After raising interest rates in March, the Fed began to shrink its balance sheet in June. The next two key time points are when to stop raising interest rates and when to stop shrinking its balance sheet. This depends on the pace of interest rate hikes and balance sheet reductions and the performance of the real economy. As inflationary pressures are still high and the normalization process is still lagging behind, the market expects the Fed to accelerate the pace of normalization, such as a larger rate hike or faster balance sheet reduction. According to the normalization principles and plans issued by the Federal Reserve, it is more feasible to raise interest rates faster, while the progress of shrinking the balance sheet is limited by the number of securities due (MBS also depends on the scale of prepayment). In practice, the Fed can temporarily absorb reserves with other debt instruments to assist in raising interest rates.
According to historical experience and the Fed’s official principles and plans, the process of normalizing unconventional monetary policy can be summarized as “four steps”: the first step is to reduce asset purchases (Taper), at this stage, the Fed’s balance sheet is still expanding , only a marginal slowdown in speed; the second step, reinvestment, rolling purchases of all maturing assets, keeping the absolute size of total assets unchanged; the third step, raising interest rates, changing the forward guidance of zero interest rates, raising federal funds through the interest rate corridor Interest rate target range; the fourth step is to shrink the balance sheet to determine the upper limit of the monthly reduction of treasury bonds and MBS, and no rolling purchases will be made after maturity (Figure 1). In this process, the Fed will continue to convert the liability side to absorb excess reserves. The commonly used tools are reverse repurchase (Reserve Repurchase, RRP) and Term Deposit Facility (TDF) – the former is Lord, the latter is secondary.
Reduce purchases + absorb reserves
At the beginning of 2013, the macro conditions for exiting quantitative easing gradually matured. In September 2013, the Federal Reserve launched the RRP and TDF after the release of the Taper signal and the market consensus that the Taper would be implemented by the end of the year. The RRP and TDF can be viewed as liability conversion vehicles that serve the federal funds rate target. Because reserves are more closely related to the federal rate than total assets. Liability structure management, Taper and shrinking balance sheets are all for interest rate hikes. In terms of liability structure management, RRP plays a more important role. Not only is there a wider range of participants, but it is also used more frequently. At the peak, the scale of reserves converted by RRP reached US$640 billion in a single month. TDF is mainly aimed at commercial banks. Commercial banks can directly convert the reserves on the asset side into RRP or TDF. Money market funds can participate in the reverse repo business, which is reflected in an increase in reverse repo and a decrease in deposits on their own balance sheets (Figure 2).
Since Taper didn’t end until October 2014, the Fed continued to expand its balance sheet from September 2013 to October 2014. Although the scale of RRP and TDF is also increasing, it is not enough to fully hedge the reserves generated by the expansion of the balance sheet, so the two (RRP+TDF and reserves) rise simultaneously. Reserves peaked at $2.8 trillion when Taper ended in October 2014. During the reinvestment phase from October 2014 to October 2017, the reserves fell to $1.9 trillion at the end of 2016, while the size of the RRP increased to $570 billion. That is to say, from the end of 2013 to the end of 2016, the RRP “diverted” nearly $500 billion in reserves.
Tradeoffs on the first rate hike
The Fed is hesitant to decide when to raise interest rates for the first time. Why is the first rate hike so late?
In the first half of Bernanke’s term and Yellen’s term, the Fed was accustomed to attributing the long recovery after the big crisis to some internal and external headwinds, such as the European debt crisis and fiscal austerity. From 2015 to 2016, changes in the external environment were indeed factors that hindered the Fed’s interest rate hike. For example, the devaluation of the RMB after the “8.11” exchange rate reform in 2015 and the Brexit referendum in 2016 had affected the Fed’s interest rate hike. decision. However, in the mid-to-late period of Yellen’s term, the Fed also began to think about the potential growth of the US economy from the supply side, which is determined by labor supply and labor productivity, which in turn depends on technological progress, the organic composition of capital, and the efficiency of factor allocation. If potential growth falls, so does the natural (or neutral) rate. This means that monetary policy is tightening even as the federal funds rate remains unchanged—how loose monetary policy is determined by the difference between the federal funds rate and the natural rate, not the absolute level of the federal funds rate.
