Focus
March 04, 2022 | 13:02
QT on the QT
Economics
Economics
Fixed Income Strategy
Economics
Economics
Economics
Economics
Economics
Economics
Fixed Income Strategy
Economics
Economics
Economics
Economics
QT on the QTQT is both an acronym for ‘quantitative tightening’ and an abbreviation for ‘quiet’. |
|
Despite the heightened risks to economic outlooks owing to Russia’s invasion of Ukraine, March is the month for policy rate liftoff on both sides of the Canada-U.S. border. The Bank of Canada hiked 25 bps on March 2, and the Fed is primed to do so on March 16. In congressional testimony (March 2), Chair Powell said that it was “still appropriate to raise rates” as signalled, despite the conflict in Eastern Europe. Both the Fed and BoC have indicated that quantitative tightening (QT) would be launched soon after rate hikes have begun. The Bank’s statement said it will now “be considering when to end the reinvestment phase and allow its holdings of Government of Canada bonds to begin to shrink”. Since October, the BoC has been buying $4-to-$5 billion per month to keep its holdings unchanged. The purchase pace was calibrated to match the large and unevenly spaced maturities over time. In the next-day speech and presser, Governor Macklem said: “When we initiate QT, we will stop purchasing… We do not intend to actively sell bonds.” Along with a follow-up rate hike, we’re bracing for the BoC's announcement of QT’s start at the April 13 meeting, which will immediately end reinvestment. Amid the FOMC’s upcoming policy pronouncements, we’re anticipating more insight into the structure and timing of the Fed’s QT plan, but Powell indicated in the recent testimony that it probably won’t be finalized at this meeting. (For the record, we expect it to be in effect by July 1.) However, from the guiding principles released at the last FOMC confab, we already know the Fed intends to reduce its securities holdings “over time in a predictable manner primarily by adjusting the amounts reinvested of principal payments”. The word ‘sell’ was not found anywhere. There is an interesting asymmetry here. Both central banks moved vociferously into quantitative easing (QE) via outright purchases of securities. It was the Bank’s first foray into QE, while the Fed's well-versed in securities buying with three rounds tucked under its belt (2008-2014). However, they have both indicated less strident styles to QT, by excluding outright sales. Powell said QT should run “in the background”. His predecessor, Janet Yellen, said QT should be like “watching paint dry”. Why the asymmetry? In the January 26 presser, Powell said: “Balance sheet [policy] is... still a relatively new thing for... the markets and for us, so we’re less certain about that.” Given this uncertainty, along with the policy bias amid economic emergencies (such as during global financial crises and pandemics) to put minimizing deflation risk paramount, QE is bound to play the policy lion compared to QT’s lamb. Also, QE occurs after policy rates have fallen to their effective lower bound (ELB), and the bond buying acts as a substitute for further rate cuts. However, QT occurs after policy rates are back above their ELB, and any bond selling acts as either a substitute for further rate hikes or a complement to them. On the latter point, a scenario in which central banks want to magnify the power of their rate hikes to an uncertain magnification, when they can simply raise rates more aggressively, seems unlikely. While bond selling in lieu of rate hikes is a distinct possibility, central banks prefer having the focus of attention on their policy rates when adjusting and communicating monetary policy. Hence, the predilection for loud QE and quiet QT. QT ≠ -QE… In the absence of bond selling, QT doesn’t work as effectively, or at all, in lifting bond yields via the channels QE worked through to lower them. First, central bank bond buying reduced the remaining supply of bonds relative to investor demand, thus contributing to higher prices and lower yields. The opposite won’t be occurring with QT. But, the absence of demand from central bank rollover should apply some upward pressure on bond yields (other things equal). How governments refinance these absent rollovers would focus this pressure on benchmark bond and bill yields, specifically. Second, the above bond buying skew reduced remaining aggregate duration, contributing to lower term premiums and yields. Again, the opposite won’t be occurring with QT. The absence of rollovers should have little impact on aggregate duration because the rolled-over amounts tended to be allocated across sectors roughly in line with their market weights. Third, QE played a forward guidance role in signalling that policy rates would be ‘lower for longer’; i.e., that rates wouldn’t rise until at least QE ended. In the current situation, the Fed quickened the QE tapering pace to signal (make room for) a March rate hike. However, quiet QT is unlikely to affect market rate hike expectations that much. |
QT, Episode 1… Compared to QE, QT (as framed) is bound to have less influence on bond yields, but it should still have some impact. This was evident, for a while, during the Fed’s first QT episode (QT1) (Chart 1). Accompanying a policy rate hike, the FOMC first detailed its QT plan on June 14, 2017, two meetings or three months before it began. Compared to the day before the plan was announced, intra-episode yield highs were hit in early November 2018 with 2-year yields up 160 bps and 10-year tenors up just over 100 bps. While QT contributed to some of this, the increases reflected more than 125 bps worth of Fed rate hikes in the interim along with a booming stock market that was initially prodded by the Tax Cuts and Jobs Act. And, afterwards, as fears of an escalating U.S.-China trade war calmed. The latter fears eventually returned to haunt the stock market, which, along with the Fed’s final rate hike (December 19, 2018), caused stocks to sell off sharply and bonds to rally bigtime. |
