The world's central banks are engaged in a major policy reversal，playing a game of risk management. At the start of the year，the Fed – the US central bank – paused on its march towards higher interest rates. In September，it will stop shrinking its balance sheet. In March the European Central Bank became the first major developed-country central bank to provide new stimulus. It unveiled plans for fresh measures to boost the eurozone's faltering economy – less than three months after phasing out its bond-buying program. Most importantly，the 2008 crisis has forced central banks to drastically change the way they conduct monetary policy. Moreover，monetary policy framework also appears to be in need of a reassessment. Just in the past several months，a very controversial heterodox monetary theory，dubbed Modern Monetary Theory (MMT)，has become a focus of policy debate. The result of this debate is crucial not only for advanced economies，but for emerging countries like China as well.

The Fed's Awkward Turn

In September 2018，Fed officials signaled they thought they would raise short-term interest rates three times in 2019 on the way to setting a policy rate near 3.5% in 2020. Before the year-end，officials shifted their outlook，suggesting they expected between one and three increases. In January 2019，the Fed left its key policy rate in a range between 2.25% and 2.5% and indicated that it is unlikely to raise rates at all this year. Fed officials recently reiterated they have paused for now，with short-term rates just below 2.5%，and may even cease the balance sheet runoff entirely in September.

The economic data remain strong，but the tone on the pace of interest rate hikes and balance sheet reduction has softened，giving the impression that Trump's frequent "critiques" have been effective.

Historically，most US presidents have been extremely respectful of the Fed's independence and have sought to avoid second- guessing the central bank about monetary policy – at least in public. But this practice was broken by Trump，and in a stunning way. From July 19 to the end of December 2018，the Fed raised interest rates twice，but Trump lashed out at the Fed no less than 20 times. Trump insists that the Fed's rate hikes were too aggressive and posed a big threat to the economy and the securities markets. In an interview with The Washington Post on November 27，2018，Trump expressed his deep dissatisfaction with Fed Chair Jerome Powell，criticizing him for not cutting interest rates. He suggested Powell was to blame for the decline in the capital markets and plant closures and layoffs at General Motors. He went on to say that the Fed was a bigger problem than China despite the prolonged trade friction. On December 18，the day before the Fed raised interest rates again，Trump warned Powell not to make another "mistake." A few days later，Trump even threatened to fire him – a power the president does not have.

In response to Trump's attacks，Powell repeatedly stated that he would resolutely safeguard the independence of the Fed and protect it from any political interference. But in the fourth quarter of 2018，as equity markets tumbled，Trump became more and more vocal about his misgivings，and Powell had to respond to market shocks. No matter how often Powell made his case that market volatility was not the result of monetary policy，he found it difficult to dispel the doubts. In the stock market，the S&P 500 reached a high of 2,931 points on September 20 before falling to 2,351 points on December 24，for a decline of 19.8%. From December 19 when the Fed raised rates to December 24，the S&P 500 lost 7.66%. In the bond market，the 10-year treasury yield reached a 2018 peak of 3.24% on November 8，before falling back to 2.68% by year-end. The Fed raised interest rates four times in all of 2018，while the 10-year treasury yield，which was 2.48% at the beginning of the year，had risen only 20 basis points by the year-end. In addition，oil prices fell sharply and gold rose rapidly.

As market volatility intensifies，Powell's wording on monetary policy is also undergoing subtle changes. At the beginning of October last year，the Fed emphasized that the federal funds rate was still far from a neutral level，and there was considerable room for an upward adjustment. On November 28，Powell said in a speech at the New York Economic Club that the federal funds rate is near a neutral interest rate level. This statement led to market speculation that monetary policy might shift，sending the Dow Jones Industrial Average 600 points higher that day. On December 19，at the press conference after the last interest rate hike in 2018，Powell said that the Fed will closely track market changes – though markets were not the only factor in interest rate decisions. The Fed also needed to prepare for uncertainties surrounding the UK's Brexit ordeal.

