QT impact: has the biting point already arrived?
Yvan Berthoux - 21 February 2019
Over the course of 2018, many practitioners were worried about the market impacts of central banks’ net purchases turning negative. The first reaction to the Fed’s balance sheet shrinkage plan, with US$450bn to US$500bn of US Treasuries and MBS expected to run off this year, was that it would immediately damage equity prices. Figure 1 (left frame) shows the co-movement between the annual change in the Fed’s balance sheet assets and the SP500. Should we expect to see a divergence in the two times series in the future? In addition, uncertainty around the economic outlook combined with an increase in Fed balance sheet 12-month volatility could both increase the term premium sharply and therefore create another ‘taper tantrum’ episode as seen in 2013. Looking at the global picture, there is no question that quantitative tightening coupled with a rising effective Federal Funds rate (EFFR) are major sources of risk for this year, especially for risky assets such as equities.
However, there has been increasing concern that reduction of the balance sheet has already reached a threshold where constraints are binding, inferring that the Fed may struggle to continue the programme when assets fall below US$3.5tr, which is roughly $500bn below the current level. What is the liquidity problem and why are reserves (held at the Fed) falling drastically?
Data Source: Eikon Retuers, Adrian et al. (2013)
In response to the global financial crisis, the Fed bought securities directly from the market to restore financial stability and regenerate economic growth and inflation. They purchased MBS and Treasuries through commercial banks, which acted as intermediaries as they were also net buyers of US Treasuries during the QE period. As a result, QE inflated the monetary base, with credit institutions’ reserves soaring from a negligible amount to peak at US$2.8tr in July 2014 (figure 2, left frame). Since October, the Fed has been paying interest on excess reserves that has usually been higher than the Effective Fed Funds (EFFR), the volume-weighted median of overnight Fed Funds transactions. However, the IOER-EFFR spread has converged back to zero this year, implying that the banks cannot benefit from the ‘seignorage profits’ offered by the Fed. Between 2014 and 2016, if banks were borrowing from the money market at EFFR to satisfy reserve requirements, they would be making a guaranteed 15 to 20bps on excess reserves.
Does it really matter whether the EFFR trades above the IOER? In theory, not. It is just that if the spread becomes positive (i.e. EFFR trading at 2.43 per cent versus the current IOER of 2.40 per cent), banks will then tap their excess reserves to meet reserves requirements and pay nothing instead of borrowing at EFFR and pay a 3bps carry. In addition, it made sense for banks to keep excess reserves at the Fed between 2012 and 2016 when the IOER was yielding higher than a 1-year Treasury (i.e. alternative risk-adjusted returns were not high enough to make these reserves flow into the economy, figure 3, right frame). However, if EFFR moves higher than the IOER and the term premium rises across the risk spectrum, some banks will also loan their excess reserves at higher rates short term and earn a higher profit. As observed in figure 2, reserves fell by US$1.1tr since their August 2014 high, resulting from a rise in higher-yielding instruments in the market.
Data Source: FRED
The problem now arises when we look at the small banks in the US, as a significant proportion is already borrowing at an overnight rate higher than the Fed’s upper boundary target. Figure 3 (left frame) shows that the 99th percentile is borrowing at 2.6 per cent (10bps above the upper band), which has very rarely happened after a rate hike. As small banks in the US are reserves poor (5 per cent of the banks in the US control 90 per cent of the excess reserves available according to JP Morgan), they will be forced to bid for required reserves at higher rates.
Data Source: NY Fed
Share on LinkedIn: