Sunday, January 26, 2014

the intergenerational mobility kerfuffle; December existing homes sales

there really weren't any important economic releases this past week....the most widely covered report, on existing home sales from the National Association of Realtors (NAR), may be important to those involved in the real estate trade and to those financial institutions who are making the real estate loans, but they entail no new investment and few additional jobs compared to what's generated by construction and sales of new residential property...and other than that, there was just the Kansas City Fed regional manufacturing report for January, wherein the composite index came in at +5, up from -3 December, indicating mild expansion by manufacturing industries in the 10th District, and the Markit Manufacturing Flash PMI for January, an early read on next week's full release, which indicated a drop from December's 55.0 to three month low at 53.7, but similarly mildly expansionary nonetheless...there was, however, something of a hullabaloo in the academic econo-blogosphere over a new report on income mobility in the US by a handful of young hotshot economists, and since everybody and his brother latched onto it to prove what they already been preaching, we'll take a brief look at that study to see what all the commotion was about...

US Intergenerational Mobility Remains Below that of Most Rich Nations

the title of the paper in question is Where is the Land of Opportunity? the Geography of Intergenerational Mobility in the United States and it's a 94 page PDF, fairly heavy on math and computed tables, so i doubt if any of those commenting on it read it in its entirety, although there's a one page executive summary (pdf) that's fairly accessible...what the economists apparently did is examine tax records of more than 40 million US children and their parents from those born in 1971 to those born in 1993 to determine how much income mobility there was between one generation an the next....basically, both parents and their children were divided into quintiles by income, which is to say the top 20% of incomes by age group in the US, the next 20% by income, and so on till the bottom 20% of incomes, and then the change in income quintile from one generation to the next for each year was computed for each group...what they found was that recent intergenerational mobility has not changed much from the 70s; when there was an 8.4% chance of children whose parents were in the lowest bracket of incomes to make it to the top 20% when they were adults; similarly, there has been a 20% chance of children born into the middle income group to make it into the top fifth as adults, also largely unchanged over the last decade...comparing their results to earlier studies of a similar nature, they conclude that income mobility in the US hadn't changed much at all between the 50s and 90s…however, intergenerational mobility in the US still remains below that of Canada, Australia, and most of the rich European nations; as one of the lead authors notes in the New York Times, the odds of escaping poverty appear to be only about half as high in the United States as in the most mobile countries like Denmark..

a few caveats: obviously, this study, ending with those born in 1993, is hardly capturing the effect of the recent financial bubble and subsequent great recession over the lifespan of those now growing up in it...it simply can't tell us anything about the income mobility of those born in the last 20 years vis-a-vis their parents...furthermore, for those born in the years after 1986, college attainment has been used as a proxy for income for the ranking of those individuals within quintiles, since in earlier years college was a fair predictor of income...with all the recent anecdotes about college graduates trying to pay off their loans flipping burgers, it's hard to say if that remains true...also, by using the top quintile, or the top 20%, as its definition of wealthy, it ignores the sharp divisions that have occurred within that quintile, wherein the incomes or the top 1% have pulled away from the pack...one of the authors of this study, Emmanuel Saez, is the co-author of another recent study that showed that 95% of the incomes gains over the 2009 to 2012 period went to the top 1%, while the bottom 99% saw income growth of just 0.4%....in addition, we should also note that by focusing on incomes, it says nothing about inherited wealth, a divide between those born to wealth and those who've worked for it that's only going to get worse with the recent cuts & exemptions to estate taxes that came with last year's fiscal cliff deal...

the one aspect of this study that's most interesting is the very wide difference in upward mobility between children born in different regions of our own county...the map below is a picture of an interactive graphic from the New York Times which was developed from the report which shows the percentage chance that someone born into the lowest quintile might reach the top quintile by adulthood; in the darkest red shade, that chance is less than 4%, and includes several pockets in the South, including Atlanta at 4% and the Memphis environs at 2.6%; those born in the areas with lighter red shadings, including much of the South and Detroit at 5.1%, Cleveland at 5.2%, and Chicago at 6.1%, all have less than 10% chance of rising to the top fifth from the bottom...the yellow shading, including most of the rural areas and cities such as Salt Lake at 11.5% and San Francisco at 5.2%, all have upward mobility chances greater that 10%, while those born in green areas with a 20% chance of making it to the top include some smaller western plains cities such as Casper Wyoming at 19% and Yankton S. Dakota at 22.8%...and in a telling revelation not specifically noted in any of the literature but clearly obvious on the map, children born dirt poor in the western counties of North Dakota shown in blue as have had as much as a 35% chance of reaching the top quintile, with those born in the heart of the Bakken oil field in Williston North Dakota having a 33.1% chance of earning top bucks as adults despite being born into families at the bottom...

upward mobility map

Contraction in Existing Home Sales Confirmed by December Data

as we mentioned, the one report this week that did get widespread coverage was on December Existing-Home Sales from the National Association of Realtors (NAR), which indicated home resales were at a seasonally adjusted annual rate of 4.87 million in December, 1.0% above November's revised level, but 0.6% below the level of December a year ago...as with any report from a trade group such as this, we have to get past the spin in the press release, which headlined that 2013 was the best year for home sales in seven years, and look at the hard data to determine what's really been going on recently...first we should note that November's sales were originally reported at a seasonally adjusted annual rate of 4.90 million, down from a 5.12 million rate in October, which itself was down 3.2% from the 5.29 million annual rate in September; which ultimately means December home sales were 8.0% lower than sales of three months ago, and down 9.7% from the 5.39 million manual rate hit in July and August, a sales downturn which is clearly visible on the FRED graph below, which tracks the annual rate of homes sold monthly over the past ten years...also clearly noticeable on that graph are the spikes in home sales in 2009 and 2010 that resulted from the first time home buyer tax credits of those years...

FRED Graph

December existing home prices home prices have been moderating seasonally along with the decline in home sales...the median home sales price in December was $198,000, up $2,500 from the $195,500 median sales price in November, but not much change from the $197,500 median realized in October or the median sales price of $198,500 in September...those recent prices are down roughly 7.5% from the peak median sales price of $214,000 of June sales, a fairly typical price change from summer to winter....the average price realized for homes sold in December was $246,800, up from the $243,600 average of November and the highest average sales price since August, so we can assume there was a greater proportion of McMansion type homes sold in the mix in December than in the months preceding it...the pie graph to the right, which was taken from the Summary Statistics graphics for December from the NAR (pdf) shows the percentage of homes sold in each of several price ranges in December; the orange wedge at the top indicates that 2% of homes sold in December exchanged hands at over $1 million, while the teal blue wedge next to it indicates that another 2% sold for between $750,000 and $1 million...on the other end of the spectrum, the dark blue wedge in the upper right indicates 18% of homes sold were below $100,000, while the largest dark red wedge indicates that 44% of homes sold in December went for between $100,000 and $250,000  (click to enlarge)

the FRED graph below shows the track of the median, or middle sales price of previously occupied homes over the past ten years in blue, and the track of the average existing homes sales price over the same time span in red…the seasonal pattern of house prices is quite obvious, wherein home prices typically peak in the summer months, fall in the Fall, tick up in December, and then bottom out in February and March….the slowing of price increases in recent months suggests that lagging home price indexes, such as Case-Shiller (which last reported prices for August to October) should start to see more modest year over year increases going forward…

FRED Graph

so it’s finally become apparent that the spike in interest rates that was precipitated by the Fed’s taper talk has had an impact on both sales and prices...average interest rates on 30 year fixed rate mortgages rose to 4.46% in December, from 4.26% in November, not much out of the 4.19% to 4.49% range they've held for the last half of 2013; over the first five months of 2013, before the June FOMC meeting when the eventual Fed withdrawal became obvious, interest rates on a 30 year fixed mortgage had averaged 3.50%...these higher rates may be part of what's reducing the percentage of first time home buyers, which fell to 27% of purchases in December, down from 28% in November and 30% a year ago...buyers classified as investors are picking up the slack; they accounted for 21% of December sales, up from 19% in November...and according to the NAR, 32% of all sales in December were all cash sales, unchanged from November but up from the 28% of all home sales that cash transactions accounted for a year earlier...this is in contrast to a report from RealtyTrac, which says all-cash purchases accounted for 42.1% of home sales in December, up from 38.1% in November, but in contrast to just 18.0% of sales that were all cash in December 2012...RealtyTrac also finds that more than 50% of homes in Florida, Wisconsin, Alabama, South Carolina, and Georgia were bought with cash only, with nearly 63% of home sold in Florida purchased on a cash only basis...