Another equally important issue is the estimate of the natural rate of unemployment, which influences the Fed’s judgment of the tightness of the labor market and thus directly determines the monetary policy stance. The 6.5% unemployment level is a threshold for the Fed to think about raising interest rates. Unemployment fell to 6.2% in April 2014, two months after Yellen took over as Fed chair, and six months before Taper ends. Before the March 2015 regular meeting, the unemployment rate had fallen to 5.4%, just 0.2 percentage points above the Fed’s estimated natural rate of unemployment (5.2%), but it took nine months for the Fed to start raising interest rates. The pace of Fed rate hikes lags significantly behind the rate dot plot projections—two in 2015 and four in 2016 (25 basis points each).
Figure 1: Normalized order and rhythm
Source: CEIC, Orient Securities Wealth Research. Graphics: Yu Zongwen
It is not the absolute level of the unemployment rate that reflects the tightness of the labor market, but the difference between the unemployment rate and the natural rate of unemployment. The larger the difference, the looser the labor market and the less likely the monetary policy will be tightened. Underpinning this behavioral rule is still the Phillips curve. In general, because of the lag in monetary policy, the Fed should act early before the unemployment rate falls to the natural rate. Only in this way can inflation be suppressed. This is typical of the Volcker-Greenspan era. It is precisely because of this that the Fed has kept inflation expectations firmly around its 2% target, adding to the credibility of monetary policy. But since the “Great Moderation” era, the flattening of the Phillips curve has eased the urgency for the Fed to “go against the wind.” And, if the natural rate of unemployment is falling at the same time as the unemployment rate, then the unemployment rate overestimates the degree of slack in the labor market. That was the reality during the Great Recession, due to rising educational attainment, aging, and more flexible jobs. Therefore, the Federal Reserve continuously revises the forecast of the natural unemployment rate downwards in the macro forecast (SEP). As a labor economist, Yellen is well aware of the structural changes in the labor market, and also understands that there will be errors in the estimation of the natural unemployment rate. It cannot rely on any single indicator, but observes a package of indicators: the employment rate of the working-age population, labor force participation rate, the number of part-time workers willing to work full-time, and the voluntary quit rate, etc., to grasp the labor diagnosis labor market from multiple dimensions. This tradition remains to this day. So, the Fed calls maximum employment a “broad and inclusive goal.” Of course, this also reflects part of the political demands.
The Fed especially emphasized that the constraint of zero interest rate on monetary policy is asymmetric. When inflation pressure is not high, the Fed can patiently wait for more positive labor market recovery signals. Because once inflation exceeds a reasonable target range and is not temporary, the Fed has unlimited room to raise interest rates. Conversely, if the Fed tightens monetary policy too quickly when the labor market recovery is not yet sound, the Fed will be “shooting itself in the foot.” It doesn’t have enough room to cut rates to deal with the situation.
Therefore, the Fed decided to raise interest rates as late as December 2015, and the upper limit of the federal funds rate target was increased from 0.25% to 0.5%. The unemployment rate fell to 5.1% in November, and core CPI and core PCE inflation rebounded to 2% and 1.2%, respectively. One year later, on December 15, 2016, the Fed raised interest rates again by 25 basis points. After a total of seven rate hikes (25 basis points each) in 2017 (3 times) and 2018 (4 times), the federal funds target range had risen to 2.25%-2.5% by the end of 2018.
All four rate hikes in 2018 occurred during Powell’s tenure. In November 2017, Trump announced that he would nominate Jerome Powell as Fed chair, largely because he was more supportive of financial deregulation than Yellen. With the fermentation of the Sino-US trade conflict, US stocks continued to retreat after the fall of 2018. December 2018 was the worst December for U.S. stocks since 1931, with the Dow Jones Industrial Average falling 8.3% in the two weeks before the regular session. The rapid pace of interest rate hikes is an important explanation. The December rate dot plot foreshadows further rate hikes in 2019. Although Powell sent a “dovish” signal in his statement and repeatedly stated publicly that the Fed will be more patient in raising interest rates and monetary policy will be adaptable. The January 2019 regular meeting will keep interest rates unchanged. The March rate dot plot shows no rate hikes in 2019 (only one rate hike in 2020). Market confidence gradually recovered. The term structure of U.S. Treasury bond interest rates continued to invert from May to June 2019, indicating a tight monetary policy stance, indicating an increasing probability of economic recession. In a speech on June 4, Powell first hinted at a rate cut. In July 2019, the Federal Reserve officially started a new cycle of rate cuts. At this time, the table is still shrinking, but the speed has slowed down. After October, signs of recession and financial instability largely disappeared. It was a “soft landing” in the Powell era.
