Economic News & Analysis—August 17, 2012
Why are rates so low—and can they stay low?By James Kochan, Chief Fixed-Income Strategist
Exceptionally low interest rates are a key feature of this economic cycle. Not only have most rates and bond yields reached the lowest levels in modern times, but they have stayed low and have done so for much longer than most observers would have expected in 2009, when the cyclical recovery began. By this time in previous cycles—the beginning of the fourth recovery year—interest rates were usually rising or about to start rising. What’s different now?
The economySevere recessions were typically followed by strong recoveries, until this cycle. For example, in the three years after the troughs of the steep recessions of 1974–1975 and 1981–1982, real GDP increased 14% and 18% (see Chart 1). By contrast, in the three years after the June 2009 trough of the current cycle, real GDP has increased only 7%—the poorest performance in the past 50 years.
Chart 1. In the three years after the June 2009 trough of the current cycle, real GDP has increased only 7%—the poorest performance for the past six prior cycles.
Strong economic growth has usually been associated with rising interest rates and market yields. That was true in the cycles of the 1960s and 1970s when, by this time in the expansion, interest rates and bond yields were above the cycle lows. This was not true in the 1982–1985 recovery because rates were declining from the record high levels reached in 1981. The recoveries of 1991–1994 and 2001–2004 were relatively slow, and market yields, on balance, declined over those periods. The fact that the current recovery is even slower is one reason why rates and market yields are lower now than in June 2009.
InflationInflation typically accelerates when growth is strong, and higher inflation typically begets higher interest rates. In the first three years of the strong recoveries of the 1960s, 1970s, and 1980s, the Consumer Price Index (CPI) rose an average of 13%. Today, the CPI is up only 6% from the cycle trough in 2009, and over the past 12 months it is up only 1.8%.
Two key contributors to inflation—labor costs and money supply growth—are much weaker in this cycle. Unit labor costs are only 1.1% higher today than in June 2009. That is the smallest three-year increase since the low-inflation decade of the 1960s. In the 1970s, unit labor costs were increasing at double-digit rates. Even in the weak recovery years of 1991–1994 and 2001–2004, unit labor costs increased an average of 3%. Intense international competition is forcing U.S. companies to keep production costs, including labor costs, under strict control.
Almost always, when Federal Reserve (Fed) policy is discussed in the media, the term “printing money” appears, and that leads to predictions of higher inflation. However, according to M2—a measure of the money supply that is used to forecast inflation—the money supply is now only 18% higher than it was when the recovery began in 2009. In the first three years of the six previous recovery cycles, the increase in M2 averaged 26%. In the two recoveries in the 1970s, it averaged 36% and 40%. Only the relatively anemic recovery years of 1991–1994 recorded slower M2 growth than the 6% average annual growth rate thus far in this recovery cycle. This indicates that monetary growth may not be strong enough to promote much future inflation.
Table 1. Previous recessions were typically followed by stronger recoveries than
the recent recession has experienced.
Credit demandsThe absence of a recovery in the housing sector has been a key reason for slower GDP growth and the persistence of low interest rates. The housing sector is a big user of borrowed money (mortgage loans), so healthy growth in housing would be expected eventually to cause market yields to rise. The collapse in mortgage lending might be the most unique feature of this cycle. During the first three years of the previous four recovery cycles, mortgage debt outstanding increased an average of 35%. Thus far in the current recovery, mortgage debt outstanding has declined 5%. A drop of that magnitude in private-sector borrowing more than offset the big increase in Treasury borrowing during this cycle. Even with the huge Treasury deficits, the total volume of borrowing in the U.S. capital markets is significantly less now than it was before the recession. If private-sector demand for credit were to remain weak, one important contributor to higher rates would remain missing.
Another offset to the heavy pace of Treasury borrowing has been very strong demand for Treasuries from overseas investors. Until the mid-1990s, foreign purchases of U.S. securities were minimal. Since then, they have been increasing steadily. By 2005, those purchases were averaging $70 billion per month. In 2010, foreign purchases averaged $65 billion per month, and that has been approximately the monthly average in 2012. Because most of those purchases are Treasury notes and bonds, foreign demand, sparked by turmoil in the European markets, has been a major factor keeping Treasury yields low. As long as the global markets remain unsettled, the demand for “safe” assets, such as Treasuries, would be expected to remain strong.
ConclusionAnyone who has experienced or studied the interest rate cycles of the postwar period would find it hard to believe that rates could stay near current levels for another year or more. The suggestion that rates could stay so low is completely at odds with past postwar cycles. Yet the differences between this and earlier recoveries are sufficiently great that past cyclical experience has not been—and probably is not now—a reliable guide to future events. GDP growth is half the average of previous recoveries, demand for mortgage loans is still moribund, inflation is not accelerating, and the world markets are so unsettled that the safety bid for Treasuries remains substantial. While these fundamentals might not be enough to keep all market yields at near-record lows, they are probably sufficient to delay the eventual cyclical rise in rates and yields beyond the foreseeable future.
Might this relatively sanguine rate outlook change? It depends on housing data, inflation figures, and the value of the dollar. A recovery in housing starts to above the 1-million-unit annual rate would suggest that private-sector credit demands had strengthened. A rise in core inflation to above 2.5% might prompt the Fed to start raising rates, even if the unemployment rate were still uncomfortably high. A prolonged downtrend in the U.S. dollar could reduce foreign demand for Treasuries, thereby weakening one of the major forces keeping Treasury yields exceptionally low. If this were a typical cycle, some of these developments would already be evident. But in this cycle, they remain below the horizon.