(the above is my weekly commentary that accompanied my sunday morning links emailing, which in turn was mostly selected from my weekly blog post on the global glass onion…if you’d be interested in getting my weekly emailing of selected links that accompanies these commentaries, most coming from the aforementioned GGO posts, contact me…)

Sunday, January 19, 2014

December retail sales and consumer prices and their impact on 4th quarter GDP; November Mortgage Monitor

this week we'll start by looking at the Advance Monthly Retail Sales Report for December from the Census Bureau, which estimated that our total seasonally adjusted retail and food services sales were at $431.9 billion, which would be 0.2 percent (±0.5%)* higher than November sales...before the seasonal adjustment, December's sales were estimated at $488.7 billion, up 13.2% from November's $431.8 billion, as we would expect for December...the adjusted increase in November sales was revised from a +0.7 percent (±0.5%) increase from October to a 0.4 percent (±0.2%) increase, so on net this report reflects a 0.1% contraction from what was previously reported...also note the range of accuracy for December of ±0.5% and the Census asterisk, which advises us that they have insufficient data to determine whether sales actually rose or not in December, but that they're 90% confident the change in sales from November to December was between a decrease of 0.3% and an increase of 0.7%, and like the November report, will likely be revised when a larger sample of retail establishment sales are included in the next estimate...despite the high degree of uncertainty in the data, this is a widely watched indicator, and ultimately will be used to generate real goods consumption for the first estimate of the 4th quarter GDP, so we'll play along and look at these estimates as if they were exact...

we'll begin by including below a picture of the table of monthly sales percentage changes by business type from the Census pdf...there are essentially three double columns in that table, and we'll use them all today; the first double column gives us the percentage change in sales from November to December in the first sub-column, and then the year over year percentage change since last December in the 2nd column; the second double column set below gives us the revision of November's advance estimates (now "preliminary") as of this report, with the October to November percentage change under "October 2013 revised" and the November to November percentage change as revised in the 2nd column of the pair (our picture of what those November percentages looked like before this revision is here)...the third pair of columns shows the percentage change of the last 3 months - October, November and December sales - from the preceding three months (July through September) for each business type listed; this is effectively the percentage change in retail sales from the 3rd quarter to the 4th quarter overall...and finally, the last column shows the year over year percentage change in retail sales from the 4th quarter of last year (Oct 2012 to Dec 2012) to the 4th quarter of this year…

December 2013 retail sales

we'll be starting with the current month over month changes, as shown in the first column above...as has been the case in the two previous months, the change in vehicle sales were a major factor in the overall December change; seasonally adjusted sales at motor vehicle and parts dealers were at $81,690 million, down 1.8% from November sales of 83,229 million; without the fall in automotive sales, all other retail sales were up a respectable 0.7%, led by a 2.0% increase to $55,580 million at food and beverage stores; there must have been more holiday partying than normal...sales at clothing stores were up 1.8%, from $21,180 million to $21,562 million, while gasoline station sales rose 1.6% to $45,327 million, with their year over year sales turning positive at up 0.6%...non-store or online retailers also had a sizable December sales increase, as their sales rose 1.4% from $38,613 million in November to $39,141 million in December...several retail sectors saw sales fall; seasonally adjusted sales at electronics and appliance stores fell 2.5%, from $8,578 million in November to $8,364 million in December; department store sales, not including leased departments, were down 0.7% to $14,453 million, although the entire general merchandise group rose 0.1% to $55,384 million...sales at specialty shops, such as sporting goods, book and music stores, also fell 0.6%, from $7,812 million to $7,766 million, while sales at furniture stores fell 0.4% to $8,703 million...and sales at building material and garden supply stores also fell 0.4%, from $25,880 million to $25,780 million...understand that these are the seasonally adjusted sales, meaning sales rose or fell by that percentage over a normal December; ie,  building material sales as reported were down 6.7% to $22,988, as you'd expect for a colder month, while electronics sales rose 24.2%, as you would expect before the holidays; what the seasonally adjusted change tells us is that building materials normally only fall 6.3% in December, while electronics sales were expected to rise 26.7%...

the pie graph below, from Robert Oak at the economic populist, puts the sales of each of these business types in this report in perspective; clearly, motor vehicles sales at 19% have the largest impact on the report; but note there are two large wedges that relate to food sales; grocery stores at 13% and restaurants and bars at 11% of sales, so almost a quarter of retail sales is food and drink…and also note that the big box stores still command a significant portion of sales at 13%, while sales at specialty stores (ie books, sporting goods), electronic and appliance stores, and furniture stores, at 2% of sales each, rarely change enough in one month to impact the overall report…

December 2013 retail pie via robert oak

Revisions Leave December Retail Sales 0.1% Below November Advance Report

as we mentioned, there were major revisions to November sales as well, as we originally reported they increased by 0.7% only to find the increase has been revised to a 0.4% gain...since the change in seasonally adjusted vehicle and parts sales was unrevised as an 1.8% increase, the retail sales change ex-autos was revised down from a 0.4% increase to just a 0.1% sales gain....major changes from what we reported last month include sales at electronic and appliance stores, which were originally reported to have shown a 1.1% increase in November, have now been revised to show a 2.1% decrease; that, combined with what was apparently a good December last year, has turned year over year sales for electronic stores negative; in November, they had been showing a 6.8% year over year sales increase, while they’re now showing a now showing a 1.4% annual decrease from December 2012…sales at furniture stores, which were originally reported up 1.2% over October, are now seen to have fallen 0.2%; building material and garden supply stores, which were originally thought to have seen sales increase by 1.8%, have now seen their November sales gain cut to 0.4%...in addition, November sales at gasoline stations have been revised to show a 1.5% decrease from the 1.1% drop first reported, the sales decrease at clothing stores was revised from 0.2% to 0.5%, drug store sales fell 0.3% instead of being unchanged, and non-store (online) sales were reported up 2.2% and they're now seen as having risen only 1.6%....however, sales at sporting goods, hobby book and music stores, which were first reported to have increased only 0.1%, have been revised to show November sales 1.1% over October's, and miscellaneous store retailers, whose sales were first reported to have dropped 1.3%, have been revised to show a sales increase of 0.8% in November..

it’s also apparent from the new sales totals of the past three months that there have been notable downward revisions to October retail sales as well, which had been revised to a preliminary $429.4 billion with the November report; they are now shown to be just $429.0 billion, which means that the percentage increases for both November and December were off a lower base...as of that November report, we were led to believe that retail sales had risen 0.6% in October and 0.7% in November, which led us to believe that we’d see a decent 4th quarter boost to GDP from personal consumption expenditures…however, as we can now see from the third column of the initial table that we included above, the total growth in sales for all three months over the third quarter is now just 1.0%, which doesn't even match the growth we previously thought we had out of just two months….of course, to determine the effect of this weaker report on GDP, consumption of goods for the 4th quarter GDP will need to be adjusted for inflation...the excel formula used by the BEA to do that is somewhat complex, however, and their process is certainly automated…we can, however, get a fair approximation of real consumption by arithmetically adjusting retail sales with the consumer price index...since that report was also released this week, we'll look at those numbers next, before we get back to adjusting retail...

Consumer Prices Rise Most in 6 Months on Higher Fuel Costs

the seasonally adjusted December Consumer Price Index for for All Urban Consumers (CPI-U) from the Bureau of Labor Statistics indicated that overall prices rose by 0.3% in December, the largest increase in the index in 6 months, mostly due to increases in fuel prices and rent...prices were up 1.5% over the year, up from an annual inflation reading of a bit over 1.2% last month...the monthly increase was entirely in the seasonal adjustment; the unadjusted price index, which is based on 1982-84 prices equal to 100, was at 233.049, down fractionally from the 233.069 reading in November; last December the index stood at 229.601...the seasonally adjusted Core CPI, which is all items except for food and energy, rose just 0.1%, with its year over at 1.72% (rounded to 1.7%), unchanged from November...similarly, the December index value for core CPI before seasonal adjustment was below November's 235.243 at 235.000...

a 2.1% increase in the energy index drove the overall CPI index higher...prices for energy commodities were 3.4% higher than November, with gasoline prices 3.1% higher, fuel oil up 2.4%, and propane and kerosene up 4.1%...prices for energy services were 0.2% higher, with electric rates up 0.4% and utility gas down by the same amount...for the entire year, however, the energy index only increased by half a percent, with the energy commodity index down 0.8% while energy services were 3.4% higher; gasoline in december cost 1.0% less than a year ago, while fuel oil was 1.8% lower...piped utility gas was 0.1% lower for the year as well, but the national average of electric rates has risen 3.2% since last December....