At the annual meeting in January of the Allied Social Science Associations in Atlanta，Georgia，The New York Times conducted a joint interview with Powell，Janet Yellen and Ben Bernanke，three current and former Fed Chairs. Powell said that monetary policy is not pre-set and the Fed will listen carefully to the voice of the market. The central bank can adjust its policy stance at any time according to changes in the market situation and make major changes when necessary. This does not exclude the possibility of suspending policy tightening – as in 2016. Moreover，he believes that this flexibility also applies to plans to reduce the size of the balance sheet each month. The weakening of inflation has provided room for flexibility in monetary policy. In summary，the author believes，as he wrote at that time，this indicates that the Fed's monetary policy has undergone major changes，and the pace of rate hikes may slow，at least for a while. In response，the S&P 500 index rose 3.4%，while the 10-year U.S. treasury yield jumped 11 basis points immediately after Powell made his remarks.

The rise in stock indices and bond yields was also related to the strong economic data released by the US Department of Labor the same day. Employment in December 2018 increased by 312,000，much higher than the market expectations of 177,000，and annual salaries increased 3.2%. At the same time，the employment data for October and November were also revised. Although the unemployment rate rose from 3.7% to 3.9%，it was still at a low level and the increase reflected a rise in the labor participation rate from 62.9% to 63.1%.

The economic data remained strong，but the tone of monetary policy had softened. At first glance，this seems difficult to understand. But Bernanke provided a very convincing explanation. At the same meeting，Bernanke said that he was not surprised by the sharp fall in the price of assets，believing it to be a normal reaction to trade tensions and uncertainties about the US and the global economy. He went on to say that much to his surprise the market turned a blind eye to these uncertainties in the first three quarters of 2018. Compared to Powell and Yellen，Bernanke was more cautious about the prospects of the economy.

The author personally agrees with Bernanke's judgment. On January 3，Apple cut its earnings forecast sharply. This was the first time in 20 years for the company. The ISM manufacturing data of the leading US economic indicators released on the same day was weak. A decline in housing，the auto market，global trade and investment，and the low level of US labor productivity growth at 1.8% was also worrying. There is still a high level of concern over the state of the economy and global political tension.

No matter what happens next，the Fed's independence has been deeply eroded. As Adam Posen of the Peterson Institute for International Economics rightly said，arguing for and protecting central banks' operational independence must be defended.

Why Did the Fed Cut Excess Reserve Rates?

At the Federal Open Market Committee meeting that concluded on May 1 this year，the Fed held the federal funds rate in a target range of 2.25 to 2.5% and set the interest rate paid on required and excess reserve balances (the IOER rate) at 2.35%.

That same day，US stocks fell，bond yields rose and the dollar strengthened. That suggests the market thinks the Fed is more tolerant of weak inflation and not ready to cut rates.

Since the end of 2018，US inflation has been falling. Core personal consumption expenditures，the Fed's preferred index for gauging inflation，excluding food and energy，was 2% in December 2018，just around the Fed's inflation target. But it fell to 1.8% in January and slipped further to 1.6% in March. The Fed sees the decline as temporary but this is not in line with market expectations. The mixed data add to the complexity of making an accurate assessment. US household spending and business investment are slowing，and the rest of the world economy is weak.

At this meeting，there was one move that seemed to have received little attention: the IOER rate was set five basis points lower than before. Even though the cut was minimal，the signal should not be ignored.

As the benchmark interest rate，the Federal Funds Rate (FFR) is the overnight US interbank offered rate，which is a market rate. Before the financial crisis in 2008，the Fed affected the reserve balances of depositary financial institutions mainly by buying or selling treasury bonds，thus reducing or increasing the FFR and bringing it close to the Fed funds target rate，or the FFTR. However，after the outbreak of the crisis，the quantitative easing of the Federal Reserve greatly increased the size of reserves，which exceeded $4.4 trillion at the peak. The flood of funds and the lack of mutual borrowing demand by market participants made it difficult for the Fed to influence the FFR through traditional operations.

In order to prevent the FFR from deviating from policy objectives and leading to the loss of control of monetary policy，the Fed began to pay interest on excess reserves in October 2008.

