. Any long-term structural fix instituted by the Fed requires a careful study of the causal factors behind the repo market spike, which will help determine the best path forward.
The repo rate is determined through the supply and demand for reserves. Due to three previous quantitative easing programs that vastly expanded the Fed's balance sheet, the market was operating in a regime of reserve abundance. In this regime, the rate for reserves is set by the Fed using the IOR, as demand for reserves is completely satiated. However, the program of quantitative tightening reduced the supply of reserves held by banks by letting Treasuries held by the Fed slowly expire. The minimum level of reserves necessary for the Fed to operate in the regime of reserve abundance is unknown, but the recent instability indicates that reserves are becoming scarce. In addition, a supply shock caused by a settlement date for corporate taxes with the Treasury reduced the quantity of reserves in the market.
The increasing scarcity of reserves set the conditions for the crunch, but what was its root cause? Why were banks unable expand their lending in the face of higher repo rates, causing the rates to spiral upward? One view is that an increase in demand for reserves has led banks to eschew lending in the repo market. As detailed in this blog post, banks are being constrained by post-crisis regulations to prefer reserves over Treasuries on their balance sheet. Although one measure, the liquidity coverage ratio, counts both reserves and Treasuries as high-quality liquid assets, other requirements cause banks to prefer holding reserves over Treasuries. This inability to substitute reserves for Treasuries prevented banks from arbitraging higher repo rates using their reserve holdings.
Another view, expressed in this post, is that the uncapped foreign RRP facility offered by the Fed is sucking in reserves and reducing lending in the repo market. The reverse-repurchase market allows banks to swap reserves for safe collateral such as Treasuries. As interest rates in the United States are above levels elsewhere in the world, foreign banks are increasingly using the RRP facility to gain higher returns. In addition, the inverted yield curve has sent banks looking for greater yields by lending on the short end of the curve. The RRP market creates a positive feedback loop, as the subtraction of reserves for collateral increases repo rates, which sends more reserves streaming into the facility. The overall effect is to reduce the supply of reserves on the market, which leads eventually to a situation of reserve scarcity and the current repo crunch.
What can the Fed do in the short and long term to ease this situation? Its current policy of OMOs which inject reserves into the market has helped send rates back down to the Fed's target range. The two views of the causes of the crisis do not preclude one another. However, the correct long-term solution requires an understanding of which factors are most affecting repo rates. If the problem is an increased structural demand for reserves, one solution is for the Fed to institute a standing repo facility, detailed here. This facility will allow banks to automatically borrow reserves in the Fed using a repurchase agreement when in need. This will prevent shocks in reserve supply from transmitting to repo rates as the Fed will be on standby to inject reserves. However, if the problem is the uncapped RRP facility further reducing the supply of reserves, instituting a standing repo facility might not be wise, as it will further encourage banks to use the RRP facility by providing an unlimited source of reserves. Instead, one solution might be to cut rates to revert the curve to an upward slope, which would reduce inflows to the foreign RRP facility and result in a greater supply of reserves. Any long-term structural fix instituted by the Fed requires a careful study of the causal factors behind the repo market spike, which will help determine the best path forward.
Comments
Post a Comment