The Fed is not in the driving seat any more

The Fed is not in the driving seat anymore

September 05 2019

Executive Summary

The recent gathering of central bankers at the Jackson Hole symposium of August 2019 was the occasion for prominent members of the Fed to confess a sense of powerlessness when confronted with the consequences of the White House’s disruptive economic policy. Several stylized facts suggest that the Fed is not in the driving seat of U.S. monetary policy any more.

  • The Fed’s independence is at risk as President Trump’s tweets and recent institutional reforms making it more accountable to public bodies and civil society, as well as higher pressure from public opinion, have significantly increased the number of constraints weighing on its decisions. A text-mining approach, based on an analysis of sentiments on social media, confirms that 2017-2018 was a turning point with regard to the perception of its policy.
  • The Fed is now too “market-dependent”. A policy reaction function with time-varying parameters shows that the Fed has increasingly attached a larger weight to the stabilization of market volatility, to the detriment of its targets on growth and inflation.
  • The Fed is less and less credible in fulfilling its dual mandate. Our proprietary indicators mirror a higher risk of persistently undershooting its 2% inflation target over the medium-term, an altered capacity to maintain a situation of full employment - should the economic cycle deteriorate again - and higher risks of policy mistakes.
  • The Fed has become a poor guide for the market. We show that cross asset correlations have reached a pretty high level in 2019. This is equivalent to a loss of directionality for the market, i.e. it suggests that the Fed has become a poor guide for investors but also shows that there won’t be any place to hide in case of a severe episode of stress.
  • The Fed is about to lose its grip on currency policy. Despite an easing of U.S. monetary policy, the USD Trade Weighted index reached a record high during the summer of 2019. This primarily reflects a loss of influence on currency issues because of the actions of the U.S. Treasury and the People's Bank of China (PBOC).

History tells us that weaker central banks are associated with a higher risk of recession. Policy mistakes range from prematurely tightening monetary policy, nurturing bubbles, or lacking the authority in circumstances requiring rapid and bold moves of stabilization. Separately, the lack of direction perceived by the market is propitious to the existence of multiple equilibria and self-fulfilling prophecies. This combination of factors is prone to increase the probability of recession, as observed today.


Today, U.S. unemployment is below 4% (maximum employment objective fulfilled) and inflation is broadly under control, close to 2%. Despite being close to its objectives, the Fed seems to be eager to ease its monetary policy in a context of high political and trade uncertainty in order to preemptively smooth an upcoming deceleration of growth. One can reasonably ask the question of a possible over-reaction of the Fed, which is possibly being influenced by the market or the U.S. government. In a context of a late phase in the cycle and increasing probability of a recession, there is a risk that a weaker Fed favors self-fulfilling phenomena and contributes to aggravate the impact of a potential correction.

The Fed’s independence is at risk due to higher pressure from external actors

Four former heads of the U.S. Federal Reserve (Ben Bernanke, Alan Greenspan, Janet Yellen and Paul Volker) recently published an article in the Wall Street Journal (“America Needs an Independent Fed”, 09 August 2019) reiterating the need for Federal Open Market Committee (FOMC) members to remain independent from any political pressure. Today there is a high probability that the Fed’s independence is at risk due to the following four factors:

  • Direct pressure from President Trump on the Fed. A succession of tweets by President Trump has clearly indicated his strong dissatisfaction with Jerome Powell’s policy. Because of his moderation in cutting the official rate, Powell was described as an enemy of the U.S. in August 2019. This strong pressure coming from the U.S. President himself (the U.S. President nominates the Fed Chairman for a mandate of four years) undeniably endangers the independence of the Fed.
  • A reinforcement of the Fed’s accountability to the public. In the years following the global financial crisis (GFC), the Fed’s accountability to the public has significantly increased. This followed criticisms that it abused its mandate by intervening in a discretionary manner to save failing banking institutions during the peak of volatility in 2009 and by injecting huge amounts of liquidity into the U.S. economic system via security purchases. Since the GFC, the tone of communication has significantly harshened during testimonies of the Chairman in front of the Congress. Moreover, the Fed has recently opted for more frequent interactions with the press regarding the orientation of monetary policy: almost on a monthly basis compared with a quarterly frequency before. The accountability of the Fed is therefore much stronger now.
  • ·A re-composition of the FOMC and regulators, which is likely to reflect a higher influence from the business sector. The new regulators nominated by the U.S. President to the top of the FOMC, the Consumer Financial Protection Bureau (CFPB) and the Federal Deposit Insurance Corporation (FDIC) have a more pro-business profile. This has definitely increased the potential of influence coming from the corporate sector.
  • A significant deterioration in the perception of the Fed’s action.  2017 and 2018 have represented a turning point in terms of the perception of the Fed’s action. Based on a text-mining approach[1], we build an indicator collecting words associated with negative sentiments about the Fed on social media. We can see in Figure 1 that we have recently experienced a change of regime compared with previous years in the sense that the deterioration of this sentiment is the largest and the longest-running since 2013. In a context where the Fed is being held much more accountable compared with pre-crisis times, this means that the level of pressure on FMOC members’ shoulders is very high today.
Figure 1: Sentiment on the Fed in social media (level)
Figure 1: Sentiment on the Fed in social media (level)
Sources: Euler Hermes, Allianz Research