The Fed now sets the IOER rate and the overnight reverse repurchase rate (ON RRP) in addition to the FFTR. It is the setting of these two rates，combined with repo operations in the open market and the encouragement of arbitrage among institutions that puts the FFR within the FFTR. (See Chart 1) In this sense，the IOER rate already has some aspects of a policy interest rate.

Since interest payments on reserves have been made，the IOER rate has been set at the upper end of the FFTR. On June 13，2018，the Fed raised the FFTR from 1.5-1.75% to 1.75-2%. According to past practice，the IOER rate would have been increased from 1.75% to 2%. In fact，the Fed only raised the IOER rate to 1.95%，five basis points below the FFTR ceiling. This was the first time in a decade. In December 2018，when the Fed raised interest rates again，it hiked the FFTR from 2-2.25% to 2.25-2.5% and the IOER rate from 2.2% to 2.4%. That left the IOER rate 10 basis points below the FFTR ceiling，up from five basis points and widening the spread further.

At the FOMC meeting on May 1，the Fed adjusted the IOER rate again.The IOER rate spread over the FFTR ceiling increased by another five basis points to 15 basis points. But there were significant differences. Whereas the previous two adjustments were made in the course of raising interest rates，this time the IOER rate was cut directly while the FFTR remained the same. So，while the IOER rate spreads over the FFTR ceiling have widened three times，the absolute level of the IOER rate had fallen for the first time.

The Fed's explanation for the IOER rate cut was that the FFR of actual transactions is often at the upper end of the FFTR and that the Fed wanted to guide it to somewhere in the middle of the FFTR range. At the post-FOMC press conference，Powell stressed that the IOER rate reduction was only a technical adjustment and that the Fed's monetary policy stance had not changed. In other words，the current interest rate was appropriate and there was no need to raise or lower the benchmark policy rate.

Powell's explanation is not hard to understand. But why would the FFR increase to the top of the FFTR? One reason is that the Fed has fewer reserves because of the balance sheet runoff，but at the same time，demand for reserves has risen dramatically instead of falling synchronously. This pressure has become so normalized that it has to be contained by widening the spread between the IOER and the FFTR cap.

Although the IOER rate reduction cannot be understood as a rate cut，future FFTR cuts are still possible. Powell maintained the fall in inflation in the first quarter of this year was temporary. If that turns out not to be the case，the pressure to cut rates will increase significantly. The Fed also discussed the possibility of changing the maturity structure of its bond holdings in the future. Given that current holdings are skewed towards the longer term，a change in direction would have to be a higher proportion of short-term paper. The effect on the yield curve，then，is that short-end rates may fall and long-end rates may rise.

Could the US See Negative Interest Rates?

Six of the world's central banks have implemented a NIR policy (See Chart 2)，but the Fed，which has always been the most innovative in monetary theory and practice，has shied away from this.

Monetary policy makers used to hold a firm belief that policy rates could not be negative，or that there is a "zero lower bound." It is said that the zero lower bound was in place because the benefit of holding cash could not be negative.

Nowadays，this seems to be a "traditional" or even outdated view as six of the world's central banks including those from Sweden，Denmark， Switzerland，Japan，Hungary and the eurozone countries，have put in place negative interest rate policies. And once the last theoretical barriers break down there will be widespread contagion. As of May 2019，the global total of negative-yield bonds was $13.56 trillion，accounting for 11.84% of the world's outstanding bonds. Japan's negative-yield bonds account for 60% of the country's outstanding bonds. (See Chart 3)

However，the Fed has been unusually quiet. Although the Fed used all its skills in response to the financial crisis of 2008，from zero interest rates，QE and "operation twist" to numerous other policies，an NIR policy so far has remained off the table.

Why is that?

One explanation is that the US economy is far stronger than the economies in Europe or Japan so there is no need. But this explanation fails to tell the whole story. Although the US recovered strongly from the global financial crisis，it still suffered serious damage. The launch of the third round of quantitative easing in late 2012 was a desperate decision made under great pressure from a serious economic situation. At that time there was no limitation on bond purchases and no declaration of when to exit.