Figure 2: Fed Absorbs Reserves Using Fixed RRP and TDF
Source: FED, CEIC, Orient Securities Wealth Research
Start and end points of the abbreviated table
The September 2014 Principles and Plans stated that “the exact timing will depend on economic and financial conditions and the evolution of the economic outlook” on the level of interest rate hikes before the balance sheet begins to shrink. Bernanke’s view is: “Once the federal funds rate rises enough to create some room for rate cuts if necessary, it is possible to start shrinking the balance sheet”. These are subjective criteria. According to the author’s observation, when the federal funds rate rises to the established path, the Fed will begin to shrink its balance sheet. The so-called “established path” can be either a prediction of an interest rate dot plot or some type of “Taylor rule”. Shrinking the balance sheet is an aid to raising interest rates, and it can also have the same effect as raising interest rates. The key is where the appropriate interest rate level is.
In October 2017, after the target range of the federal funds rate was raised to 1%-1.25% (4 times of interest rate hikes), the Fed began to shrink its balance sheet: the starting point for the monthly reduction of treasury bonds was 6 billion US dollars, and the quarterly increase was 60%. $30 billion cap; Institutional MBS starts at $4 billion, increasing by $4 billion quarterly, capping $20 billion. In October 2018, the Fed began to shrink its balance sheet by the $50 billion/month cap. As the pace of interest rate hikes has slowed, the FOMC decided at its regular meeting in March 2019 to slow down the reduction of the balance sheet in May and stop in September. At that time, the total assets of the Fed dropped from $4.5 trillion at the peak to $3.7 trillion, and the proportion of gross domestic product (GDP) dropped from 25% at the peak to 17.5%. Among them, national debt dropped from $2.5 trillion to $2 trillion, and MBS dropped from $1.8 trillion to $1.5 trillion. The size of reserves has also dropped from $2.8 trillion to $1.4 trillion, and the proportion of GDP has dropped from 16% at the peak to 7% – providing a reference indicator for when the shrinking of the balance sheet will end in the post-pandemic era.
Since there is no experience to follow, the principles and plans do not make it clear when the balance sheet shrinking will end, saying only: “In the long run, the Fed will not hold more securities than is necessary to effectively conduct monetary policy, and it will mainly holding Treasuries, thereby minimizing the impact of the Fed’s holdings on the credit allocation of various sectors of the economy.” According to the 2016 annual report on open market operations, total assets should fall to $3.2 trillion in the base case, about 15% of GDP. The long-term equilibrium level of reserves is about $500 billion, to be reached around 2020-2021. According to the results of the Federal Reserve’s survey of commercial banks, the “minimum comfortable reserve size” of banks is about 800 billion to 900 billion US dollars. Add in some buffers, and a decent reserve would be around $100 billion (or a little more). It can be seen that, like the early termination of interest rate hikes, the termination of the balance sheet reduction plan is also slightly ahead of schedule.
In hindsight, it was correct to terminate the shrinkage early. Because, starting from mid-September, the liquidity in the money market has become increasingly tight due to the impact of national debt issuance and corporate tax payments. EFFR once exceeded the upper limit of the range (2.25%) by 5 basis points, and the AA-rated non-financial corporate overnight commercial paper rate and the repo rate jumped to 4.3% and 5.25%, respectively. The Federal Reserve immediately began to purchase Treasury bills, restarted the repurchase facility, and injected reserves into the market, marking the end of the process of shrinking the balance sheet. The federal funds rate is back in the target range. The balance sheet enters the stage of “organic growth”. As of January 2020, before the outbreak of the new crown pneumonia epidemic, the Fed’s total assets had recovered to $4.2 trillion, and its reserves had increased to $1.6 trillion.
