While some have talked of the 'credit-easing' possibility a la Bank of England (which Goldman notes is unlikely due to low costs of funding for banks already, significant current backing for mortgage lending, and bank aversion to holding hands with the government again), there remains a plethora of options available for the Fed. From ZIRP extensions, lower IOER, direct monetization of fiscal policy needs, all the way to explicit USD devaluation (relative to Gold); BofAML lays out the choices, impacts, and probabilities in this handy pocket-size cheat-sheet that every FOMC member will be carrying with them next week.
The Fed's Arsenal
BofAML: Stage 3: Nuclear options
Nuclear options, by definition, are large scale approaches that are employed to lift “animal spirits", confidence or inflation expectations by credible means. As we noted before, they are unlikely to be adopted unless the economic outlook deteriorates dramatically. We would expect all of these measures to lead to a rise in TIPS breakevens and inflation risk premiums, resulting in a steepening of the long end of the Treasury curve. We discuss market implications of each of these options in detail below.
1. Imposing a ceiling on yields
The Fed could resort to targeting Treasury or MBS yields by committing to maintain long term yields at or below a certain level for a specified period of time. For example, the Fed could announce that it intends to buy Treasury securities in order to maintain 10y Treasury yields below 1.25% over the next two years. Such a move, which would be an unrestricted commitment to expand the balance sheet, has been suggested by Bernanke as an antidote to deflation in the past2. We believe that the Fed would choose a below-market rate since the aim would be to provide net new stimulus.
In our view, the Swiss National Bank (SNB) decision to announce a floor on EUR/CHF at 1.20 last August – shortly after it had traded as low as 1.03 – serves as an interesting case study. The spot exchange rate has traded consistently above the 1.20 floor and 1m implied vol on the EUR/CHF currency pair has fallen from a peak of more than 27 last August to just 2.4 currently, suggesting the floor is seen as credible, at least in the short term (Chart 16). However, this has come at great cost: since July 2011 the SNB’s FX reserves have increasd by 190bn francs (or 82%) to approximately 240% of GDP (Chart 17).
- We would expect 10y yields to rally slightly below the ceiling set by the Fed. However, we are more skeptical that yields would rally to levels significantly lower than the ceiling, as higher inflation risk premiums and a positive stock market reaction should be limiting factors.
- The curve reaction would likely depend largely on which rate(s) the Fed targets. We would expect the curve to flatten until the point of the target (say the 10y) and then for the 10-30y sector to steepen considerably as inflation expectations should rise. This could prompt the Fed to cap rates at multiple points on the curve.
- The commitment to expand the balance sheet without a limit would likely increase inflation risk premiums sharply across the curve, though presumably this would be a desired effect. Assuming that yield target is around the 10y, we would expect 10y TIPS breakevens to rise close to 2.5% real rates to decline so as to bring nominal rates below the yield ceiling.
- Volatility should decline significantly as the Fed cuts off the higher rate distribution, at least temporarily. We would expect short term payer skew to cheapen while longer term payer skew should rise given higher inflation risk premiums.
2. Mandate targeting
Chicago Fed President Charles Evans has proposed that the Fed keep policy accommodative until certain conditions or “triggers” for a change in the policy stance have been met. In other words, make policy “state dependent” (i.e., a function of economic conditions and/or forecasts) rather than “time dependent” (expiring at a given point in time). Both the forward guidance and the balance sheet expansion could be made state dependent.
Evans has proposed a pair of triggers, based on the dual mandate. The first is for the unemployment rate to fall below 7%, which appears to be well above the Committee’s estimate of the NAIRU inflation trigger. The second is if the inflation rate rose above 3%. The rationale for this number is symmetry: Evans noted that the Committee was willing to let the inflation rate drop to 1% before aggressively responding in late 2012; thus, they should be willing to let inflation rise to 3% before strongly countering it.
Importantly, these triggers are designed to be simple but temporary means of communicating the likely stance of Fed policy when it is significantly missing its dual mandate objectives. In more “normal” times, these triggers would not be used. In particular, the 3% trigger is supposed to be consistent with a longer-term 2% inflation target. That is, should the unemployment rate be sufficiently high, the Fed might tolerate inflation up to 3% for a period of time, but would still aim to keep inflation around 2% in the long run. The hope is that with clear communication, the Fed’s anti-inflation credibility would not be damaged.
- Committing to an unemployment trigger would likely to lead a large decline in rates, in our view. Given our current forecasts, we do not project unemployment levels to reach even 7% before 2015 or 2016. This would likely mean continued purchases of Treasuries, which could likely take Fed ownership of the 6y-30y sector above 50%. 10y rates could decline close to 1% in this scenario, in our view. Further, the first Fed hike would be extended out much more, thus helping lower 5y rates.
- TIPS breakevens are likely to increase as the indefinite expansion of the balance sheet is likely to raise inflation risk premiums. However, the effect on the nominal curve would likely depend largely on the sector purchased by the Fed in order to meet its mandate. Any distribution similar to Operation twist would likely to lead to a large flattening of the curve.
- The move lower in rates should lower volatility in the rates market. However, volatility could significantly increase around key data releases such as payrolls.
3. Raising the inflation target
A more extreme policy than temporarily establishing an above-inflation target trigger for policy is to just raise the central bank’s long-run inflation target itself.