Another explanation is that the NIR policy is unpopular. No one，whether depositors，financial institutions or borrowers，likes NIRs. Taking depositors for example，Bernanke says in his memoirs that the Fed once tried to stimulate demand by dragging down interest rates but it made the relationship with depositors nervous，and they had considerable political influence. If NIRs had been adopted，it would have been worse. However，many of the measures taken by the Fed during the crisis were unpopular. There was so much criticism and second guessing that Bernanke said he kept an index card with an inspirational quote from Abraham Lincoln to help him through the crisis.

In my opinion，the fundamental reason lies in the fact that the Fed was restricted by the monetary policy operation framework. If it had imposed an NIR policy，the Fed would have needed to give up its long-used benchmark interest rate，the FFTR. It would have been too hard to make such a big decision，especially in the midst of a crisis.

In March 2016，Bernanke also acknowledged，when writing to discuss the NIR policy，that if the IOER rate reached zero or became negative，the lending market would shrink sharply. And as a result，the FFTR benchmark would have been lost，making it impossible to serve as the Fed's policy rate. He also argued that it would not be disastrous if the FFTR was abandoned and the IOER rate was raised to the new policy rate. But this statement should be viewed as more of a theoretical discussion. The US economy was booming at the time – much like today – and we do not know whether Bernanke would be so calm in times of crisis such as a few years ago.

Bernanke has never rejected NIRs. In his view，the perception of the zero lower bound is biased. And further，he thinks the NIR policy is useful，but its usefulness should not be overstated.  Former Fed chair Yellen seems more cautious than Bernanke on this point. She is concerned about the "unintended consequences" of adopting the NIR policy.

NIRs still remain unknown and closely tracking the experience of central banks in Europe and Japan will help us to solve their mysteries. It's generally expected that the long-term normal FFR is only around 2.75%. It is clear there will be less scope for interest rate cuts as a response to recession in the future than in the past. If it's hard to eliminate the effective lower bound and nominal rates cannot go too negative，so other solutions must be found. But QE and forward guidance might not be sufficient tools，because in the event of a recession，long rates would be expected to "fall to very low levels without any help from forward guidance or QE." (See Blanchard，O.，& Summers，L. (2019). Evolution or Revolution? Rethinking macroeconomic policy after the Great Recession. Cambridge，MA: MIT Press.) So monetary policy will likely lack the space and tools to deal with the challenge. In some cases，the central bankers are boxed in，without sufficient tools available to cushion their economies. To generate expectations of higher inflation could be an alternative solution，which requires a reassessment of the current monetary policy framework.

Re-examination of Monetary Policy Framework

Ten years ago，few would have predicted that inflation would still be below target today for most of the advanced economies. This has amplified a critical question for central banks，especially the Fed，the ECB，and the BOJ: are they suffering from a "credibility" issue with inflation? The low-inflation，low-interest-rate，higher perceived risks environment in which we now live also calls into question the ability of central banks to deal with future economic downturns.

As part of the Fed's Congressionally-mandated mission of achieving price stability，the FOMC first published in January 2012 a "Statement on Longer-Run Goals and Monetary Policy Strategy" and it referred to a 2% target for inflation tied to total personal consumption expenditures. It was the first time that the Fed explicitly defined the stable-prices component of the mandate in numerical terms. The FOMC has reaffirmed the Statement each January since then.

But measures of inflation have drifted below the 2% level quite consistently，despite the Fed's insistence that the target is "symmetric" and not a ceiling，raising questions over whether the Fed can credibly get inflation to its target.

At issue is the framework known as the Phillips curve，which is very flat and makes the relationship between employment and inflation more tenuous. It has been commonly agreed that the existing monetary policy tools – both conventional and unconventional – may prove to be insufficient to address current problems and offset a potential severe economic slowdown.

Alternative approaches have to be found. One solution is to generate expectations of higher inflation when they are needed，namely，when policy rates have hit the zero lower bound.

"Price-level targeting，if credible，does achieve that," Blanchard and Summers (2019) wrote. "If the economy is in a recession and inflation is low，the commitment to return to the price-level path implies a commitment by the central bank to generate inflation later."