Why are short-term rates still outside the target range when reserves are so abundant? Why did the arbitrage mechanism fail? – Banks can fully arbitrage between the 2% interest rate on reserve balances (IORB) and the 10% interest rate for overnight repurchase operations (ON Repo). In fact, the so-called sufficient reserve size can only be found through trial and error, and no a priori rule can provide a standard answer. Not to mention it is a dynamic concept. Perhaps it is not the arbitrage mechanism that has failed, but the Fed’s estimate is wrong. According to Bernanke, “perhaps due to concerns about financial regulation, banks felt that there were not enough reserves in September 2019, at least not enough to lend in the repo market. The Fed’s balance sheet shrinks. , and the resulting decline in bank reserves, has gone too far.” Whether it’s new financial regulatory requirements, or precautionary needs from banks, or the Fed’s new framework for operating rates, all mean larger Fed assets Balance sheet and reserve size.
On the whole, although Yellen delayed the timing of raising interest rates and shrinking the balance sheet and slowed down the pace, her dominant exit strategy still inherited Greenspan’s pre-emptive thinking, especially compared to the post-pandemic situation. In terms of times. Low interest rates boost risk appetite and, along with excess liquidity, fuel the inflation of asset price bubbles. So the Fed should act before private sector funding needs recover. If you wait until the economy is growing strongly and demand for private-sector loans recovers, the balance sheet reduction could lead to a spike in interest rates. That’s why the Fed started scaling back asset purchases when inflation was just 1.1%, and it started raising rates when core PCE inflation was just 1.2%. In her first rate hike in December 2015, Yellen explained: “If the FOMC delays the start of policy normalization for too long, we may end up having to tighten policy abruptly at some point to prevent The economy is overheating and inflation is well above our target. Such a sudden tightening could increase the risk of pushing the economy into a recession.”
Easing inflationary pressures and a flattening of the Phillips curve does increase room for monetary policy easing, but the Fed must also Focusing on financial stability, this policy objective is not reflected in the “dual mandate”. In addition, early action can provide a first-mover advantage. One side effect of the Fed’s fears of raising interest rates is to trigger a rise in interest rates in the bond market, especially long-term rates. This will increase financing costs and be detrimental to the real economy. Compared with other countries, raising interest rates in advance will attract international funds to flow into the United States, which will play a certain buffering role.
Post-pandemic era: V-shaped recovery and the Fed’s ‘sudden stop’
The “long recovery” in the post-crisis era is the basis for the Fed’s methodical normalization push (Figure 3). The post-pandemic era is markedly different. The V-shaped recovery of the economy requires the Federal Reserve to accelerate the pace of normalization.
The Fed took only five months to complete Taper (November 2021-March 2022) and immediately began raising rates, eliminating the “reinvestment” phase. In March and May, interest rates were raised twice in a row by a total of 75 basis points, and the target range was raised from 0%-0.25% to 0.75%-1%. At its regular meeting in May, the Fed decided to start shrinking its balance sheet in early June. As the May inflation data exceeded expectations and hit a record high, the market expects the Fed to raise interest rates by 50 basis points (75 basis points are not excluded) in the regular meeting in mid-June, and 75 basis points in the regular meeting at the end of July. No matter from the rhythm of the rate hike or the time interval between different links, this normalization is a “sudden stop”. Even so, the Fed is lagging behind. According to our fitted new Taylor rule, the federal funds rate, which is appropriate for current unemployment and inflation, is about 3%.
Generally speaking, in the process of raising interest rates, the short-term interest rate rises faster, which leads to an inversion of the interest rate term structure, which indicates the risk of recession in the future. Of course, shrinking balance sheets and forward guidance can increase term premiums and mitigate inversion risks. Alternatively, if inflationary pressures do not ease in the short term, long-term interest rates will reflect a higher inflation risk premium. This reflects stagflation expectations. In our view, the most likely deductive path is from stagflation to recession, as inflation is still dominated by the supply side. The Fed’s process of raising interest rates and shrinking its balance sheet is still in the early stages, and the contraction will accelerate in the future until the inflation inflection point is established, and there are more and more recession signals, such as rising unemployment, declining consumer momentum or investment demand, or shrinking overseas demand or financial markets. Only when risks accumulate, the Fed will slow down the pace of normalization. For the United States, the risks of a hard economic landing are much higher than those of a soft landing. According to research by former U.S. Treasury Secretary Summers, if monetary policy tightens under the inflation and unemployment levels in the first quarter of 2022, the probability of a recession in the next 12 months (Q1 2022-Q1 2023) is 100% .