Across the Curve

JPMorgan on Industrial Production

  Industrial production increased 1.3%m/m in October, considerably above
   expectations (JPM: 0.6%, consensus: 0.2%). However, September was
   revised down from -2.8% to -3.7%, which is the biggest monthly decline
   since 1946. The net result is that the 4Q run-rate on IP was tracking
   -9.8%q/q annualized through October, close to our original expectation.
   Growth in 3Q was revised down to -7.8%q/q from -5.8%.

   The big swings in IP in September and October are the result of certain
   technical factors. According to Federal Reserve estimates, the combined
   effects of September hurricanes and a strike at Boeing reduced September
   IP by 3.0 %-points. The post-hurricane recovery in selected industries
   then lifted output by about 2.0 %-points in October. Excluding all these
   distortions, the Fed estimated that total IP would have fallen by about
   two-thirds of a percent in both September and October. Had that actually
   occurred, IP would have fallen 3.7%q/q annualized in 3Q and be on track
   for a 8.7% fall in 4Q. Thus the decline in IP was clearly intensifying
   in 4Q when one-off factors are stripped out.

   ·  Manufacturing production increased 0.6%m/m in October but the
   run-rate for 4Q was -12.7%q/q annualized, considerably worse than the
   -7.8% reading in 3Q. The quarterly run-rate is somewhat misleading
   though, because it only includes data through October and thus
   completely misses an expected bounce in aircraft production in November,
   the result of the end of the Boeing strike.  Because of the recovery in
   aircraft output, we still expect that 4Q manufacturing output will have
   fallen around 8.0%. If our forecast is realized, the combined
   two-quarter drop in output in 3Q and 4Q would have been the worst since
   1981-82, though a two-quarter stretch during 1990-91 would have been
   almost as bad.

   ·  Business equipment output, a correlate of capital spending, fell
   2.2%m/m in October and was tracking a huge 35.4%q/q annualized drop in
   4Q. One of the reasons the 3Q figure looks so bad is because this
   category includes aircraft. But there was also very significant weakness
   in other categories. Business vehicles were tracking a 35.9% 4Q decline,
   high-tech equipment a 5.2% decline, and industrial and other equipment a
   14.4% decline.

   ·  Computer and peripheral output, the one part of this report which is
   used directly in the GDP report, was revised down to -3.2%q/q annualized
   in 3Q from +4.3%. Combined with the results of the September
   international trade report, this suggests that 3Q real equipment and
   software investment will get revised down from the originally reported
   -5.5% to about -7.5%. This also leaves our tracking estimate of 3Q GDP
   at -0.7%, compared to -0.6% before this report. The BEA’s original
   advance print was -0.3%. Computer output also looks like it will be even
   weaker in 4Q. It fell 1.1%m/m in October, the lowest reading since 2004,
   meaning that 4Q output was tracking a decline of -10.5%. That would be
   the worst quarterly result since 2Q04.

   ·  Motor vehicle and parts production declined 3.5%m/m in October as
   vehicle assemblies dipped 3.8%.  Despite very weak auto sales numbers,
   production schedules provided by Wards Autos show an 11.9%m/m increase
   in assemblies in November (seasonally adjusted). However, we would not
   that actual production has often undershot the schedules in recent
   months.