the food price index rose 0.1% in December, the same tiny increase as logged in November and in October; the food at home index was unchanged for the third time in four months, while the food away from home index rose by 0.1%, as full service restaurants saw prices rise 0.2%, while prices at fast food restaurants rose just 0.1%...of the major food at home groupings, prices for cereal & bakery products were 0.1% lower, as prices for white bread fell 1.0%, rice, pasta and cornmeal were 1..4% lower, while flour was 0.9% higher and the price of crackers rose 1.4%...the price index for meats, poultry, fish, and eggs rose 0.3% on pork prices that were 1.1% higher with breakfast sausage seeing a 4.7% price jump, while fish and seafood prices were 0.4% lower on a 1.7% drop in prices for fresh seafood...dairy products also rose by 0.4% in December as milk prices rose 0.9%, cheese and ice cream prices rose by 0.5%, and prices for other dairy products were unchanged... the fruit and vegetable index, however, was down 1.5% for the month, its largest drop in five years, as prices for fresh vegetables fell by 2.7% led by a 4.2% drop in lettuce prices, while prices for processed fruits and vegetables rose 0.4%..meanwhile, beverage prices rose 0.5% as noncarbonated juices and drink prices rose 1.8% while coffee prices were off 0.4%, and prices for other food at home rose by 0.3% on butter prices that were 2.6% higher, prepared frozen foods that rose 1.2%, and sugar prices that were 1.2% lower...for the year, the cost of food at home has risen by a very modest 0.4%, with declines in prices for dairy products, fruits and vegetables, and beverages...only meats, poultry, fish, and eggs saw a sizable rise rise of 2.9% for 2013, with pork up 4.5%, poultry up 3.0%, seafood up 4.6% and egg prices 5.8% higher than a year ago...

other price changes in December were modest, with prices for many commodities drifting lower...the index for shelter, which is nearly 32% of the CPI, rose 0.2%, with rent of shelter rising 0.3%, as rents rose 0.3% and owner's equivalent rent rose 0.2%...prices for apparel, which had been down 0.3% year to date, rose 0.9% on a 2.1% increase in prices for women's clothing while footware prices fell 0.5%...meanwhile, the index for medical care was unchanged in December for the third month in a row , with medical care commodities falling 0.8% on drug prices that were 1.0% lower, while medical care services rose 0.3% on 0.4% higher fees for physicians services and hospital services that were 0.5% higher in price....prices for transportation commodities not including fuel were 0.1% lower as prices of new cars and new trucks fell 0.1% and used car and truck prices were 0.2% lower, while prices for transportation services were off 0.4% on vehicle leasing costs that were 1.4% lower than November and air fares that fell 4.7%...in addition, the recreation price index was down 0.3% as recreation commodities fell 0.7% on TV prices that were 1.6% lower and toy prices that fell 0.8% while recreation services fell 0.1% as prices for video and audio services fell 0.3%, gym dues fell 0.2%, and pet services rose 0.6%...lastly, the education and communication index rose 0.2% on the month, as education and communication commodities rose 0.3% on an 0.8% increase in prices for personal computers and peripheral equipment and educational books that were 0.6% higher, while education and communication services rose 0.2% on a 0.4% rise in college tuitions...

the pie graph below from Doug Short shows the weighting of the CPI index components that go into calculating the overall CPI…obviously, the housing price index, at just over 41% of the CPI, dominates this calculation, and includes not only rent equivalents, but also utilities and maintenance, home insurance, and a handful of other housing associated expense categories…the apparel index is the only major category below without a service index component; the food wedge combines indexes for both food at home and away from home, and each of the other wedges are the representative size of the major indexes we described above except transportation, in which the index weighting below includes the prices for motor fuels…notice there’s no energy index; in this take, energy costs are mostly included in the housing and transportation indexes…

CPI-categories via d.short

We Find Real Retail Sales Up 1.4% in the 4th Quarter and Estimate They Add 1.32% to 4th Quarter GDP

next, we’ll take a stab at adjusting retail sales over the entire 4th quarter for inflation over that same time span…we want to do that because we want to estimate the impact of personal consumption expenditures on 4th quarter GDP….it’s important to understand that GDP is not a dollar measure, although it is usually quoted in dollars; it’s a measure of the growth rate of our national product in UNITS of goods and services produced…every dollar denominated component of GDP is adjusted with an inflation gauge in order to arrive at the quarter over quarter change in units of production; in the case of goods sold at retail, the BEA computes a PCE price index for goods to do this, which as we’ve noted is a complex calculation…to simply our calculation, we’ll use the simple one decimal place CPI fractional changes as adjusters, and round number percentages as shown in Robert Oak’s retail sale pie graph above for the percentage of each of the retail components included in the total….imposing any more math than that on figures that are subject to further revision would be a fool’s errand…

we'll start with the largest component of retail sales, which is motor vehicle and parts dealers, which according to the table above, were up 1.2% from the third quarter (July thru Sept) to the 4th quarter (Oct to Dec); we'll adjust that with the price index for "transportation commodities less motor fuel" which includes weighted prices for new and used vehicles and tires and parts; that index was unchanged in October and November and down 0.1% in December...using our simple math, we'll adjust the 1.2% dollar value change in motor vehicle and parts sales with the cumulative 0.1% price decrease to find that real unit sales of motor vehicles and parts were up 1.3% in the 4th quarter...going down the table of the quarter sales changes, we find furniture store sales in dollars were up 2.2%; we adjust that with price changes for "household furnishings and supplies" over three months and find that unit sales of furniture were up 2.4%; next, we have sales at electronics and appliance stores up 0.4% for the quarter; adjusted with the weighted price indexes for appliances and for video and audio products, we find that unit sales at electronics and appliance stores were actually up 1.6%; meanwhile, dollar value sales at building material and garden supply stores were down 2.1% for the quarter; we'll adjust that with the price index for "tools, hardware, outdoor equipment and supplies" for 3 months and find that unit sales of building material and garden supplies were only down 0.5% over that period...next on the table we see sales at food & beverage stores increased 0.9% for the month; since the food at home index was up 0.1% over three months, those unit food sales are reduced to an 0.8% increase...next, sales at health & personal care stores, which are more commonly referred to as drug stores, were up 1.7% for the quarter; adjusting those sales with the price index for medical care commodities indicates that unit sales at drug stores were really up 2.2%...next are sales at gas stations, which were down 0.8% in dollars for the 4th quarter, while the monthly price changes for all types of gasoline were down 2.9% in October, down another 1.6% in November, but up 3.1% in december; with a bit of compounding, we'll say gasoline prices fell 1.5% for the quarter and thus unit sales at gas stations were up 0.7% over the quarter...clothing store sales were up 2.0%, adjusted with the apparel index which was essentially unchanged leaves us real sales of clothing at 2.0%...sales at sporting goods, hobby, book & music stores were up 3.1% in the 4th quarter; we'll adjust them with the recreational commodity price index and find that real sales of sporting goods, books & music were up 3.8%....sales at general merchandising stores were up 0.4% in the quarter, adjusting them with the quarterly change in the price change for retail commodities less food and energy commodities indicates real sales at general merchandising stores rose 0.6% for the quarter, we'll adjust the nominal 0.5% increase in sales at miscellaneous stores with the same index and call their real sales up 0.7%...meanwhile, sales at non-store retailers were up 2.3% for the quarter; we've previously decided to adjust those mostly online sales with a weighted combination of software and book prices, consumer electronics prices, and the apparel price index which would result a 2.9% increase in real online sales ..lastly, sales at bars and restaurants were up 3.0% during the 4th quarter; adjusting that for the half percent rise in prices for food away from home means real restaurant sales rose 2.5%...

now, we have to add all those real price changes together with their proper weighting to get the real price change for all retail sales...as we originally said, instead of using exact percentages for weighting, we're using the whole number percentages shown in Robert Oak’s pie graph; thus, as vehicle sales are 19% of retail, we multiply the 1.3% increase in real motor vehicles and parts sales by that to get their portion of the increase in retail...so, to take a shortcut directly to the math, the change in real retail sales = 19% * 1.3% + 2% *  2.4% + 2% * 1.6% + 6%* -.5% + 13% * 0.8% + 6% * 2.2% + 10% *.7% + 5% * 2%  + 2% * 3.8%  + 13% * 0.6% + 2% * .7% + 9% * 2.9% + 11% * 2.5% equals 1.407% if we've done our math right - someone check that and get back to me...so our real personal consumption of goods appears to have risen 1.4% in the 4th quarter, which would be at a 5.75% annual rate....since  personal consumption of goods is approximately 23% of GDP, we can estimate that real retail sales will add roughly 1.32% to 4th quarter GDP...

