The Federal Reserve Open Market Committee decided to keep the federal funds rate at 5-¼ % again because “the moderation in economic growth appears to be continuing” and “inflation pressures seem likely to moderate over time”. This sentiment flies in the face of history. Over the last twenty years the inflation rate has accelerated almost half of the time when the growth of industrial production slowed. Nonetheless, someone is buying the idea that core inflation rates will decline. Since the end of June, longer-dated bond yields have fallen (i.e. prices have risen) substantially on the belief that a significant slowdown is in the offing and that inflation will soon begin to decline. On the other hand, stock investors have rallied prices back up near their May highs because they see little or no slowdown occurring in 2007. In fact, 2007 earnings estimates have increased over the last three months. These opposing opinions might be summed up in the first lines of Buffalo Springfield's 1967 rock anthem “For What It's Worth”:

“There's something happening here
What it is ain't exactly clear”

Just what is going on here? How should one view the current economy and how do we know that inflation will go down? Although the Fed controls the level of short term interest rates, it is just one piece of the puzzle. It is the “yield curve” that gets the most press and it should. The simplest way to depict the yield curve is to compute the “yield spread” or the difference in yield between 3 month Treasury bills and 10 year Treasury notes. This metric has had a good track record of calling recessions. When short rates (yields) are equal to or higher than longer dated bond yields, the curve has become “inverted” and the “spread” become flat or negative. Such a spread has proceeded all recessions over the past 40 years. Nonetheless, not all inverted yield spreads have created a recession!

Is the current inverted yield curve predicting another recession? Many economists and investment strategists believe so. Even some of the top fixed-income investment managers are calling for an extremely weak economy next year and are buying long term bonds because they believe that yields will go down and prices will go up. However the inverted yield curve signal is not invariably correct. For example, the yield curve flattened to narrow spreads in late 1995 and 1998 with no recession occurring; and the associated slowdowns were rather mild. According to Princeton economics professor Burton Malkiel, the inverted yield curve signal “has predicted eight of the past six recessions”. How can we confirm that the index is operating correctly this time?

The best leading indicator of economic activity other than the yield spread itself is the Institute of Supply Management 's (ISM) New Order diffusion index. The ISM new order index leads industrial production by about 5 months and employment by 9 months. Typically, the yield spread leads the ISM new order index by 3 months and the two have a fairly good record of tracking well. (See Chart 1)

As you can see, the ISM new order index is unusually high relative to the current yield spread. There is not an historical period where this relationship was so “out of kilter” when the yield spread was this small. Most of the other ISM diffusion indexes, including production and employment are above 50 (indicating expansion) and were stable during the month of August. This suggests a manufacturing economy that is still expanding and stable.

There have been many anecdotal reports of a slowing consumer economy. Housing, casual dining, and large ticket recreational goods have all been suffering. These are the marginal impacts of high gasoline prices and higher short term interest rates. The core driver of consumption, however, is real compensation which accounts for 68% of personal income. Compensation is composed of wages and salary disbursements and supplements to wages and salaries. The components of price-adjusted compensation are aggregate hours worked, compensation per hour, and the inflation rate. Aggregate weekly hours worked and the ISM employment index is steady as a rock. (See Chart 2). With hours worked growing at 2.6% and compensation per hour at 3.9%, total compensation is expanding at 6.9%. When inflation is deducted, real total compensation is expanding at 3.2%. This is a healthy level of income for consumers. Nonetheless, if the realized inflation rate (which includes food and energy costs) declines, it will create more real disposable personal income. From this source come personal outlays and savings. Although the savings rate should increase over time, personal outlays should remain strong enough to support a healthy economy.

Although the news of a lower rate of inflation for the consumer price index excluding food and energy (commonly referred to as CPI core) in August was most welcome, we don't expect many repeats of that for the remainder of the year. We analyze core CPI inflation by separating it into its two major components: goods and services. Goods inflation is primarily correlated with (1) the producer price index excluding food and energy (PPI core) inflation for finished goods and (2) import prices for autos and non-auto consumer goods. PPI core inflation for finished goods tends to lead CPI core goods by 6 months. (See Chart 3) The change in U.S. dollar tends to lead import auto and consumer non-auto prices by 6 months as well. (See Chart 4) It should come as no surprise that the market share of imported consumer goods has been growing as well. Over the last decade imported consumer goods as a percent of personal consumption expenditures for goods (excluding food and energy) increased from 23.2% to 31.7%. If the dollar were to significantly weaken, import prices would evolve from the deflating prices observed during the 1996-2004 strong dollar period to a period of inflating prices . Put another way, a strengthening Japanese yen or Chinese yuan would have a deleterious impact on inflation.

Given the recent weakness in the dollar, we would expect some upward pressure on imported consumer products over the next six months. On the other hand, the recent decline in PPI core inflation (caused by the decline of core intermediate producer goods) should have a beneficial impact on the core CPI goods inflation through March of 2007. However, goods only account for only 29% of core CPI inflation. The other 71% is accounted for by services.

The core CPI service index is dominated by “rent of primary residence” and “owners' equivalent rent of primary residence” (OER). Combined these two items account for 53% of core CPI services (and 38% of core CPI). Both have been trending upwards since 2004. Owners equivalent rent (the dominant of the two) rose from a three month rate of change of 2.7% ending last February to a 5.0% rate of change as of last month. This is the source of much of the “bulge” in the core CPI. A great deal of this strength comes from its mode of calculation. The Bureau of Labor Statistics (BLS) calculates the “pure rent” by removing the value of any landlord-provided utilities and furniture. Thus when piped natural gas or electricity prices rise it suppresses the OER figure reported by BLS. Conversely, when energy prices fall, it enhances the figure. As Chart 5 shows that while this inverse correlation is not always consistent in terms of magnitude, there is a definite relationship.

Of even greater importance, however, is the causation of other changes in OER, i.e. true rent. During the housing boom of the mid-1990's to early 2000's, rental vacancy rates increased steadily from 6.9% (1993-IV) to 10.4% (2004-I). This was particularly evident in single family rental units where the vacancy rate more than doubled. Since last year's fourth quarter the vacancy rate has begun to decline. The National Association of Home Builders reports that “builder confidence in current rental apartment market conditions climbed to a new high in the second quarter of 2006”. The best index that illustrates what is driving rental rates is the Housing Affordability Index (produced by the National Association of Realtors). Over the last 18 months this index has fallen to levels not seen since 1991. When plotted inversely against the change in owner's equivalent rent, it paints a picture of potentially higher rental rates over the next few years. (See Chart 6)

Our purpose in developing these insights into the lead indicators of inflation is to model them in such a way that we have a reasonable idea of how the next twelve months of inflation is likely to play out. Chart 7 depicts our best estimates of core inflation for goods and services through 2007. When these components are combined our model suggests that core inflation, which has increased at a 2.74% rate over the last 12 months, will increase at a 3.1% rate by next spring. If this occurs it should have a negative impact on long term bond yields; however, there are other forces at work which could mitigate this reaction. (We'll write about this in Part II of For What It's Worth).

As for the interest rates set by the Federal Open Market Committee, it appears that they have returned to the neutral ground of the last twenty years. In other words, the current short term interest rate (using the constant one year Treasury note index as a proxy) relative to anticipated inflation (the next 12 months) is exactly at its historical norm. (See Chart 8) And if lower prices for gasoline offset the depressing impact of lower housing prices, the Fed will have no urgency to reduce interest rates in the near future.