New York Fed President John Williams calls for reassessing the inflation-targeting framework，and has advocated for price-level targeting，in which the Fed allows inflation to fluctuate as long as price levels themselves remain stable. "The persistent undershoot of the Fed's target risks undermining the 2% inflation anchor," Williams said. He argues that in low-interest rate environments，the 2% stated target has a "downward bias" to stay below the 2% level，and suggests that price-level targeting would allow for more dynamism in inflation expectations. He writes that price-level targeting would be able to compensate for a negative supply shock by allowing an overshoot of inflation，whereas average-inflation targeting would not.

However，price-level targeting has a major shortcoming. "It is symmetric: if the economy is operating at potential but inflation has been too high in the recent past，the central bank must be willing to return to the price-level path，and thus must be willing to tighten monetary policy and risk a recession for no good reason beyond the previous commitment." Blanchard and Summers wrote，"This may be politically difficult and，by implication，not fully credible."

Bernanke expanded on this theory，proposing a "temporary price-level targeting" approach that would apply price levels only at times when policy rates are at or very close to zero. At other times，the Fed would rely on its existing standard inflation targeting. "This approach involves only modest changes to the current framework，and it is consistent with the Fed's current mandates for maximum employment and price stability," Bernanke argued.

Another alternative，suggested by Lawrence Summers in 2018，is a shift to nominal GDP targeting calibrated to ensure nominal interest rates in normal times in the 4% range.

Eric Rosengren，the president and CEO of the Federal Reserve Bank of Boston，has long been an advocate of reviewing monetary policy frameworks，suggesting that any reassessment might include discussion of whether an inflation range (setting an inflation range with an adjustable inflation target，such as a 1.5% to 3% range，for example)，nominal GDP targeting，or price-level targeting would help the Fed better achieve its Congressionally-mandated goals. In October 2018，he delivered a speech at Harvard University on broad policy issues，proposing that the assessment "should be focused on an inflation range，with the potential to move within the range as the optimal inflation rate changes." (Also see Eric S. Rosengren (January 8，2018. "Reviewing Monetary Policy Frameworks"，remarks at a forum on the Fed's inflation target – the Hutchins Center on Fiscal and Monetary Policy，the Brookings Institution. Https://www.bostonfed.org/site-search.aspx?q=review%20monetary%20framework)

Keeping the current inflation-targeting regime but raising the inflation target to more than the 2% level，say 3 or 4%，is another option. Adam Posen argues that central banks should not be reluctant to change the inflation targets，and suggests that central banks raise inflation targets simultaneously in a coordinated fashion. "That would be the most credible，simple，lasting，and obvious way to make up for the price-level undershoots of the past," he wrote in a book chapter of Evolution or Revolution? Bernanke disagrees，saying it raises a number of concerns，including costs，uncertainties，delays，and so on.

Last November，the Fed announced a plan to conduct a comprehensive review of the strategies，tools，and communications practices used to pursue monetary policy goals. This review will include outreach to a broad range of stakeholders across the country.  The Fed has reiterated that the inflation target of 2% will not change，though some argue that it should convey the message that it wants to modestly overshoot its target. But the Fed said it is open to tweaks in the way it gets to the 2% target.

"[Economists have been] thinking of ways to make that inflation 2% target credible — highly credible — so that inflation kind of averages around 2% rather than only averaging 2% in good times and then averaging way less than that in bad times，which would drag expectations down," Powell told the Congress in February while testifying before the senate. He also admitted this will be a huge effort.

As the Fed maintains its rate hike pause，the central bank's approach to inflation could prove more important than the inflation target itself.

MMT Debate: Sense or Nonsense?

Surprisingly，just in the past several months，a very controversial heterodox monetary theory，dubbed the Modern Monetary Theory (MMT)，has become a focus of economic and policy debates.