|
On March 20, 2019, QT was scaled back and scheduled to end in October. The Fed ended QT two months early on July 31, accompanied by a rate cut. On the policy easing, it became clear that a 2.375% fed funds target was proving to be too onerous for the economy, particularly given the ongoing trade war with China (the manufacturing sector spent the final five months of the year in recession). On the early end, QT-reduced bank reserves were increasingly becoming ‘scarce’, causing pressure in overnight funding markets. The issue came to a head mid-September when SOFR topped 5% at one point. The Fed started regular (technical) T-bill purchases the next month to rebuild bank reserves, which continued until QE ramped up in the wake of the pandemic. Some Symmetry... Despite their different effects on interest rates, QE and QT have comparable impacts on central bank balance sheets and liquidity in the banking system. QE purchases, a central bank asset, are paid for by creating bank reserves (Fed) or settlement balances (BoC), which are liabilities. The balance sheet and liquidity grow. But, QT involves an extra step. When matured bonds are redeemed (lowers assets), government deposit balances are reduced (lowers liabilities). The balance sheet shrinks. Then, as the government refinances the redemptions, their deposit balances are replenished at the expense of lower reserves or settlement balances. Liquidity shrinks. Along with the interest rate channel, central banks’ increasing and decreasing of assets and consequent expansion and contraction of liquidity also influence financial conditions. Recent Fed policy statements, for example, assert that “ongoing purchases and holdings of securities will continue to foster smooth market functioning and accommodative financial conditions”. Thus, as assets decrease and liquidity drains, QT should have a negative effect on financial conditions. |
Both the Fed and BoC boast large balance sheets supporting lots of liquidity (Chart 2), so QT has its work cut out for it. One barometer of ‘excess’ liquidity is the degree to which both bellwether overnight interest rates, the effective fed funds rate (EFFR) and the Canadian overnight repo rate average (CORRA), trade below their respective targets. On March 3, EFFR was 8 bps, compared to the 12.5-bp midpoint of the Fed’s 0%-to-0.25% target range. CORRA was 43 bps versus the Bank’s 50-bp overnight target. Run ’n’ roll… With outright sales not part of QT plans (for the time being), the amount of maturing securities not being rolled over, a.k.a. the ‘cap’, will be the valve to increase upward pressure on bond yields. The Bank of Canada is going to allow “full run-off” (Chart 3), unlike the Fed. Meanwhile, some Fed officials have been talking about “big caps”. The Fed’s previous QT plan began with caps of $6 billion for Treasuries and $4 billion for MBS. They were raised every three months by $6 billon and $4 billion, respectively, until they reached their final levels of $30 billion and $20 billion after a year. However, compared to QT1, holdings of Treasuries are starting out 133% higher with MBS holdings up 52%. Our working assumption is for caps to commence at $12 billion for Treasuries and $6 billion for MBS, increasing by these amounts every two months until reaching their final levels of $60 billion and $30 billion, respectively, after eight months. Given the profile for the Fed’s maturities of Treasuries (Chart 4), the net risk is for a higher cap profile and a shorter phase-in period, given that there are some meaty maturities looming. For MBS principal payments, the latest monthly amount is just over $40 billion, but it was topping $65 several months ago. The pullback reflects the impact of rising mortgage rates on refinancing activity. With higher longer-term interest rates likely looming, contributing to cooler refinancing activity and potentially dampened home sales, principal payments will probably slip further. An eventual $30 billion cap could potentially cover all principal payments. For Sale? Outright sales have not been completely ruled out by either central bank. The Bank of Canada said it was not going to “actively” sell bonds. The Fed said it would shrink its balance sheet “primarily” by adjusting reinvestment. Indeed, with the Fed wanting to hold only Treasuries over time, and MBS holdings (even at full roll-off) likely to take a long time to recede, the FOMC is entertaining the notion of selling MBS outright to expedite the process. |
|
However, as both central banks face pressing inflation problems and perceptions of being ‘behind the curve’, they are sounding increasingly hawkish on the rate hike front. Powell and Macklem have both not ruled out 50-bp moves. Powell has even admitted the Fed “should have moved earlier”. As such, the risk of the yield curve (2s-10s) inverting earlier in this tightening cycle is mounting. Both central banks, at least at this stage, likely don’t want to cause a recession to bring inflation down. And, an inverted curve would not only signal recession in the market’s mind, but it would also likely contribute to one directly by encumbering the credit creation process. Substituting outright bond sales for some rate hikes would apply pressure for the yield curve to re-steepen. Bottom Line: On both sides of the Canada-U.S. border, coming QT (as framed) should apply upward pressure on yields along the curve, complementing central bank rate hikes. Furthermore, QT should have a negative effect on broader financial conditions. Although the magnitudes of these pressures and effects are uncertain, the larger they are, the more they’ll limit cumulative rates hikes in the policy battle against inflation. |