Tim Duy looks at rising long rates and the yield curve. Is a steeper yield curve good news or bad news?:
Morgan Stanley strategist Stephen Roach sees bad tidings in the steepening of the US yield curve in recent days. But I seem to recall considerable debate that the opposite, a sustained and increasing inversion of the curve, was a bad thing as well. Indeed, James Hamilton at Econbrowser has been relatively vocal with his concerns. Is this yet another case of “two economists, three opinions?” Should we be more worried that the curve is steepening, or that it is inverting?
This is a tough question.
Consider first the inversion debate. Many economists believe that an inverted yield curve is a good – or even excellent – predictor of recessions. This has been met with some skepticism on the part of Federal Reserve officials, who tend to view low long term rates as the result of structural shifts in the global economy. From Chairman Bernanke’s recent Congressional testimony:
The premiums that investors demand as compensation for the risk of unforeseen changes in real interest rates and inflation appear to have declined significantly over the past decade or so. Given the more stable macroeconomic climate in the United States and in the global economy since the mid-1980s, some decline in risk premiums is not surprising. In addition, though, investors seem to expect real interest rates to remain relatively low. Such a view is consistent with a hypothesis I offered last year--that, in recent years, an excess of desired global saving over the quantity of global investment opportunities that pay historically normal returns has forced down the real interest rate prevailing in global capital markets.
Your assumptions about the nature and timing of a structural break in the US economy will greatly influence your view on the usefulness of the yield curve in predicting future economic activity. For example, I ran a simple probit model of the NBER recession dates against a constant and the 10 year – 2 year yield spread lagged 12 months. I ran two models using start dates of 1978:1 and 1984:1, respectively. In both cases, the initial end date was 1994:1. After running the model, I forecast the probability of recession 12 months ahead. I then add a month to the end date, rerun the model, and then make a new 12 month ahead forecast. The final end date is 2006:2, with a final forecast of the recession probability for 2007:2. The results, graphically:
If you use the 1978 start date, the odds of a recession beginning in February 2007 are low, just 16.5%. Of course, you would have completely missed the recession that began in 2001. If instead you use the 1984 start date – a cheap and easy way of accommodating a structural break at that time – the probability of recession next February is 45% - quite high, but still below the 50% cutoff. The last time the model predicted a recession was for March 2001 (using March 2000 data), with a probability of 53%. The probability for April 2001 was 69%. Of course, you can argue the model was lucky. After all, only a single recession was in the data at that point.
For more detailed treatments of this topic and methodological approach, see Estrella and Mishkin (1998) or Chauvet and Potter (2002). I will leave the reader to decide which model is more accurate, only adding that from my perspective, I would feel better seeing rising long rates prior to additional Fed rate hikes since a rising long rate would indicate a strengthening economy. A more cautious approach, yes, but I tend to be somewhat risk averse.
Does a steepening yield curve imply that we shouldn’t be concerned with rising long rates? In Steve Roach’s view, low rates are a symptom of the excess liquidity that has kept the global economy afloat over these past years. In this view, significant dislocations will occur as a sudden renormalization of interest rates asserts itself. A pressing concern – see Brad Setser – is the possibility that higher rates disrupt the apparent soft landing of the housing market. Another concern is disruption to the idyllic conditions in the corporate bond market (WSJ subscription). The list goes on and on; you get the point.
So maybe a steepening in the yield curve isn’t so great after all. Can we untangle these two views? We can at least try…
Let’s assume, talk of dark matter aside, that the US current account deficit (CAD) is an artifact of an important global imbalance, a mismatch between saving and investment. Global central banks created a vast amount of liquidity to offset the negative impacts of this mismatch, and much of that liquidity found its way into the US, stimulating the housing market (and, arguably, the earlier tech boom). Foreigners were willing to accept pieces of paper in exchange for their goods, and in the process the American consumer helped support global economic activity. Another consequence of these events is that the US consumption propensities exceed domestic production capacities.
I have simplified the story, of course, and many versions have been told. For my purposes, the exact details are not that important. What is important is that no one really believed that this situation would continue indefinitely. Eventually, the mismatch between savings and investment would narrow, and the need for central bank liquidity would decline as well. Refer again to a recent speech by Richmond Fed President Jeffrey Lacker:
Policymakers will need to be alert for movements in economic fundamentals that shift the relative pressure on current versus future resources in ways that require changes in real interest rates, even if inflation pressures subside.