Mortgage Delinquencies Rose 2.6% in November as Foreclosure Starts Fell to a Post Crisis Low

another monthly report that was released this week that we want to review is the Mortgage Monitor for November from LPS (pdf) which includes quite thorough and detailed graphics covering the spectrum of information on US mortgages...as per usual, we'll be focusing on mortgage delinquencies and foreclosures, which are the crisis aspects of his report...with this release, LPS (Lender Processing Services), a company involved in mortgage servicing, has split off their Data & Analytics division, which publishes the Mortgage Monitor, into a separate unit that now goes by Black Knight Financial Services...however, as this report still being published from the LPS website, we'll continue to refer to this report as from LPS until such time as the new moniker becomes entrenched..

as of November, LPS reports that 1,256,000 home mortgages, or 2.50% of all mortgages outstanding, remained in the foreclosure process, meaning that a foreclosure notice had been served but the title had not yet been auctioned back to the bank; this is down from 1,276,000, or 2.54% of mortgages, that were in the foreclosure process at the end of October, and down from 1,767,000, or 3.51% of loans that were in the foreclosure process in November a year ago...in addition, 3,241,000​ mortgage loans, or 6.45% of all mortgages, were at least one housepayment delinquent but not in foreclosure in November, up from 3,152,000 homeowners, or 6.28% of those with a mortgage, who were delinquent in October, but that's still down from the 3,583,000, or 7.12% of mortgages, reported delinquent but not in foreclosure last November...of those who were delinquent in November, 1,283,000 properties were seriously delinquent, which means they were 90 or more days delinquent, but not in foreclosure at the end of the month, which when combined with foreclosure indicates that 5.05% of homeowners with a mortgage remained in trouble at the end November...new foreclosure starts fell to 104,939 for November, which is the lowest number of new foreclosures since 2006...

the first graph below, from page 4 of the mortgage monitor pdf, shows the percentage of active loans that have been delinquent monthly since 1995 in red and the percentage that have been in the foreclosure process in green over that same time period…clearly, the percentage of homes in foreclosure has been falling fairly rapidly over the last year and a half and at 2.50% is now down 42% from the October 2011 peak of 4.29% of mortgages in the foreclosure process, which is also known as the foreclosure presale inventory…but that’s still more than four times the pre-crisis foreclosure inventory of 0.44% of December 2005 highlighted on the graph, so we still have a long way to go to return to normalcy…similarly, with delinquent mortgages shown in red at 6.45% of all mortgage outstanding, we’re down significantly from the 10.57% of mortgages that were delinquent but not in foreclosure in January of 2010, but still 51% above the precrisis delinquency percentage of 4.27% of December 2005...

Nov LPS delinquencies and foreclosures monthly

the next graph, from page 25 of the Mortgage Monitor, which is the data summary page, zeroes in on the percentage of mortgage that were delinquent each month over the past year…it’s clear that the percentage of mortgages delinquent has been bouncing up & down over the past 6 months, after falling quickly from 7.17% in February to 6.08% in May, and then jumping back up to 6.68% in June…there is a seasonality to mortgage delinquencies, which you can also see clearly on the graph above, wherein they usually peak at year end, when most people get overextended on their payments during the holidays, and then decline over the first few months of each year as homeowners catch up, but LPS doesn't seasonally adjust their data...we’d expect a higher delinquency rate with some new delinquencies in December, and then expect a decline in total delinquencies in the new year…

Nov LPS summary stats

the next graphic, from page 5 of the pdf, is a map with breakdown of the total percentage of non-current mortgages by state; these percentages include those mortgages that are in the foreclosure process as well as all delinquent mortgages…the darkest red indicates states where the total percentage of delinquent mortgages is above 12%, while the lighter red indicates states where 9% to 12% of mortgages are late on their payments…for the states shaded light green, the delinquency rates range from 5.9% for Nebraska to 8.7% for Washington, while the darkest green states all have total mortgage delinquency rates lower than 5.5%…those low current rates doesn't mean those states haven't had a problem; in the case of Arizona and California, for instance, it just means that those homeowners who fell behind on their mortgages were quickly foreclosed on...

Nov LPS percentage non current state map

after years of Florida being ground zero for the mortgage crisis, you can see Mississippi now has the most non-current mortgages at 15.4% as Florida has also cleared much of its backlog through foreclosure and now sees 14.6% of its homeowners not current on their mortgage payments, which is now the same percentage as New Jersey…but unlike Florida and New Jersey, which still have over 7% of their mortgaged homeowners in foreclosure, Mississippi’s problem seems to be persistent late payments, as only 2.1% of their homeowners have progressed to foreclosure, which you’ll see in the table showing the exact percentages delinquent and in foreclosure that follows, which comes from page 26 of the pdf...

NOv LPS states non-current table

in the above table, the first column shows the delinquency rate (Del%), for each state, which on this table is the percentage of mortgages in each state that are at least one month behind and not yet in foreclosure...the second column is the percentage in each state that are in foreclosure (FC%), while the total percentage of mortgages that aren't current with their payments (NonCurr%) is shown in the 3rd column, which is the sum or the first two....then in the last column, they've included  the year over year change in the total percentage of non-current mortgages...note that those states that have a judicial foreclosure process, where the bank must prove their right to foreclose on a homeowner in court, are marked by a red asterisk...most of the states with the largest foreclosure backlog are judicial, led by Florida at 7.5% and New Jersey at 7.2%...although this mortgage monitor did not include data on the pipeline ratio, as of last month they showed that at the rate foreclosures are being processed in judicial states, it would take 47 months to clear the backlog..

the last graph, below, is from page 6 of the pdf, shows the number of foreclosure starts and foreclosure sales monthly since the beginning of 2005; foreclosure starts are indicated for the month that the loan servicer refers a delinquent mortgage to its attorneys for foreclosure, and are usually initiated with an official notice of delinquency or foreclosure depending on state regulations...for the months after the foreclosure start and before the foreclosure sale, the mortgage is then counted in the foreclosure inventory, or as we've referred to them as "in the foreclosure process"(terms are on page 30 of the pdf)....the definition of a foreclosure sale is less intuitive; it's the legal auction wherein the bank buys the title after the foreclosure completes; after a foreclosure sale, the home moves into the bank's property inventory...you can see that foreclosure starts have been well ahead of foreclosure sales from the duration of the mortgage crisis, which means a lot of foreclosures were started that were never completed...the national average length of time for homes remaining in the foreclosure inventory rose to a record 905 days in November...typically, that means the homeowners have continued to live in their homes without making payments for the duration...for the 1,283,000 seriously delinquent homes, the average length of time they have remained delinquent without moving into foreclosure is now at 500 days; these homeowners have also not been making payments over that period...as noted earlier, the total of these two stranded groups of seriously compromised mortgages is still greater than 5% of all mortgages outstanding nationally...

Nov LPS foreclosure starts and sales

(the above is my weekly commentary that accompanied my sunday morning links emailing, which in turn was mostly selected from my weekly blog post on the global glass onion…if you’d be interested in getting my weekly emailing of selected links that accompanies these commentaries, most coming from the aforementioned GGO posts, contact me…)

Tuesday, January 14, 2014

change in payroll jobs vs change in employed differs by nearly 3 million since July

  while covering the Employment Situation Summary from the BLS on Sunday, we neglected to take an in depth look at the unadjusted employment data from which the widely reported seasonally adjusted numbers are derived; as you should all know, the summary employment report that's released on the first Friday of most months is actually two reports generated from two surveys; the first, the Current Employment Statistics (CES), which is more commonly known as the establishment survey, is a monthly polling of roughly 145,000 businesses and government agencies, or roughly 26% of all US employers, is conducted by the BLS and it gives us a reasonably close estimate (+/- 90,000) of how many jobs they've added each month, how many hours were worked, and how much the workers were paid…the second report, the Current Population Survey (CPS), more commonly known as the household survey, is a telephone poll of roughly 60,000 households typically representing 110,000 working age individuals conducted by the census bureau for the labor department, which despite being quite detailed as to the reasons each individual's employment status, is subject to sampling and other errors that result in an 90% confidence level of +/- 300,000 in the monthly change in the number unemployed, and +/- 0.2% in the unemployment rate....queries to either businesses or individuals for both of these reports focus on employment conditions as of the week that includes the 12th day of the surveyed month, but that's where the similarity ends; the establishment survey only covers those workers who are on a business or government payroll, so it doesn't include the employed farmers or farm workers; it also doesn't include the self employed or individual proprietors, while the household survey includes all of these as employed, and in fact anyone over the age of 16 who is reported as working for pay even one hour during the reference week, ie, even a teenager who's mowing lawns for neighbors…

now, after the employment data is collected by both of those surveys, the BLS runs each of them through a program which compares that month's data to the changes for the same months from each survey over the past several years and adjusts the results for seasonal factors, holidays, and other unusual factors, and ultimately generates the press release and summary using only that adjusted data; the reason this is done is so we can compare one month to another on a similar basis, without unusual factors such as the beginning and end of the school year or holiday hiring that would create large and not very meaningful changes in employment...so there is no mention of the original actual employment numbers or other survey data in the report as released, every statement in the summary fairly much assumes that readers know it's been seasonally adjusted, and only a few of the tables at the end of the summary include unadjusted data....thus, when it's reported by the business press and most economic blogs, none of the actual raw employment data from which the report is generated is even mentioned...

