Rates Spark: Fed’s balance sheet glides lower, again

The Federal Reserve's balance sheet has fallen for the second week running, down by another USD75bn, to stand at a little over USD7trn. This is still a huge balance sheet when compared with an economy of around USD20trn, but the fall goes against worries in some quarters that it was on a continuous rising process in coming months, and potentially years.

The Fed is still buying bonds, an additional USD36bn for the latest week. But this is more than offset by a USD47bn fall in repo support. Confirmation of reductions in dollar swap lines to global central banks can also be gleaned, now down to USD275bn, having been as high as USD450bn only a month ago. This is indicative of reduced stress in the system, correlating with reduced demand for the safety of USD.

The Fed also continues to reduce its spending on support for money market funds, down by an additional USD2bn. There has been an equal but opposite increase in spending on the paycheck protection programme though. Spending on the latter Main Street programme now stands at USD60bn, versus USD23bn on the money market fund, as a further indication of an increased Fed focus on Main Street, and away from the need to support Wall Street.

The Fed started buying corporate bonds direct on Tuesday last week. Before that they had been exclusively buying corporates through ETFs, some 17% of which was in high yield. News that the Fed would commence buying corporate bonds outright caused a stir, and led to a risk-on reaction. We had questioned this outsized positive reaction, as there was no suggestion that this meant a material increase in Fed buying in the corporate space.

Latest data confirm our reservations. These show that there was no material change in volumes bought in the corporate space which for the latest week ran at USD1.8bn. This is up on the previous week's USD1.3bn, but clearly not by much. Extrapolating this type of buying would see the Fed buying a cumulative that is not too deviant from the USD50bn to USD75bn range. This is not nothing, but is a fraction of the full potential of closer to USD750bn.

All that the Fed has done is substitute (some) ETF buying with corporate bond buying, with the latter done in a way where the Fed replicates a balanced market-wide portfolio of corporate bonds. In the end this is not much different from buying ETFs, as ETF funds themselves would match Fed buying by buying corporate bonds direct in any case. It would have been a significant switch if it had meant bigger volumes, but the hard data tells us that has not been the case.

More evidence here then of reduced need for the Fed to aggressively support the market; which makes sense given the self-healing that has occurred in the corprate bond space (underpinned by the record primary market activity seen). Our early observations on the corporate buying programme, as outlined here, have not materially changed - the positive effect from Fed buying is more qualitative than quantitative. Big investor inflows have had the bigger effect.

In Europe, a fairly quiet events slate today might be an opportunity to reflect on the prospect of further central bank intervention given the rapidly deteriorating state of risk sentiment. The minutes of the June 4th ECB meeting are adding some elements to this. As our economics team noted, the minutes seemed aimed at addressing the concerns of the German constitutional court’s ruling on PSPP. Together with reports that the governing council decided on a course of action this week, including routing some documents to the court, this should remove lingering doubts about the ECB’s ability and willingness to act if necessary.

The first half of the year drawing to a close, a good time to take stock of the ECB's interventions in government bond markets. To date Eurozone countries and supranational agencies have issued more than €775bn in EUR bonds. Compared to previous years that is almost €200bn more than ususally printed by this time of the year and not far from the €980bn issued on average per year since 2017. That is a notably steeper trajectory of bond issuance, but at the same time the ECB stepped up its purchases via PEPP and PSPP: We estimate that the ECB has bought just shy of €400bn in the public sector to date via the two programmes, absorbing more than just the new funding for crisis measures (for the chart above we have assumed that 80% of PEPP purchases are directed at the public sector).

Detailed data available for those programmes had shown that the ECB has skewed purchases heavily towards buying bonds of peripheral issuers, deviating decidedly from its capital key benchmark for purchase allocations. Yesterday's minutes indicated that this deviation could go on for longer, as it was noted that the reinvestment period was extented until at least end-2022, and could be used to "to reduce possible deviations from the capital key that might arise during the net purchase phase."

May US income and spending are arguably the main data point of an otherwise quiet calendar. Eurozone M3 growth is also due.