Arguably, that story is unfolding, as central banks in Europe and Japan gear up for additional tightening in light of changing economic fundamentals. Of course, you have to believe that economic fundamentals are indeed changing to buy into this story. Here I am sympathetic. Rising energy and other commodity prices and brisk global economic activity have the feel to me of investment opportunities on the rise. Think of China’s plans to build 30 nuclear reactors. Think of Jim Hamilton’s questions about the current oil price assumptions with regards to investment planning.
Presumably, as the savings/investment imbalances narrows, so too must the US current account deficit. This requires that the US reduces the gap between its production capacities and consumption propensities. Optimally, this can be achieved by an internal rebalancing of economic activity. Household consumption grows by a slower rate such that the investment share of GDP rises. Rising long term rates are consistent with the scenario – the market is signaling the need to shift resources from households to firms. On net, GDP grows at slightly below potential, the gap between consumption and production falls over time, and so does the CAD. This process may already be in play – I noted in my Fed Watch earlier this week that the housing market appears to be slowing while the need for new investment looks solid. So far, so good.
But when does it turn bad? As always, this is a speed of adjustment issue. The imbalance grew over a decade – it may take that long to unwind. If the unwinding happens too quickly, things get messy. If foreigners become unwilling to finance the US current account deficit, the internal imbalance of overconsumption relative to production would need to be forcibly resolved. Structural rigidities make it very unlikely that the resolution would result from increased production. No, I can only think of one way in which to resolve a 7% imbalance; interest rates (and taxes) would need to rise sharply to curtail domestic demand via a significant recession.
My null hypothesis is that the adjustment will occur gradually over time. Two factors appear relevant here. First, the global savings/investment mismatch will likely unwind over a process of years – investment plans today, like natural gas pipelines from Alaska, don’t require the resources until sometime in the future. Second, central banks, particularly in Asia, have an incentive to maintain stability, and I believe would be willing to gradually reduce the acquisition of dollar denominated assets. Of course, should falling bond values prove too much for the Bank of China’s balance sheets to handle….for that, I will refer you back to Brad Setser.
This sounds like – and is – a relatively soft landing story for the US economy. Note that I am masking some potentially painful dislocations under the surface. A slowdown in consumption spending might fall disproportionately on the auto industry, for example. There will be winners and losers in the adjustment process.
The Federal Reserve will be setting policy in somewhat dangerous waters. My suspicion is that they will behave exactly as Lacker suggests – largely following the long end of the yield curve to align short and long term borrowing. They will be hesitant to allow the yield curve to steepen dramatically as that may stimulate excessive lending activity given the resource constraints that rebalancing implies. They will be hesitant to invert the yield curve because they want to maintain sufficiently liquidity to support investment.
Moreover, they will (or should) recognize the need to maintain relatively high short term rates to maintain capital inflows. Indeed, this is where the real problems can emerge. If the dollar falls sharply, inflationary pressures could emerge quickly (especially if the era of low inflation has been the result of easy access to cheap foreign goods). This is where the Fed would have to bring out the big guns to both attract capital flows and resolve the consumption/production imbalance via a rapid unwinding. A sharp inversion of the yield curve would likely emerge – and I think one that would accurately signal a recession.
One final thought. Although this is an area I engage cautiously, I suspect that such an environment would be challenging for US equities, at least short term. True, the estimated PE on the S&P500 is somewhere around 15, continuing it steady downward trend. But this is still historically a bit high, while the dividend yield is just 1.8%. Moreover, the scenario I just sketched out above implies rising labor and capital expenses, suggesting falling profit growth. Compare this to the 4.5% rate currently offered on overnight money, likely rising to 5% over the next couple of months….something to chew on at least.
Whether or not the economy can safely traverse this knife edge set of risks remains to be seen. There are a lot of “ifs” and “maybes” in this story. And I would prefer to enter such a transition with a more balanced federal budget. Bad policy or negative shocks could initiate a more rapid unwinding of the imbalances than I anticipate. Or bond yields could retreat and this whole process could be delayed yet another year.
Still, I admit to being cautiously optimistic; I tend to see the US economy as inherently stable and resilient. But, then again, I am just another economist entitled to his 1.5 opinions.