The Fed is now too “market-dependent”

Over the last few months and days, the Fed has often given the impression of bowing to the market. It was particularly rapid in interrupting the normalization of its monetary policy and re-envisaging rate cuts when confronted with a sharp inversion of the U.S. yield curve (synonymous with an increasing probability of recession) or with a sharp correction of the equity market, as seen in the end of 2018 or during the summer of 2019.
Today, the Fed owns close to USD 2 trillion of U.S. government bonds, representing a non-negligible part of the market (at its peak it reached 19% of the publicly traded U.S. Treasury securities). The persistently high exposure of the Fed to Treasuries probably explains the lack of aggressiveness of the central bank when normalizing its monetary policy. A too rapid implementation of rate hikes or reduction of its balance sheets would have considerably disrupted the market of U.S. sovereign bonds and could have significantly impaired the value of its portfolio. The destinies of the Fed and the U.S. fixed income market are now deeply intertwined.

In order to test this over-sensitiveness of the Fed to the market’s variables, we build a Taylor Function augmented by market volatility, i.e. we study the reaction of the Fed, in terms of interest rate decisions, when inflation deviates too extensively from its 2% target (the inflation gap), when unemployment deviates too much from its equilibrium value (the so-called NAIRU gap) and when the market’s volatility deviates too extensively from its average value [2] (we use a z-score of the St Louis Fed’s volatility index). Estimates of coefficients are presented in Figure 2 below.

Figure 2: Time varying coefficients of the Fed’s Taylor function
Figure 2: Time varying coefficients of the Fed’s Taylor function
Sources: Euler Hermes, Allianz Research

The signs of the coefficients attached to inflation and unemployment stabilization are conventional as they are, respectively, positive for the inflation gap (a level of inflation above the target of 2% triggers a tightening of monetary policy) and negative for the NAIRU gap (a level of unemployment above its equilibrium value triggers an easing of monetary policy).  The sign of the coefficient attached to the stabilization of the market’s volatility has turned from being positive to negative, showing that the Fed now cuts the official rate when observing a surge of market volatility. In parallel, the coefficient attached to its inflation stability mandate has significantly declined in a particular context where inflation is definitely under control.

The Fed is less and less credible in fulfilling its dual mandate

In a context of persistently low inflation, a frequent undershooting of the Fed’s target of inflation at the rate of 2% (“as measured by the annual change in the price index for personal consumption expenditures, or PCE”), could damage the credibility of the U.S. central bank. Different filtering techniques based on Allais transformation (see our upcoming publication by Eric Barthalon, Allianz Research capital markets) suggest that long-term core inflation expectations (breakeven rates extracted from long-term inflation-linked bonds) are likely to stay within a 1.20%-2% range in the next three to four years, this band of fluctuation being characterized by a low elasticity. This means that the risk of a de-anchoring of inflation expectations (even if it is too early to conclude this) has increased, which would represent a failure of the Fed in its inflation stabilization mandate.

The other aspect of the Fed’s policy, maximum employment, is less frequently questioned, especially when the U.S. economy is experiencing a situation of full employment as it is today (the U.S. unemployment rate reached 3.7% in July 2017). However, it is more the capacity of the Fed to stabilize growth and employment, should the macroeconomic situation rapidly deteriorate, that we should interrogate instead of today’s situation in the job market. To this regard, it is interesting to compare the estimate of the output gap by the NY Fed and the Federal Reserve Bank of St Louis (Figure 3).

Figure 3: Estimated output gaps (deviation as % of potential GDP)
Figure 3: Estimated output gaps (deviation as % of potential GDP)
Sources: Euler Hermes, Allianz Research
They differ significantly, meaning that deciding whether to ease (or tighten) the monetary policy in such a context is particularly difficult. We build a Fed dissension index, a proxy of the difficulty surrounding the Fed’s ability to make the right decision, based on the absolute value of the 12-month moving average spread between the two estimates of the output gap (figure 4). It shows that the difficulty around making the right decision for growth stabilization has been close to a record high over the last few years. These considerations are consistent with numerous critics and divisions regarding the correct value of the natural rate of interest. There are therefore strong doubts over the Fed’s capacity to correctly analyze the current economic cycle. In the case where the monetary policy would have been over-expansionary, because of an output gap being much higher in reality than perceived by the Fed, it could nurture the risk of bubbles, as it is often mentioned regarding the U.S. credit market. In the case where the monetary policy would have been restrictive because the real output gap is much smaller than estimated by the FMOC, it could create too tight credit conditions and contribute to increase the risk of a recession. A large degree of dissension regarding the right position in the cycle, and  the estimate of potential GDP, therefore significantly increases the risk of policy mistakes.
Figure 4: Fed’s dissension index in interpreting the growth cycle (level)
Figure 4: Fed’s dissension index in interpreting the growth cycle (level)
Sources: Euler Hermes, Allianz Research