Recently, this approach has been advocated by Princeton’s Paul Krugman and the IMF’s Olivier Blanchard — not to mention a certain Ben Bernanke some years ago when advising the Bank of Japan. Such a policy is most appropriate when a central bank has been unsuccessful in preventing a deflationary psychology from setting in. In this case, the central bank commits itself to being “irresponsible,” in the words of its proponents, in order to raise inflation expectations above a deflationary level. This should stimulate the economy by reducing real interest rates, thereby encouraging borrowing and investment and jump-starting the economy from a liquidity trap.
However, many central bankers see this option as playing with fire — the current chairman of the Federal Reserve included. In his recent Humphrey-Hawkins testimony, Bernanke argued that raising the inflation target was unlikely to pass a cost-benefit analysis: “I am very skeptical that it would increase confidence among businesses and households and increase economic activity. I think it would create a lot of problems in financial markets as well,” he stated. Bernanke also suggested that it might be hard for a central bank like the Fed to easily unhinge currently well-anchored inflation expectations — and then, once doing so, it would also be hard to re-anchor them at, say, 4% without potentially drifting higher (Chart 19). Once a central bank decides it won’t stick to one previously announced inflation target, why should the public believe it will stick to a different one?
- We would expect long end rates to rise considerably, led by inflation expectations. However, if the Fed combines its inflation target along with a QE program, purchases would likely offset the rise in inflation expectations by lowering real rates. The net effect on nominal rates would likely depend on how aggressively the Fed set out to achieve a higher inflation target. If the Fed combined a higher inflation target with a ceiling on yields, it could drive real rates deeply negative.
- TIPS breakevens and breakeven volatility would likely shift much higher as the risk of the Fed losing control of inflation expectations increases. This would also likely cause a shift in base inflation expectations to 3% or above.
- Vega on 30y tails would likely increase significantly and payer skew should richen as inflation expectations head higher
Coordinated nuclear options
The Fed could also embark on a number of actions with coordinated support from the Treasury and/or Congress. These options are likely to have largely similar impacts on the rates market given how large they are likely to be. First, the market would presumably be at extremely stressed levels for the Fed to even consider any of these options. This would likely be manifested in much lower rates, flatter curve, wider credit spreads and a lower stock market. So we expect the nuclear policy response to come as a relief and should result in rising rates, higher inflation expectations, a steeper curve and higher vol. Note that the Fed could combine one of these policies with a yield target to limit the rise in rates.
Money financed fiscal expansion
This option is the realization of Milton Friedman’s rhetorical “helicopter drop” of money to stimulate spending. When the Fed engages in traditional expansion of the money supply, money slowly and rather indirectly works through the financial sector and into the hands of consumers and businesses though an increase in lending. However, a money financed fiscal expansion would directly stimulate aggregate demand by increasing government consumption expenditures and/or putting the money in the hands of those who would spend it.
Instead of financing expansionary spending or tax policies through debt issuance or tax increases, the government would rely on seignorage (i.e., printing money). For example, the government thus could finance tax cuts — an expansionary fiscal policy — via Fed purchases of Treasuries funded by an increase in excess reserves. (Note that the Federal Reserve Act prohibits Fed purchases of government debt in the primary market, so the Fed would still have to purchase Treasuries in the secondary market as in the case of QE). Alternatively, the government could increase transfers or rebates to households or businesses paid for through seignorage. Presumably, there is a psychological impact of a temporary income windfall that results in a higher than usual propensity to spend and thus a larger stimulative impact upon the economy. Moreover, such a policy would likely weaken the dollar, potentially providing a boost to net exports (Chart 20).
There are several challenges to this approach. First, it runs the risk to generating higher inflation. Thus, it is most likely to be tried — and to be effective – during a recession. Second, it strongly blurs the line between monetary and fiscal policy, and could raise concerns among investors that the government might try to coerce the central bank to repeat such a policy when it is not appropriate, such as ahead of an election. In that case, the independence of the central bank would be questioned, and investors would price in a greater probability of higher inflation over time. That said, it seems extremely unlikely that Congress would agree to such a policy given the current political environment in Washington DC. Also, it is likely to be extremely difficult for the Fed and Congress to agree upon how much money financed fiscal expansion would be appropriate.
FX market intervention to weaken the dollar
One of the channels through which monetary policy affects aggregate demand is through the exchange rate. Expansionary monetary policy typically is associated with a weaker currency, which should bolster exports and discourage imports, all else equal. However, the Fed’s approach toward the dollar is best described as one of “benign neglect”: the Fed does not typically try to weaken the dollar as an explicit goal of policy, but it recognizes that dollar depreciation is a normal consequence of expansionary monetary policy and does not try to counteract it either. (Indeed, Fed officials rarely discuss the dollar, as by convention foreign exchange policy is the authority of the Treasury Department.) In this plan, the Fed would instead directly (perhaps in conjunction with the domestic fiscal authorities or with other central banks) sell the US dollar and buy currencies of other regions in order to weaken the international value of the dollar.
This approach could be inflationary if imports become more expensive as a result.
Given calls to avoid “dollar debasement” in the Congress currently, this policy would also contain significant political risks. Although highly unlikely to be implemented today, it is worth noting that Bernanke’s academic work on the Great Depression has discussed the reflationary effects of the 40% dollar devaluation against gold that was engineered by President Roosevelt in 1933. This coincided with a sharp rebound in US industrial production, which had fallen by more than 50% over the prior four years.
The question of what assets the Fed would purchase with its intervention proceeds is also relevant. Section 14 of the Federal Reserve Act authorizes the Fed to purchase "obligations of, or fully guaranteed as to principal and interest by, a foreign government or agency thereof." In other words, the Fed has the statutory authority to purchase foreign currency-denominated government debt in unlimited quantities should it choose to do so.