in a similar manner, we ourselves typically report just the adjusted data and only mention the raw, unadjusted data from the two surveys in passing, when it seems that it might be useful for some perspective...it was in so doing that we first noticed a major discrepancy between the number of new payroll jobs indicated by employer responses, and the number of employed as reported by households when covering the August report, which we headlined as a seasonal adjustment discrepancy, because the seasonal adjustment subtracted more than 200,000 jobs from the establishment survey and added nearly 500,000 to the household survey in August...what happened in the August report was that the unadjusted data from the establishment survey indicated payrolls jobs increased by 378,000 from 135,583,000 in July to 135,961,000 (it’s since been slightly revised) while the not seasonally adjusted household data indicated that the count of the employed dropped 604,000, from 145,113,000 in July to 144,509,000, and the seasonal adjustments brought them into approximate alignment....even though we thought this to be an aberration, after consulting with several economists, we felt confident that the August misalignment would subsequently reverse itself in the next month or two, and the two surveys would be brought back into sync...however, the September report compounded the discrepancy, as it again showed a payroll job gain of of 612,000, which was lowered by the seasonal adjustment to 148,000, while the unadjusted household survey count of 142,000 employed was little changed by the seasonal adjustment...the October report was even more alarming, in that there was a difference of 1,162,000 jobs in the seasonal adjustments between the the two surveys; the establishment survey seasonal adjustment subtracted 682,000 payroll jobs, while the household survey seasonal adjustment added 480,000 to the count of employed...thus, after three months we found that the unadjusted payroll job count went up by 1,963,000, while the unadjusted count of the employed went down by 967,000, only to have them brought into approximate alignment by the seasonal adjustments...still, because that data was affected by the shutdown, we expected this to subsequently correct, but as of the November and December reports it has not to any degree...so this post is to update where the two unadjusted surveys stand as of the end of December, taking the minor revisions of previous months since into account…we’ll start by looking at the recent data as it’s available from FRED:

first, we have the unadjusted non farm payrolls as reported by employers for each month from July through December (000's):
2013-07    135577
2013-08    136002
2013-09    136612
2013-10    137523
2013-11    137999
2013-12    137753

then, we have the raw unadjusted count of employed extrapolated from those who reported they were employed in the household surveys over the same time frame (000's):
2013-07    145113
2013-08    144509
2013-09    144651
2013-10    144144
2013-11    144775
2013-12    144423

it’s clear from the above that from July to December, the unadjusted count of the employed from the household survey fell by 690,000, from 145,113,000 in July, to 144,423,000 in December...over the same time frame, the unadjusted non-farm payrolls rose by 2,176,000, from 135,577,000 in July to 137,753,000 in December...after seasonal adjustments, the household survey was changed to show an increase of 301,000 employed over those same months, while the increase in non-farm payrolls was cut to 928,000... for a visualization of what has been happening, we are including below a FRED graph which shows the historical track of the jobs or employed count from two surveys, before seasonal adjustments, since the beginning of the last decade…in blue, we have the unadjusted count of those self-reporting as employed from the household survey, with the count in thousands of employed on the graph’s right margin; in red, we have the unadjusted count of payroll jobs as reported by employers taking part in the establishment survey, with that count on the left margin, such that the two tracks tend to overlap…there’s obviously a lot of noise in the unadjusted data series; what we want to focus on is the last five monthly changes, where we see the payroll jobs in red moving almost straight up until December, while the count of the employed in blue has been trending downward; notice the red line is relatively well above the blue one for the first time in years...ominously, the two other times that there was such a large increase in payroll jobs that was not accompanied by an equal increase in the number of employed were in 2008, and 2000-01, just preceding the recessions which are shown as grey bars on this graph…we are not suggesting that this predicts a recession, however; we're just noting that it’s inconceivable for these two lines, which are in effect measurements of the same function of employment, to continue moving in different directions indefinitely…

FRED Graph

Sunday, January 12, 2014

December employment summary and November international trade

there isn't much good to be said about the two December employment reports released by the BLS on Friday....against a median forecast of 200,000 new jobs, the establishment survey from Friday's employment summary indicated just 74,000 jobs were added in December, the slowest payroll jobs growth in 3 years, while the weekly hours for those working decreased...and the headline result from the household survey showed that the unemployment rate dropped to 6.7% from 7.0% in November, almost entirely because over a half a million of us gave up looking for work and dropped out of the labor force, and hence weren't counted when the jobless percentage was calculated...

Establishment Survey Shows December Added Least Lousy Jobs in 3 years

FRED Graph

with the addition of just 74,000 jobs in December, and the revision of the change in November job additions from 203,000 to 241,000 (October's job count was unrevised at 200,000), the total job creation for 2013 is now in the books at 2,186,000, which means, unless there are further revisions, the total job creation this year was actually less than the anemic 2,193,000 jobs created in 2012; although that was less than the 2,395,000 increase in the working age population, roughly one-third of that population increase represents retirees, students, and housewives or those who wouldn't consider working otherwise, so that means on net our economy only generated 586,000 more jobs in 2013 than needed to keep up with the increase in the labor force, or less than 49,000 a month...at that pace, it would take another 14 years to make up the 1.3 million jobs lost and the 6.6 million jobs needed that were not created since 2008 to return to the level of labor force participation we saw in 2007...our adjacent FRED bar graph shows the monthly number of payroll jobs gained or lost since the beginning of 2008; clearly, it takes a better pace than we’ve seen just to recover from the 12 month period in 2008-09 when monthly job losses approached the 400,000 to 800,000 level…

the types of jobs created in December were mostly of the low paying variety we've been seeing most of this year as well...on a seasonally adjusted basis, there were 55,300 additional jobs created in the retail sector; of those, 11,800 were in food and beverage stores, 11,600 were in clothing and accessories stores, 7,600 were in general merchandise stores, and 7,300 were with motor vehicle and parts dealers; these are retail jobs in excess of what would normally be added in December; the unadjusted data indicates there were 176,500 additional retail jobs over the month...an additional net of 19,000 jobs were generated in professional and business services, with a gain of 40,400 temporary help service jobs offset by a loss of 24,700 accounting and bookkeeping jobs...15,400 more jobs were created in wholesale trade area, with most of those in electronic markets, working with or as agents and brokers, while the manufacturing sector added 9,000 jobs, 5,000 of which were involved in fabricating metal products, while manufacturers of transportation equipment added 3,800 and electronic instruments makers cut 3,500 ...(note that there may be an undercount of jobs in manufacturing in BLS data, as many manufacturers are now hiring agency temps as virtual full time worker as a way to avoid union scale wages and benefits that would be due full-time employees; BLS counts these as temporary help service jobs)...the leisure and hospitality added 9,000 jobs, as 14,400 were added in accommodation and food services and 11,600 were lost in performing arts and spectator sports....an additional 5,100 jobs were created in mining, with 3,500 of those in mining support activities...

areas where payroll employment declined in December included construction, where there were a seasonally adjusted 16,000 less jobs, largely due to the loss of 12,900 working for nonresidential specialty trade contractors; a number of analysts are blaming this on the weather; in fact, some call the whole report into question due to the cold & snowy weather...but this is December, and the data is seasonally adjusted, so the weather affected this report only insofar as it was that much worse than a normal December; the unadjusted data shows 216,000 less construction jobs, so in effect, a reduction of 200K in the construction labor force in December is considered normal...for perspective, consider that a seasonally adjusted 146,000 new jobs were created in the month that hurricane Sandy impacted the jobs report...December also saw a shrinkage of 13,000 government jobs, with a loss of 14,900 in local school districts offsetting gains of 5,000 other jobs at the local level....there was also a lost of 12,000 jobs in information sector, with cuts of 13,700 in the motion picture and sound recording industries overwhelming small gains elsewhere in the sector...and there was a shrinkage of 6,000 jobs in health care, the first time that sector has seen job losses in over ten years; there were 4,100 less jobs in  ambulatory health care services, 2,400 less in hospitals, while 5,000 more were working in social assistance...payroll employment in the remaining sectors, finance, transportation and warehousing, and private education, was little changed for the month...