The Fed has become a poor guide for the market

The loss of directionality given by central banks to the market is a traditional characteristic of times preceding major disruptions in financial and economic variables. Bernanke has repeatedly put forward this weakening of the central bank’s role as being an important factor contributing to instability. This is visible today as traditionally unrelated segments of the market move in the same direction in a more frequent manner. For example, the number of days during which the 11 sectors of the S&P500 index have moved in the same direction in 2019 is at its highest point since 2015. This means that the risk-off / risk-on dynamic prevails. In the same vein, demand for diversification has significantly increased in 2019, as mirrored by the significant increase of multi-asset strategies’ performances. The correlation between the S&P 500 index and multi-asset index ETFs is currently at extremely high levels (Figure 5).

Figure 5: Correlation between S&P500 and ishare multi-asset income index ETF (level)
Figure 5: Correlation between S&P500 and ishare multi-asset income index ETF (level)
Sources: Euler Hermes, Allianz Research

The Fed is about to lose its currency policy tool

The Dollar Trade Weighted Index is today at a record high level (Figure 6). Post Tokyo’s G7 summit, Steven Mnuchin confirmed that the U.S. Treasury could envisage currency interventions with or without an agreement with the Fed. Under a proposed new rule that could come into effect in the coming months, the U.S. may also impose tariffs on countries perceived as manipulators of their currencies. In August, China was officially identified as being a currency manipulator. The U.S. Treasury is therefore about to use the currency tool without real consultation with the Fed. There was until today no real interference of the Treasury on currency initiatives, with the Fed implicitly having the initiative, via the orientation of its rate policy, to influence USD fluctuations. Today, the initiative on currencies is more on the White House’s side, with negative consequences as the current strength of the USD broadly neutralizes the effects of tariffs. The U.S. current account balance has indeed fluctuated between -2% and -2.5% of GDP since President Trump started his mandate. On the external side, the huge amount of reserves in the PBOC’s balance sheet allows the Chinese central bank to use its currency to absorb external shocks. This is today the case: The depreciation of the CNY is managed by the Chinese monetary authority and significantly impacts the value of the USD at a global level. Therefore, both external and internal forces have taken the tool of currency policy away from the Fed.

Figure 6: Fed’s Trade Weighted Dollar index (level)
Figure 6: Fed’s Trade Weighted Dollar index (level)
Sources: Euler Hermes, Allianz Research

Policy mistakes as a possible trigger or aggravator of recessions  

Ben Bernanke[3] has extensively studied the central role of the Federal Reserve during the Great Depression. One key argument deals with the loss of confidence in the monetary institution and the erosion of its authority, which led to policy mistakes. Today, these types of policy mistakes remain conceivable, i.e.

  • Prematurely tightening monetary policy following a wrong assessment of the real strength of the U.S. economy. (This is a criticism currently expressed regarding the Fed’s recent initiatives in terms of monetary policy normalization.)
  • Aggravating the impact of a crisis due to the lack of decisive action. At this level, the lack of authority could hamper the central bank’s capacity to take rapid and bold decisions to rescue potentially failing banking institutions. The major risk for U.S. companies in these are related to refinancing needs. We know that the degree of leverage of U.S. companies is probably under-estimated[4]. There is therefore a need for them to reduce their refinancing needs as any policy mistake of the Fed could be the trigger for a significant tightening of credit conditions.
  • Nurturing risk and bubbles. Being too obedient to the White House could lead to an over-expansionary stance of monetary policy. The credit bubble would probably be the main beneficiary of this. This again points toward the need for U.S. companies to reduce their leverage and build up liquidity buffers.

[1] Analysis of a large sample of tweets containing the words "@federalreserve","Federal Reserve","Janet Yellen","Jerome Powell","FOMC","FED" and then analysis of the sentiment based on Loughran-McDonald dictionary.

[2] We use a time-varying equation (state-space model using a kalman filter, following the methodology of Strohsal & all, 2016, Journal of Macroeconomics) to study the evolution of coefficients attached to inflation spread, NAIRU gap and St Louis financial stress indicator in the Fed’s reaction function. We also use a dummy variable corresponding to the implementation of QE starting in November 2008.

[3] Bernanke, Ben, "Non-Monetary Effects of the Financial Crisis in the Propagation of the Great Depression," American Economic Review, 1983.

[4] US corporate leverage is probably under-estimated, Euler Hermes, 2019.

Head of Economic Research