Over the past couple of years, there has been a decent amount of chatter about the Fed’s not being able to ‘reload’ in order to have enough tools to deal effectively with the next cyclical trough. And while the fear of running out of runway during the next downturn should not effect present policy, it likely is as it is a background consideration for the Fed as it thinks about liftoff. Below, I want to talk about this reload process. But I also want to move forward to the end of the cycle and give some thought to what happens to monetary policy when the cycle turns down with rates pinned near zero percent. I believe unconventional policy actions like quantitative easing – potentially including short-term municipal debt – will be used.
A friend of mine who runs a small hedge fund started a group thread via email on the Fed’s present interest rate conundrum. Several of us on the thread had previously said in various ways that the Fed wanted to raise interest rates for a multiplicity of reasons but that the data were so weak that it was not able to. Some Fed officials are concerned about financial instability, some about wage growth and inflation. Yet others still are concerned about ‘reloading’ for the next cyclical downturn. And some Fed officials are concerned that rates at zero for so long leaves the Fed open to criticism it has distorted the economy. There are different reasons for each member to want to raise rates. At the same time, Fed members are in dispute about how close we are to the 2% inflation target and how much slack there is in the labor market at present. And so the Federal Reserve has coalesced around the 5.0 to 5.2% long-term range of unemployment as a sort of happy medium trigger/threshold.
I have likened the target to the Wolfowitz rationale for the search for weapons of mass destruction in Iraq. Wolfowitz famously said about WMDs that for bureaucratic reasons “we settled on the one issue that everyone could agree on which was weapons of mass destruction”. And so that target took on more meaning than it should. The same kind of logic is operative with unemployment as a trigger for interest rate hikes: there are multiple reasons for wanting to hike, but the U-3 unemployment level represents a good identifiable trigger around which everyone can coalesce.
The Dallas Fed has a chart of Fed Funds targets from 1971 that give us a good sense of how much the Fed raised rates during each cyclical upturn from then onward. The first campaign of hikes began in November 1971 at 5.125%. And the Fed continued a gradual increase to late August 1973 at 11%. The Fed came down from there to 5.25% in 1975 before it raised rates and then quickly reversed back down to 4.75% in Jan 1976. Rates rose from there to 15.5% in Oct 1979, before being cut and raised again to 20% in March 1980. And on it went for several cycles. Each time that we hit a cyclical trough, rates took a nosedive and we are talking 500 basis points and more every time.
Here’s the question though: How does monetary policy actually work?
All evidence is that corporations do not take on capital expenditures in greater measure because of the interest rate cycle. The FT’s Matthew Klein writes for example:
Corporate executives still require new investments to generate the same double-digit returns as they did in the mid-1980s even though interest rates have dropped by about 10 percentage points. [James] Montier’s use of survey data fits with Warren Buffett’s old argument that nominal returns on corporate investment don’t tend to change much despite massive variations in the rate of inflation. Montier’s chart also fits with an important empirical study showing that capital expenditure “responds strongly to prior profits and stock returns but, contrary to standard predictions, is largely unrelated to changes in interest rates, market volatility, or the default spread on corporate bonds.”
In other words, if there is any linkage at all between real interest rates and real business investment, it is much more subtle than the standard forecasting economic models suggest.
I don’t have the data on this unfortunately. But, given the study, one could make the case that actually capital expenditure responds in the opposite direction of rates i.e. that rates go down when the markets go down and so, capex also goes down. Capital expenditure is correlated to markets and not to the federal funds rate or corporate risk spreads. What about households and interest rates then? Here’s Klein:
Professor [Paul] Krugman responded by noting that the relationship between interest rates and housing — both prices and construction activity — has been more robust, because houses last longer than typical business investments. That isn’t unreasonable, and there is some evidence that housing drives the business cycle. Also, recall the Bernanke and Gertler claim from 1995 that monetary policy affects the economy, first and foremost, because of its impact on residential construction. It’s worth noting that, if these arguments are right, you have to assign monetary stimulus a decent chunk of the blame for the excesses of the 2000s.
So household investment activity may be correlated to interest rate cycles, whereas non-residential capital expenditure is not correlated.
I mentioned a bunch of this on Twitter already, so let me repeat it here. On the consumer side of things, households don’t necessarily save or consume more based on interest rates either.
Montier: There is no identifiable correlation between real interest rates and household savings rates pic.twitter.com/fLxjsiVXlw
— Edward Harrison (@edwardnh)
Let’s remember from a savings and consumption perspective we also have the lost interest income of some to measure against less interest debt burden for others. Let’s remember that when mortgage rates declined during this up cycle, mortgage borrowing for those with high credit scores also declined. The NY Fed’s research on mortgage origination suggests lower rates have had the (perverse) effect of accelerating pay downs among creditworthy mortgage borrowers.Now, if high credit score mortgage borrowers have low marginal propensity to consume, lower rates could actually lower consumption. You’re getting a net shift to borrowers with low marginal consumption propensity away from others with higher consumption pattern. It is not clear to me that lower rates have the longer-term impact intended by policy because they have distributional effects.