our usual FRED bar graph below shows the monthly change in payroll employment in selected sectors since the beginning of 2012, with the scale on the left indicating the increase of payroll jobs in thousands when the bar extends upwards, and the decrease in jobs when it points downwards..in each monthly grouping of 8 bars, the monthly change in manufacturing employment for that month is indicated in blue, the change in monthly construction employment is in red, the monthly change in retail employment is in dark green, the monthly change in government jobs is in yellow, and the change in employment in professional and business services, which includes everything from legal services to hauling out the trash, is in grey; also included are the CES employment subcategories of jobs in bars and restaurants in light green, and new health care jobs in orange, with private education jobs shown in violet...December is represented by the bar cluster on the far right; the big jump in retail jobs (dark green), and the loss of construction (red), government (yellow) and health care (orange) jobs is fairly obvious, as is the change from November when all eight categories gained jobs…(click to enlarge)

FRED Graph
as we mentioned in opening, the report from employers also indicated that the average workweek for all payroll employees fell from 34.5 hours to 34.4 hours; in addition, the average workweek for production and nonsupervisory employees fell to 33.6 hours from 33.7 hours...the largest hour cutbacks were in mining and construction, which could be weather related, while hours for utility workers rose 41.1 to 41.7 hours...the manufacturing workweek was unchanged, at 41.0 hours, and factory overtime was up by a tenth of a hour to 3.5 hours...the average hourly pay for all workers was up 2 cents to $24.17, with utilities workers, whose hours rose from 41.6 a week to 42.0, apparently seeing an 33 cent an hour gain to $35.55 an hour, the highest average pay for any sector, while the lowest paid workers in leisure and hospitality gained 8 cents and hour to average $13.66 an hour...meanwhile, average pay for nonsupervisory workers was up 3 cents to $20.35, with utility linemen again making the most at $32.77 an hour, with pay for nonsupervisory leisure and hospitality workers also up by 10 cents to $11.93 an hour...

Labor Force Participation Rate at 36 Year Low Drives Official Unemployment Rate Down to 6.7%

the bad news from the Current Population Survey (CPS), more commonly referred to as the household survey, was that the labor force participation rate fell to a 36 year low at 62.8% in December, fractionally eclipsing the new low that was set just two months ago in October, when everyone thought that new low was a one time aberration caused by the government shutdown during the reference week...this 'participation rate' is the percentage of the population either working or looking for work, and it's continual descent is indicative of many who would be willing to work if they thought they could find a job just giving up even looking...at 0.8% lower than a year ago, this represents nearly 3 million of us who've given up entirely over the past year alone; even the number of those not in the labor force who will still admit they want a job if they could get one has fallen by more than 600,000 over that span, as many of those who are discouraged start to tell themselves that not participating is where they want to be...we can imagine that the longer these long term unemployed are out of the labor force, the more likely they'll slip from the radar entirely...we’ve stretched out our FRED graph below to take in 1978, the last time the labor force participation rate, which is shown in red, was as low as today…for those of you too young to remember, that was an era when not many women worked outside of their own homes…and on the same graph we show the employment to population ratio in blue over the same time span; in December this ratio, which could be thought of as the employment rate, was stuck at 58.6%, the same as it was in November, and the same as it was a year ago, and the same as it was over four years ago…

FRED Graph

as we look at the data from this survey, we should recall that it's extrapolated from a telephone poll of roughly 60,000 households representing around 110,000 workers, so with that small sampling there's a margin of error of +/- 300,000 in the monthly change in the number unemployed, and +/- 0.2% in the unemployment rate...in December, the change of both the number of employed and the unemployment rate was greater than that margin of error, so at least we can be sure of the direction of the change in both...according to the seasonally adjusted employment data from the household survey, 144,586,000 of us reported being employed in December, which was 143,000 more than November; another 10,351,000 of us reported being unemployed, which was a decrease of 490,000 from the previous month; thus the number of us counted as being in the labor force fell by 347,000 to 154,937,000, and the percentage of the labor force that was unemployed fell to 6.7%...since the working age population increased by 178,000 over the month, that meant the total of those who were not in the labor force and hence not counted as employed nor unemployed rose by 525,000 to a new all time high of 91,808,000...

the number of us working part time declined by 89,000 to 27,372,000 in December; however, the number of us who were working part time who indicated they wanted to be working full time but who could only find part time work, or who had their hours cut, increased by 49,000 to 7,771,000...even with that increase, however, the alternate measure of unemployment known as U-6 remained unchanged at 13.1% in the face of a smaller labor force divisor...the number of us unemployed for more than 27 weeks and who are thus now cut off from further unemployment rations fell by 166,000 to 3,878,000; as a percentage of all unemployed, however, the long term jobless are now 37.7%, up from 37.4% in November...recall that some of the numerical reduction represents those among the long term unemployed who didnt look for a job in December and who hence were not counted....among those not officially in the labor force and hence not counted, an additional 5,932,000 reported that they still want a job; of those, 2,427,000 were categorized as "marginally attached to the labor force" because they've  looked for work sometime during the last year, but not during the 30 day period covered by the December survey...917,000 of those were further characterized as "discouraged workers", because they say that they haven't looked for work because they believe there are no jobs available to them...

Trade Deficit Smallest in Four Years on Declining Oil Imports

the other important release of the past week was on on our International Trade in Goods and Services for November (full pdf), which showed our seasonally adjusted trade deficit was the smallest in more than four years, mostly due to lower oil imports...the Department of Commerce reported that record November exports of $194.9 billion, $1.7 billion more than revised October exports of $193.1 billion, and imports of $229.1 billion, $3.4 billion less than revised October imports of $232.5 billion, resulted in a goods and services deficit of $34.3 billion, down from the revised trade deficit of $39.3 billion in October...seasonally adjusted October exports were revised up by $0.4 billion and imports of goods were revised down by $1.1 billion, and thus October's deficit was down $3.6 billion from the near $43.0 billion trade deficit of September....our FRED bar graph below shows the monthly change in exports in blue and the monthly change in imports in red since November 2011, with the net of them resulting in the change in the balance of trade, which is shown in brown; each group of three bars represents one month’s of trade data over the past two years, with positive changes above the ‘0’ line and negative changes below it; when exports (blue) increase in a given month, they add to the trade balance change in brown; conversely, when exports decrease, they subtract from the brown trade balance bar; the action of imports on the balance is just the reverse; when imports increase in a given month, they subtract from the brown trade balance for the month, but when imports decrease, the balance of trade rises as a result….with two months of significantly better trade data in the books, we'd guestimate that the change in trade could add more half a percent to 4th quarter GDP, before any adjustment for inflation is applied....

FRED Graph

our November deficit in goods shrunk by $4.9 billion from October to $53.9 billion as our exports of goods increased $1.5 billion to $137.1 billion, and our imports of goods decreased $3.4 billion to $191.0 billion, while our trade surplus in services increased $0.2 billion to $19.7 billion as our imports of services increased $0.1 billion to $38.1 billion and our exports of services services increased $0.3 billion to $57.8 billion....as you see from the FRED graph, monthly changes are quite volatile; year to date, our trade deficit has been $435.1 billion on exports of $2,079.1 billion and imports of $2,514.2 billion, in contrast to the trade deficit of $496.3 billion on $2,021.9 billion in exports and $2,518.2 billion in imports over the first 11 months of 2012, so generally, our exports have been increasing over time while our imports have remained stable....

the end use categories of imports that saw increases in November included imports of automotive vehicles, parts, and engines, which increased by $1,059 million to $27,229 million, and imports of capital goods, which were up $925 million over October...the capital goods contributing most to that increase were imports of computers, which increased by $686 million to $6,033 million in November, and imports of computer accessories, which increased by $472 million to $5,203 million, which were partially offset by decreases of $361 million in imports of telecommunications equipment and $237 million of semiconductors...the import category that accounted for the lions share of the import decreases in November was industrial supplies and materials, which fell $4,297 million to $53,804 million on a $2,547 million decrease in imports of crude oil, a $826 million decrease in imports of fuel oil, a $221 million decrease in imports of other petroleum products and a $270 million decrease in imports of non-monetary gold...our imports of foods, feeds, and beverages also fell by $288 million to $9,538 million on a decrease of $158 million in imports of fruits and fruit products and a $115 million decrease in imports of fish and shellfish....our imports of consumer good fell by $119 million to $44,752 million on a $565 million decrease in imports of pharmaceuticals and and a $426 million decrease in imports of artwork, antiques and collectibles, which was mostly offset by increases of $740 million in cell phone imports and $284 million in imports of apparel other than cotton and wool, and our imports of other goods not categorized by end use fell $764 million to $5688 million...