MMT is a school of thought that's been gaining ground. Its advocates argue that budget deficits don't matter much so long as states can print money，because if a country，such as the US，UK，or Japan，borrows in its own currency，it can always print more money to pay its bills and cover its obligations. So，there is no great worry about accumulating too much debt. As a result，the thinking goes，the government should use fiscal policy to achieve full employment. The only constraint on spending is inflation once the economy reaches full employment or the public and private sectors spend too much. Inflation risk is often grossly exaggerated. When inflation is low，as it is now，there's space for larger government deficits. Without igniting inflation，the government can spend whatever it wants to maintain employment and achieve other goals. Inflation can be addressed by raising taxes and issuing bonds and removing excess money from the system. These tenets challenge the mainstream economics view that government spending should be funded only by taxes and debt issuance，not by printing money. As L. Randall Wray，a long-time MMT advocate says that MMT reaches conclusions that are shocking many conventional economists，challenges the orthodox views about government finance，monetary policy，the so-called Phillips curve trade-off，exchange rates，and current accounts. (See Chart 4)

MMT is quite controversial，with active debate about its policy effectiveness and risks.

MMT proponents expect central banks to keep interest rates low to contain the cost of paying bills，and create new money to fund government purchases. But Fed Chair Powell says the concept of MMT is "just wrong," and government deficits matter. Powell was asked about MMT in February when he testified before the Senate，and he answered: "The idea that deficits don't matter for countries that can borrow in their own currency I think is just wrong." He went on to say: "US debt is fairly high to the level of GDP – and much more importantly – it's growing faster than GDP，really significantly faster. We are going to have to spend less or raise more revenue.  And to the extent that people are talking about using the Fed – our role is not to provide support for particular policies," Powell said. "Decisions about spending and controlling spending and paying for it，are really for you."

Lawrence Summers，the former Treasury secretary under Bill Clinton and director of the National Economic Council for Barack Obama，also turned his fire on MMT，calling the doctrine “fallacious at multiple levels” and accusing its supporters of holding out the promise of a “free lunch.” “I don’t know what so-called ‘Modern Monetary Theorists’ have in mind，but the idea that we should abandon the tool of monetary policy and somehow make the central bank subservient，seems like we’ve seen that movie before,” Summers said in an interview with Yahoo Finance. But he and Olivier Blanchard all agree that when investment return or GDP growth is higher than interest costs，fiscal stimulus could be effective.

Generally speaking，critics say MMT is a recipe for reckless spending and hyperinflation. BlackRock CEO Larry Fink called MMT "garbage," while former New York Fed president Bill Dudley agreed with Summers and said MMT supporters want a "free lunch."

Japan's case has often been raised as an example by MMT's proponents. Japan has run up a public debt over the past years，mostly financed with the help of its central bank，yet it has very low inflation and can borrow money now virtually for free，and sometimes even paying negative interests. It's a puzzle which has long confused economists and policy makers. But there are also other cases such as some European and Latin American countries which experienced hyperinflation in the past.

The author would argue that though appropriately managed fiscal stimulus could be effective，as Lawrence Summers once said，the manifesto for deficit-friendly and free-spending governments is a terrible idea -– and if it's ever implemented，it would only end in tears one day. The theory which holds the point that governments can always run sustained budget deficits and rack up an ever-increasing debt burden is only "voodoo economics," a term used famously by George H.W. Bush to describe Ronald Reagan's policies when the two were rivals for the Republican presidential nomination. The broad trend since the 2008 crisis is toward coordination of monetary and fiscal policy，but budget deficits should not be simply financed by printing money.

The US budget deficit reached $779 billion in Trump's first full fiscal year in office. It's now on track to exceed $1 trillion in 2022，according to an estimate by the US Congressional Budget Office. Total public debt in the US had climbed to more than $22 trillion as of Feb. 11，2019，according to the US Treasury Department. (See Chart 5) Outlays are rising，while Trump's tax cuts have subtracted revenue. Some worry that the rising budget deficits could pose a potential threat to financial stability.

Should policymakers apply the MMT framework with more ambitious fiscal policies? It will remain a debate for a long time. The result of this debate is crucial not only for advanced economies such as the US，EU，and Japan，but for emerging countries like China as well.

The author is a visiting scholar at Harvard University and a former managing director at CITIC Securities