Comment 43: Inflation fear: the outlook of CPI-core in 2007 (May 19, 2007)

Comment 43: Inflation fear: the outlook of CPI-core in 2007 (May 19, 2007)

April 2007 CPI (consumer price indexes) report was released by U. S. Bureau of Labor Statistics on May 15, 2007. After the release Dow Jones Industrial Average surged more than 100 points. Various financial media, quoting experts, reported that the stock price surge was the result of a subdued inflation picture and the hope that FED will lower short-term interest rates soon. At the same time of the stock price surge the yield of long-term U.S. treasuries jumped more than a few basis points. The bond market has apparently viewed the report as painting a picture of a worsening inflation. Observant readers must have noticed the clear contradiction between the view of the bond market and the reports of financial media already. In this comment we will analyze the trends of CPI-all (CPI including food and energy) and CPI-core (CPI excluding food and energy). We will confirm the view of the bond market and estimate the outlook of CPI-core in 2007. At the end of this discussion we will also discuss the seemingly odd behavior of the stock market, that is, celebrating a worsening inflation picture.

The major argument of the financial media for a subdued inflation picture is based on the year-to-year comparisons of CPI-core. The CPI-core of April 2007 advanced by 2.4% compared to the CPI-core of April 2006, whereas the CPI-core of March 2007 advanced by 2.5% compared to the CPI-core of March 2006, thus a picture of calming CPI-core. However, year-to-year comparisons of CPI's are looking at the inflation picture of the past 12 months in average, not the current trends of inflation. In Comment 30 we have used year-to-year comparisons of CPI-all and CPI-core to establish an observation that sharp moves of CPI-all, mostly caused by violent movements of the price of crude oil in recent years, take about 6 months to be reflected in CPI-core; interested readers are referred to the section that contains an updated graph of year-to-year comparisons of CPI-all and CPI-core. According to this observation the sudden descend of CPI-all in the fall of 2006, prompted by the sharp drop of crude oil prices, is already reflected in the quieter rate of growth of CPI-core in recent months. The problem is that this calming of the growth rate of CPI-core in March and April of 2007 is not large enough to prevent an expected sharp upturn of CPI-core in coming months, reflecting the strong rebound of CPI-all from its nadir in the fall of 2006.

Let us look into the actual data of CPI's to see how we have arrived in our conclusion. The normal way to study the current situation of CPI's is to look at their month-to-month percentage changes. As pointed out in Comment 30 the month-to-month changes are very jumpy so that the derivation of the underlying trends is difficult. On the other hand any attempt to smooth out the gyrations in the month-to-month comparison data, like using the year-to-year comparisons, will sacrifice the timeliness. The best way to study current trends of CPI's is to plot the indexes directly on a graph in logarithmic scales as done in the graph at the right. In the graph CPI-all is plotted as the blue curve, and CPI-core as the red one. Since a straight line on a logarithmic scaled graph means a constant rate of change, the graph is ideal for drawing trend lines. The green trend line B is for CPI-all. The trend line dates back to December 2003, and represents a growth rate of 3.2% per year. Data points of CPI-all from December 2003 until May 2005 all lie above but close to this trend line B, implying that during the period CPI-all in average grew 3.2% per year. Another trend line C can be drawn for CPI-all starting from June 2005 and representing a growth rate of 4.0% per year. All CPI-all data points from June 2005 to August 2006 lie above this trend line C, indicating that CPI-all in that period grew in average 4.0% a year. This heightened growth rate of CPI-all is due to the rapid rise of crude oil price. The sharp drop of crude oil price in the fall of 2006 brought down CPI-all with it, and caused CPI-all to break through both the trend line C and the trend line B; CPI-all hit a bottom at October 2006. CPI-all bounced back from that bottom and rose consecutively since November 2006. It has moved above the trend line B in March 2007 and moved up further in April 2007. The CPI-all is under strong influence of the crude oil price, and thus is not predictable for us since we do not possess the capacity to predict near-term gyrations of the price of crude oil.

A trend line D can be drawn for CPI-core. This trend line D dates back to January 2004 and represents a growth rate of 2.2% per year. CPI-core grew close to this trend line D from January 2004 to February 2006. From February 2006 to March 2006 there was a sharp rise of CPI-core. After that one-month sharp rise, CPI-core grew parallel to the trend line D again, implying that for all the time except that one-month sharp rise CPI-core has grown roughly with a rate of 2.2% per year. A close inspection of CPI-core reveals that the tiny peak at February 2007 probably corresponds to the peak of CPI-all in August 2006, exactly a 6 month delay. If that is so, then the sharp drop of CPI-all should have been felt in CPI-core in its March and April 2007 data already. We can see from the graph that the effect of CPI-all's sharp descend in the fall of 2006 is having only a minor effect on CPI-core. From May 2007 the effect of the bounce-back of CPI-all from its October 2006 nadir should be felt in CPI-core. If this bounce-back of CPI-all causes CPI-core to rise faster than 2.2% per year, a new trend line with a steeper slope than the trend line D can be drawn for CPI-core, starting from December 2005. This new trend line, if can be drawn, implies that CPI-core is growing with a rate larger than 2.2% per year. Then this new rate of growh of CPI-core will be substantially above the comfort zone of FED, the upper limit of which is considered to be 2.0% per year.

As has been discussed in Comment 41, stock markets in recent months have been under the spell of yen carry trades; some speculative hedge funds borrow yens at very low interest rates, sold those borrowed yens for dollars, and then use those dollars to buy stocks. Therefore, as dollar becomes stronger against yen, a sign of the installment of yen carry trades, stock market rally follows. Higher yield of long-term U. S. treasuries is usually the catalyst to induce yen carry trades. Once the general yen-carry-trades start, those speculative hedge funds then piggy-back on the general move to undertake their own yen-carry-trades for the purpose of buying stocks. That is the reason why stock markets rise in tandem with higher yield of long-term treasuries, looks like celebrating higher inflation.