end use categories that saw exports increase in November included industrial supplies and materials, which were up by $707 million to $43,815 million on increases of $264 million in both exports of chemicals and finished metal shapes, a $263 increase in exports of crude oil, and a $240 increase in exports of fuel oil, while exports of non-monetary gold fell $796 million; also capital goods, in which our exports increased $336 to $45,350 million on a $390 million increase in exports of civilian aircraft and a $163 million increase in exports of engines for such aircraft, which was partially offset by a $210 million decrease in exports of semiconductors...November also saw a $141 million increase in exports of autos, trucks engines and parts to $13,091 million and a  $483 million increase to $5,552 million in exports of other goods not otherwise categorized... decreasing in November were exports of consumer goods, which fell $515 million to $15,943 million on $492 million less exports of artwork, antiques and collectibles, $172 million less diamonds, and $158 million less exports of jewelry, partially offset by a $191 million increase in exports of pharmaceuticals; also lower by $124 million were our exports of foods, feeds and beverages despite a $212 increase in soybean exports as wheat exports fell $190 million and fish, corn and rice exports fell by lesser amounts...

the graph below, from Bill McBride, is illustrative of the impact that our reduction of oil imports has had on our trade overall; reading from the top $0 line down, the black graph line tracks our deficit in petroleum trade only in billions of dollars since 1998; over the same span, the red graph shows our trade deficit for everything else except oil; combined together, those two are of course our total trade deficit, which Bill has graphed in blue...it's pretty clear that as our oil deficit has fallen, so has our overall deficit, and by virtually the same amount...meanwhile, our deficit in everything else has tracked at roughly the same magnitude, near $20 billion, over the entire four years of our recovery…

McBride TradeDeficitNov2013 with oil

it wasn't only trade in petroleum that improved in November, however...our bilateral goods trade deficit with China, which had reached a record $30.5 billion in September, fell to $26.9 billion, from $28.9 billion in October...nonetheless, that was still nearly half of our overall $53.9 billion deficit in goods for the month...our bilateral goods deficit with the European Union also shrunk by $4.2 billion to $10.1 billion in November as our imports decreased and our exports to the trade block increased...other large bilateral goods deficits in November were with Germany at $5.9 billion, Japan at $5.8 billion, OPEC at $4.8 billion, Mexico at $4.1 billion, Saudi Arabia at $2.9 billion, Ireland at $1.8 billion, Venezuela at $1.5 billion, Canada at $1.5 billion, Korea at $1.2 billion, and India at $1.0 billion...meanwhile, we ran small surpluses in November goods trade with Hong Kong at $2.9 billion, Australia at $1.2 billion, Singapore at $1.2 billion and Brazil at $1.7 billion...

the Bloomberg graphic below was taken from the Zero Hedge post on November’s trade, where it was presented without comment or a link; on the left, we see the top ten countries we import from, and within each country’s block the percentage of our total imports that come from that country, the change in imports from that country on a year over year basis, with green indicating an increase and red a decrease in imports from that trade partner, and apparently on the top line is the change in our imports from that country on a GDP basis as of the last report, which would be adjusted for inflation using chained 2009 dollars…similarly, on the right are the top ten countries we export to, the percentage of our exports that goes to each, and the year over year change in our exports to each on both a current dollars and inflation adjusted dollars basis…meanwhile, the center block shows our goods imports on a current dollar basis over the past year at $2,294.2 billion, down 0.4%, and our goods exports at $1,585 billion, up 1.9%, giving us a negative trade balance in goods over the past year of $706 billion..

Trade November 2013 BLM via ZH

(the above is my weekly commentary that accompanied my sunday morning links emailing, which in turn was mostly selected from my weekly blog post on the global glass onion…if you’d be interested in getting my weekly emailing of selected links that accompanies these commentaries, most coming from the aforementioned GGO posts, contact me…)

Sunday, January 5, 2014

November construction spending, October Case-Shiller home prices, December manufacturing PMI

it's been a fairly slow week for economic news releases...the December unemployment report, which would have typically been released on the first Friday of January, was rescheduled to be released on the 10th, and the GDP revision, usually at the end of the month, was released on the Friday before the holidays...one monthly report we normally dont cover that we could take a look at is Construction Spending for November from the Census Bureau, which goes into more important detail than the popular monthly report on new housing starts...since this report includes monthly data on not only private residential construction, but also private non-residential construction as well as construction spending for public use by governments, it's an important input into the investment component of GDP...and since we did not cover the October release of this report, we'll look at those numbers too, with an eye to gleaning some insight on how the quarterly data to date might effect 4th quarter GDP...

November Construction Spending Near Five Year High

as with all census reports on construction, we have to caution that this preliminary data still has a significant margin of error and is subject to revision; the report's opening lines put this in perspective, noting that seasonally adjusted construction spending during November was estimated at a annual rate of $934.4 billion, 1.0 percent (±1.6%)* above the revised October estimate of $925.1 billion...as you should recall, an asterisk in Census data leads to the footnote warning than since the 90% confidence range contains zero, it's uncertain whether construction spending increased or decreased month over month, that it could have risen by as much as 2.6% or fallen by 0.6%...they further explain that it may take 2 months to establish an underlying trend for total construction, and as long as 8 months for specific categories of construction; to that point we'd note that revised October data showed construction also rose by 1.0%, from $916.5 billion in September to $925.1 billion, and the revised September rise was even more, at 1.4%, so we can assume the increasing trend has been established, and if November’s numbers hold, it will be the highest construction spending since March 2009...the unadjusted data from which these annualized figures are extrapolated from indicates that November construction spending was at $77,965 million, down from $85,101 million in October, and down from $85,383 million in September, as we all know construction slows heading into winter; the seasonal adjustments make up for that normal downturn by comparing this year to the historical change...the headline data, and what we'll be looking at, are seasonally adjusted and given at an annual rate, ie, what construction spending would be if that month's change were extrapolated over an entire year....year over year figures show that  November construction was estimated at 5.9 percent (±2.0%) above last November's annualized $882.7 billion, and year to date construction spending amounted to $828.4 billion, 5.0 percent (±1.3%) above the $788.8 billion for the first 11 months of 2012...none of these figures are adjusted for inflation, although costs for labor and material price changes have been subdued, so there’s probably not much inflation to adjust for….

breaking down recent months further, seasonally adjusted private construction rose at a 2.2% month over month rate in November, which when expressed as an annual rate would be at $659.4 billion, up from the revised October estimate of $644.9 billion; however, from September to October, the increase in private construction was statistically unchanged...of the November increase in private construction, $345.5 billion of the annualized figure was residential construction, 1.9% up from the revised October estimate of $339.2 billion, and $313.9 billion at an annual rate was non-residential private construction, which was up 2.7% from October's revised estimate of $305.7 billion..revised figures show residential construction for October fell 0.4%, while non-residential construction spending was 0.5% above September... types of private non-residential construction spending seeing the largest seasonally adjusted monthly increases included commercial space, where construction spending rose 4.7% to an annualized $52,391 million, after rising 5.8% in October and which is now 20.7% ahead of the level of a year ago, private transportation construction, which also rose 4.7% to $13,492 million and is now 18.3% ahead of last year's pace, office buildings, which saw a 4.6% increase to $32,493 million in November after a 7.1% increase in October and which was 11.5% above the November 2012 level, and construction of communication infrastructure, which rose 11.2% to $ million but which is nonetheless still 10,5% off last year's pace...only construction of health care facilities, which slipped 1.2% to $30,149 million, and private education construction, which fell 3.2% to %16,900 million, saw less construction in November than October, but both of those as well as every other private construction type were at higher levels in November than at the end of the 3rd quarter...

however, the estimate of seasonally adjusted public construction, which is construction spending by federal state and local governments, fell by 1.8%  (±2.5%) to an annualized $275.0 billion in November after rising 3.1% to an estimated $280.2 billion in October; although this is still positive for the 4th quarter, public construction still remains 0.2% below its year ago level, and in real terms is at the same level as 2001....two major types of public construction saw increases in November; educational, which rose 1.1% to an annualized $65,176 million after increasing 7.1% in October, and spending on public power utilities, which rose 1.6% to $13,279 million after rising 3.6% in October..neither are up on a year over year basis, however: public construction for education is statistically unchanged since last November, and public power construction is still 2.2% below its level of a year earlier...

most other types of public construction spending were lower in November...public outlays for sewage and waste disposal facilities shrunk 8.1% to $20,482 million after falling 1.0% in October...construction of public health care facilities was off 7.3% to $10,158 million after rising 1.2% in October, while construction of government office space fell 4.6% to $7,985 million; outlays for public conservation were off 4.9% to $6,204 million, spending for construction of water supply infrastructure fell 4.8% to $12,720, spending for buildings used by public safety forces fell 1.2% to $9,590 million, public transportation construction spending fell 2.4% to $29,948 million, and road construction was down 0.4% to $82,049..

