Recent Housing and Inflation Data Support Fed Pause
By Chris Puplava, September 20, 2006
The Fed lowered interest rates to a 46-year low of 1% and the prime rate stood at 4% before the Fed began a two-year credit campaign to raise rates that began in June 2004. The Fed boosted rates a record 17 consecutive times before deciding to pause at the last meeting on Aug. 8. Today the Federal Reserve decided to keep the Fed Funds rate at 5.25% with ample economic data to support their decision.
"With the lower inflation numbers and the weaker housing data, the odds have shifted towards the Fed doing nothing for the rest of the year," said David Wyss, chief economist at Standard & Poor's in New York. Wyss expects that the next Fed move would likely be a rate cut sometime next year. In its statement, the Fed continued to voice its concerns about inflation by repeating a phrase it had used at last month's meeting, that the Fed's rate setting panel "judges that some inflation risks remain."
On the inflation front, the PPI release yesterday showed that the overall PPI rose 0.1% in August, below the consensus of a 0.3% rise, with the core PPI falling 0.4% (Figure 1). The year-over-year (YOY) rate in finished goods has decelerated from the 5.7% rate seen in January to 3.6% in August with the core finished goods falling to a 0.9% YOY rate (Figure 2).
Figure 1. PPI
Figure 2. PPI: Finished Goods
Source: Dismal Scientist
Further support of an interest rate pause is the continued decline seen in housing. The National Association of Home Builders HMI report Monday showed a decline to its lowest point since February of 1991, marking its eighth consecutive fall and down more than 50% since its peak in June 2005. The correlation between the HMI and the S&P 500 is startling, and the 50% drop seen in the HMI does not bode well for the S&P 500.
Source: John Mauldin, "Fingers of Instability - Thoughts from the Frontline"
Further support of a rapidly declining housing market came from yesterday's release of housing starts. Housing starts decreased 6% in August from July's levels to 1.665 million units, down 19.8% from August of 2005. Total housing starts are lower by more than 20% compared to a year ago in three of the four regions of the nation. The drop is largest in the Midwest (-29%) and West (-28%), followed by the Northeast (-24%); while the South is down a much more moderate 10% from year-ago levels.
In my last article ("Housing and Retail Trends Likely to Weigh on Future Employment Growth") I analyzed previous housing and employment cycles using existing home sales data. Below is a similar analysis using housing starts with data from 1960 to the present. Two separate time periods are shown below with housing starts and the YOY percent change.
Figure 5. 1960-1982
Figure 6. 1982-Present
Looking at the cycles over the charts above shows that housing starts have averaged 2133 thousand units since 1960 at their top, with 893 thousand units at the bottom with a lag of 40 months from peak to trough (3 years and 4 months). The YOY percent change typically peaks after breaching 30%, with the average since 1960 at 61% and the bottom occurring between 40-50%, with the average at 45% and a peak to trough lag of 51 months (4 years and 3 months). Looking at the charts above shows that the YOY percent change typically peaks prior to the housing starts' peak and bottoms fairly close to the bottom in housing starts. The data is given in the table below.
Table 1. Housing Starts and YOY % Change
Note: Current data is given as the bottom for the current cycle only for comparison.
As seen in the table above, housing starts decline by 58% from the average high of 2133 thousand units to the low of 893 thousand units. August housing starts came in at 1665 thousand units, down 26% from the January 2006 high, a little less than half way of the average decline of 58%, indicating more slowing is expected. The current cycle that began in January 1991 and peaked in January 2006 was the longest of the cycles since 1960. The previous cycle, which began in November 1981 and peaked in February of 1984, took nearly seven years to finally bottom. Two separate bottoms of roughly 20% declines in 1984 and 1988 were seen before the final bottom in 1991, with a 49% decline.
The months supply of houses was also analyzed since the 1960s to determine peak and trough points in housing cycles. The months of supply of unsold homes bottoms prior to housing starts peak as demand is unable to keep up with new construction (supply) and begins to rise. The rising supply of homes and weakening demand increases the months supply of housing inventory until housing starts decelerate to reduce the number of homes for sale and demand picks up. As seen from the charts below, the months supply of houses peaks coincidently with the bottom of housing starts.
Figure 7. 1962-1982
Figure 8. 1982-Present
Looking at the charts above shows that the peak months of supply of homes from the housing cycles since 1960 ranges from seven to eleven and a half months. The two housing cycles in the 1960s peaked with housing starts less than or equal to 1800 thousand units, and bottomed with the peak supply in homes at just over seven months. Housing cycles since then have peaked equal to or more than 2200 thousand units with the inventory of unsold homes peaking at nine and a half to eleven and a half months. The data is given in the table below with current housing starts and supply given for comparison, although a true bottom has not been seen.
Table 2. Housing Starts and Inventory
As the current housing cycle's peak came above 2200 thousand units, the peak months supply of housing inventory should range between nine and eleven months, with the current inventory at 6.5 months. We will not likely see a bottom in housing starts until we see a peak in the months supply of housing inventory, which would indicate that demand is outstripping supply for new homes. Housing starts are currently down 20% from last year's levels and a bottom may form near a 40% YOY% as it has in previous cycles, indicating there is still room for more damage to be done in the housing sector.
The markets moved into positive territory in early morning trading as crude oil fell on a larger than expected build in distillate inventories to their highest levels since January 1999. Other support lifting the markets in morning trading was earnings from Oracle Corp. that topped analyst estimates and earnings for Morgan Stanley, which beat expectations with its best third quarter ever.
The markets continued their upward advance leading into the announcement of the Fed meeting but declined after the decision was released. The markets regained traction in afternoon trading and reversed course recovering most of the initial reactionary drop.
Advancing issues represented 62% and 60% for the NYSE and NASDAQ respectively, with up volume representing 63% and 77% of total volume on the NYSE and NASDAQ.
All of the broad market indices were up, with the DJIA posting a gain of 72.28 points to close at 11,613.19. The S&P 500 was up 6.87 points to close at 1325.18, and the NASDAQ was also up, rising 30.52 points to close at 2252.89. The 10-year Treasury note yield fell to 4.729%, and the dollar index posted a loss on the day, falling 0.09 points to close at 85.86.
Overseas markets were mostly up with Brazil's Bovespa and Japan's Nikkei stock average posting losses of 1.92% and 0.98% respectively. Some of the biggest moves came from France's CAC-40 and Germany's DAX Index, up 1.50% and 1.38%. Mexico's Bolsa Index and London's FTSE 100 posted gains of 0.84% and 0.59% respectively.
Eight out of the ten S&P 500 sectors were up on the day with the two declines coming from energy and utilities, down 2.35% and 0.20%. The largest gain came from the technology sector on the Oracle Corp. news, up 1.65%. The telecommunications and consumer discretionary sectors posted strong gains, up 1.60%, 0.89% and 0.71% respectively.
Have a pleasant weekend,
© 2006 Chris Puplava