The bottom line here is that zero rates could just as easily be deflationary over the long run by sucking interest income out of the private sector as they could be inflationary in supporting consumption. To me, it is wholly conceivable that low rates will not have the intended consequence of creating more consumption and higher GDP. Japan is the model here. Nothing I am seeing says that deflationary pressures are relenting in the US. And we should remember that quantitative easing replaces interest bearing assets in the private sector with non-interest bearing reserves, reducing interest income in the private sector outright. It can only be stimulative if it has portfolio balance effects or if it has credit-stimulating properties – which we already know from the research cited above, it may not have.
This is the context you have to see the interest rate decision in.
Yesterday, I was talking to The Wall Street Journal’s Jon Hilsenrath about the Fed’s conundrum in dealing with the aftermath of this rate cycle given the extensive use of unconventional policy when the Fed hit the lower bound. Jon said this rate cycle is unique among post-War cycle’s, making comparisons to the past difficult. But he also suggested that the first post-World War 2 cycle, when the Fed was cautious in removing accommodation, was perhaps the best guideline for thinking about how the Fed conducts policy as it seeks monetary policy normalization. He suggested this benchmark showed the Fed moving slowly over a period of decades rather than years.
When I met Jon, I had just finished this email thread about the Fed’s desire to ‘reload’ and was reading the FT’s Robin Wigglesworth’s piece on the taming of the Fed’s balance sheet and was starting to look at the Fed’s experience in 1937 with excess reserves. So my head was in a different place. I was thinking about the removal of excess reserves that coincided with gold sterilization and tightened fiscal policy in 1937. This combination of policy moves ended up causing industrial production to fall some 30%. And so the account gives me great pause about the ability of policy makers to navigate post-debt deflationary crisis choppy waters at the zero lower bound.
I believe what is likely to happen is that the Fed will indeed take its time in removing accommodation. It will raise rates first and deal with excess reserves later. So, 1937 – with its simultaneous doubling of the reserve requirement, gold sterilization and fiscal tightening – is not a baseline scenario. It is an outlier. At the same time, the Fed is working under the assumption that low rates are stimulative for the economy over the long haul despite the fact that empirical evidence suggests low rates have portfolio rebalance effects and that’s it. Once the portfolio rebalance effects fade, the stimulus will be non-existent. And the economy will be on its own.
In thinking about this cyclical upturn, then, we should be thinking about the ability of consumer demand to withstand shocks from exports, currency, inflation, capital expenditure or inventory purges. Right now, we are at a level of wage and income growth that I believe supports a baseline of 2 to 3% real GDP growth. And so we would need to see all of these factors come to bear negatively on the US economy for it to tip into recession. In Q1 we saw these come together to produce what is likely to be a drop in GDP quarter-on-quarter. And so, to the degree these factors continue to impinge on the US economy, we should expect weakness – a weakness that delays rate hikes. I don’t think we are there yet. I believe some of these factors will wane, and that GDP growth in Q2 and Q3 will improve enough that the Fed can consider raising interest rates.
I do believe we are closer to the end of the business cycle than the beginning, however. Think of this year as analogous to 2005 or 2006. Interest rate impacts on household savings and investment may be unclear but the effects on the balance sheets of debtors is more clear. Lower rates are good whereas higher rates are bad. Thus, in thinking about the next cyclical downturn, we should think about the impact on the banking sector and on deleveraging as we did last cycle.
Because of the banking sector’s leverage, it is a debtor. And so interest rates have a big impact on financial institutions.The banking sector always benefits from a steep yield curve and increased net interest margin due to rate cuts. And this will be notably absent this go around. I believe the yield curve will flatten and remain flat, making it difficult for financial institutions to recoup losses through high net interest margins. The lower net interest margins will limit credit, delaying any subsequent up cycle and introducing greater susceptibility to financial distress. At the same time, we should also be cognizant of other debtors, who will find the full weight of interest rates come down to bear on them without any relief from cuts. Last rate cycle, the relief from cuts in terms of reduced mortgage interest costs was significant. So again, the lack of interest rate relief will mean susceptibility to increased financial distress.
The Fed will want to counteract this in the only way it knows possible – and that is by moving to unconventional policy measures quickly. And because quantitative easing will not be effective in this environment except to the degree it is done as a vehicle for credit easing, I believe we will see the Fed turn to credit easing. Mortgage bonds will not be the vehicle of choice since that is not where the distress will be. More likely are government-backed student loans or short-dated municipal bonds, both of which the Fed is authorized to buy.
We are still a ways off from this outcome but I hope this post will be food for thought on how monetary policy is likely to develop as we head there.