our first FRED graph below shows the track of each of the 3 major categories of construction spending since the beginning of 2005, with millions of dollars spent for each shown on the left margin; blue shows the annualized value all residential construction monthly, including apartment buildings, and clearly shows the housing boom and subsequent bust; red shows the track of all private nonresidential construction spending, which as we’ve seen is mostly for commercial and industrial development, over the same time frame, and also shows that this type of construction spending, with long lead times, was actually highest during the recession, marked by the grey vertical bar, and collapsed thereafter as projects already underway wrapped up; finally, the green track is for public construction spending, which also shows a similar pattern, as larger projects and those initiated as the result of the 2009 Federal stimulus kept public spending high throughout 2010, only to taper off thereafter, as lower tax receipts constrained much of such spending at the state and local level…

FRED Graph

our second FRED graph for this data focuses on the recent changes in public construction spending, as this sector has been a drag on GDP throughout the recovery; each group of six bars indicates changes in millions of dollars for each type of construction for each month since the beginning of 2012, with blue indicating the monthly change in highway and road construction spending, red the monthly change in outlays for school buildings, green the change in spending for public water supply, orange the monthly change in spending for public transportation, violet the change in public spending for sewage and waste disposal facilities, and grey bars indicating the monthly change in construction spending for public power utilities…clearly, the series is quite volatile, although the contraction seen earlier in the recession seems to have abated… (click to enlarge)

FRED Graph 

Case-Shiller Says Homes “Return” 13.6%  Year over Year in October

also released this week, on Tuesday, was the S&P Case Shiller home price index for October, which compares a 3 month unweighted average of prices received for homes sold in August, September and October to prices received for the same houses the last time they were sold, and generates an monthly index such that prices for homes in 2000 = 100...for this October release, both the original ten city index and the expanded 20 city index increased 0.2% from the September report, with the 10 city index at 180.27 and the 20 city index at 165.91...both Composite indexes also saw a year over year increase of 13.6%, which was the greatest one year jump in prices since the height of the bubble in 2006 and the 17th consecutive month that the two indexes have posted a year over year increase...our FRED graph below shows the track of both indexes since their origination; the Composite Index for the ten original metro areas (Boston, Chicago, Denver, Las Vegas, Greater Los Angeles, Miami, New York, San Diego, San Francisco, and Washington DC) is shown in red, and Composite 20, which includes the metro areas of the ten city index and added prices for Phoenix, Tampa Bay, Atlanta, Detroit, Minneapolis-Saint Paul, Charlotte, Cleveland, Portland, Dallas, and Seattle metro areas to the original 10 in 2000...note that although we and others call these city indexes, they are actually tracking prices for the entire Metropolitan Statistical Areas as defined by the Census bureau...also note that as these are not seasonally adjusted, we can see the obvious seasonal variation as home prices typically rise in spring and fall after the school year begins...

FRED Graph

according to S&P, ten metropolitan areas "posted positive monthly returns" in October, nine saw home prices fall, and home prices in New York City were statistically unchanged...Las Vegas showed the largest increase as home sales prices rose 1.2%, followed by Miami with a 1.1% monthly increase, and Phoenix and Los Angeles, where home prices rose 0.9%...Chicago home prices fell by 0.5%, and both Denver and Washington DC prices fell 0.4%, leading those cities where prices declined, which also included Atlanta, Boston, Cleveland, Dallas, San Francisco, and Seattle...only Denver saw net prices decline over the past two months, while Cleveland home prices were essentially unchanged since August...

all 20 metro areas saw prices rise year over year, with Las Vegas prices up 27.1% topping the chart, and with San Francisco 24.6% higher and Los Angeles up 22.1% also posting price increases of over 20%...San Diego saw home prices rise 19.7%, Atlanta prices were up 19.0%, Phoenix prices rose 18.1%, and Detroit prices were 17.3% higher than October 2012...on the other end of the spectrum, home prices in Cleveland and New York were only up 4.9% over one year, Washington DC area saw prices rise 7.4%, Boston prices were up 8.6%, Charlotte prices rose 8.8%, Denver prices were up 9.5% and Dallas prices were 9.7% higher than a year earlier...the bar graph below, from Robert Oak's coverage of this month's report, shows the Case-Shiller index as of October for each of the 20 cities in the Composite 20; recall that all Case Shiller indexes were set to 100 in the year 2000, so these indexes in effect measure the price change since then, such that Phoenix, the first bar on the graph with an index level of 144.49, means that the average home in that city has seen 44.49% price appreciation since then…note that these indices are not adjusted for inflation; however, we coincidentally benchmarked our own inflation gauges of CPI components to the year 2000, the same year as these indexes are benchmarked to, so we can see that the housing component of the CPI, shown in red, is up 38% since then

October 13 Case Shiller Indexs

it's important to understand that these indexes are not prices, nor are they relative to each other...again, each city index was set to 100 in 2000, and thus so were the composite averages of 10 or 20 cities....but homes in a given city with an index number of 150 just implies that home prices have risen 50% in that city over the last 13 years; as it would imply for another city with the same index value...but home prices could still be half or twice as much in a city with a 150 index value as another city with the same index reading...to illustrate that, we can look at some of the median home prices for cities from the Zillow Home Value Index, which is calculated from the online Zillow real estate database, where we find that the median price for a home in New York City was $354,100 in October, while the median price for a home in Miami was $175,000, even though they both have an Case Shiller index value of just over 173...all the index value for both means is that average prices in both cities are 173% of what the prices were in 2000; in New York's case, that would be something around $204,600, while prices in Miami were likely near $100,900 13 years ago... 

ISM Manufacturing PMI at 57.0% in December, Indicating Continued Expansion

the other widely followed index that was released this week was the December Manufacturing PMI (Purchasing Managers Index) from the Institute for Supply Management (ISM); which was down 0.3% to 57.0% but still at a level that indicated a decent level of expansion in the manufacturing sector...the manufacturing PMI is a composite index based on equal weights of the new orders, production, employment,  and inventories indexes, which themselves are diffusion indexes calculated from responses to questionnaires sent to purchasing executives on the current conditions in their industry; wherein each response of "better" out of a each hundred queries adds one to the index; each response of "the same" adds a half point, and responses of "worse" are not counted; no weighting is given for the different sizes of businesses polled, or whether conditions are a whole lot better, just marginally so, or a whole lot worse...thus any reading over 50 simply means more purchasing managers polled indicated that sector of business was expanding than those who said it was contracting, and vice versa for readings under 50...

of the five indexes that are composites of the PMI, the new orders index was at 64.2%, up from 63.6% in November and the highest since April 2010, indicating ongoing and significant increases in new orders; the production index was at 62.2%, down 0.6% from November's 62.8%, but still indicating brisk production increases, while the employment index was at 56.9%, up from 56.5%, indicating a pickup in hiring; the suppliers deliveries index was at 54.7%, up 1.5% from November, meaning that deliveries were even slower, presumably indicating suppliers were busier and hence business was better, and the inventories index, which slipped significantly into contraction at 47.0% from a barely expansionary reading of 50.5% in November...

other indexes included in this report that are not part of the composite PMI included the exports index, which fell 4.5% to 55.0%,  indicating export business what growing somewhat slower, the imports index, which was unchanged at an expansionary 55.9%, the prices index, which increased 1.0% to 53.5%, a positive as higher prices are indicative of stronger demand, the backlog of orders index, which declined 2.5% from 54.0% to 51.5%, which nonetheless still indicates growth, and customers inventories, which increased from 45.0% in November to 47.5% in December but were still too low... 

of 18 manufacturing industries whose purchasing managers contribute to this survey, 13 indicated expansion; the ISM does not give a numerical value to these, but simply lists them in order of fastest growth; leading the list of fastest growing manufacturers in December was furniture & related industries; plastics & rubber products; textile mills; and apparel and leather goods, while the four industries reporting contraction in December were nonmetallic mineral products, machinery; chemical products; and electrical equipment, appliances & components...

the ISM claims their survey is consistent with the results of the Full Report on Manufacturers’ Shipments, Inventories, and Orders from the Census Bureau, which we've occasionally reported on as the advance report on Durable Goods or as the full report on Factory Orders, and since they often report well ahead of the Census, many analysts believe that this release gives them an advance insight into the hard data that wont be released for another month...on our FRED graph below, we've included the track of the ISM manufacturing new orders index in blue since the beginning of 2007, and the percentage change in the value of new orders for all manufacturing monthly in red over same span….theoretically, the new orders from the PMI should lead the factory orders change; we’ll let the viewers decide if it appears that the blue purchasing manager’s new orders index has predicted the actual direction of real new orders in red over time or not…

FRED Graph

(the above is my weekly commentary that accompanied my sunday morning links emailing, which in turn was mostly selected from my weekly blog post on the global glass onion…if you’d be interested in getting my weekly emailing of selected links that accompanies these commentaries, most coming from the aforementioned GGO posts, contact me…)