Introduction

Balance-sheet unwinding was the week's big but completely expected announcement, one really guaranteed by the auspicious gain in core consumer prices during the prior week. Yet what moved the bond market in the latest week wasn't the Federal Reserve's balance sheet but the rising likelihood of a pending rate increase. The week's economic data may have been mixed and did take a backseat to the Fed, but a close look probably justifies the rising odds for a December hike. We start, however, where strength is the least but we'll end where it is the strongest.

The economy

Existing home sales slowed to a 5.350 million annualized rate in August for the 3rd straight dip and 4th dip in 5 months. The trend is strictly flat, up only 0.2 percent over the last year. Harvey did hurt sales as the South saw a sharp 5.7 percent August drop, but the weakness is fundamental: the result of high prices, which are running at a 6 percent annual pace, and also low supply both of which limit buyer choices. New home sales have been doing better in part because of greater supply in their market.

Supply of existing homes relative to sales held at 4.2 months in August which is where this reading has been stuck for the last 4 months. New home supply has been doing better, at 5.8 months in the latest data which are for July. In numbers, resale supply is at 1.880 million for a 6.5 percent annual dip. In contrast, new homes on the market were up a yearly 16.5 precent in July, to 276,000. But both new home sales and also supply, to be posted in the coming week, could see August hurricane effects.

Higher new home supply reflects what has been a good year for permits and completions, at least for permits which, at an 800,000 annual rate in August, were up  a yearly 7.7 percent. Completions also had been doing well, at a yearly 11.6 percent rate in July before, however, Harvey hit Houston. Pulled down by a 7.9 percent August drop in the South, completions fell 13.3 percent to a 724,000 rate for a yearly decline of 2.7 percent. Irma will also be a setback for new home supply beginning with September's data.

Housing has been uneven all year as has the factory sector. But you wouldn't know it based on anecdotal reports which have been surging. One that isn't is Markit's PMI which has been no better than moderate, coming in at 53.0 at mid-month September. In contrast, the ISM index for August rose to a 6-year high of 58.8. What are the government's actual numbers? Factory orders are up a yearly 4.9 percent while manufacturing output is up only 1.5 percent. When it comes to anecdotal data, readers beware!

There are fewer measurements of services than either housing or manufacturing, a reflection of the sector's abstract quality compared to structures and goods. This adds special interest to the Markit and ISM reports on services which have been showing solid strength perhaps a bit in excess of GDP. Here, strength in Markit's report, at 55.1 in mid-month September, has been moving ahead of ISM which, however, was unusually strong earlier in the year. Neither report has shown any substantial hurricane effects.

Together, housing, manufacturing and services point to moderate growth for the economy as a whole. But growth isn't the swing factor right now for Fed policy, inflation is. The week's data include import prices where a Harvey effect on energy helped make for strength. Less apparent is the effect of dollar weakness which raises the cost of imports. As of August, the trade-weighted dollar was down a yearly 2.4 percent, giving a boost to the price of consumer imports which moved over zero to 0.4 percent.

Strength through the breadth of import prices was impressive in August. Consumer goods were not the only ones moving into positive ground, joined by capital goods at a yearly plus 0.3 percent and vehicles at 0.1 percent. This may not look striking but the trend is favorable. All these goods are finished goods where movement may be glacial but where gains, unlike monthly ups and downs in commodities, are decisive for final selling prices. The prior week's strength in the core CPI was likely no fluke.

Speaking of flukes brings us to jobless claims where forecasters are getting rattled. Claims rose far more than expected immediately following Harvey's Texas hit, then fell back unexpectedly in the 2 following weeks. Claims in Florida did double in the latest week to nearly 10,000 with claims in Puerto Rico (and this is before Hurricane Maria) surging to 2,500. But 2 other hurricane states, South Carolina and Georgia, have shown no effect. What all this volatility means for September payrolls is anyone's guess.

Markets: Rates begin backing up

The Fed's decision to begin trimming down its $4.5 trillion balance sheet will mean one less buyer, and a big one at that, in the bond market which in turn looks to result in rising rates. The hope among policy makers is that the rise will be gradual, orderly and limited. So far so good though the 2-year yield did gain a sizable 6 basis points in the week to 1.44 percent. And looking over the past 2 weeks, the 2-year is up 19 basis points. But this surge is tied less to unwinding and more to improvement in inflation.

The rising odds for a rate hike at the December meeting, now well over 50/50 by most measures, have yet to pull down the stock market though further gains may be limited. On the week, the Dow managed only a 0.4 percent gain to 22,268 with the Nasdaq down 0.3 percent at 6,426. Oil ended over $50 for the first time since June, gaining 1.5 percent on the week to $50.64. Demand for gold eased in the week, down 1.8 percent and testing support at $1,300. The dollar index rose slightly in the week to 92.15.

The bottom line

Markets are likely to increasingly turn on inflation, especially core data that exclude what promises to be temporary hurricane-related pressure in energy prices. Hurricane effects in fact are very likely to scatter far more than energy prices, perhaps including the September employment report as well. The more volatility in the general data, the closer the focus on core inflation which takes us to the coming week and the most important inflation reading of all – the core PCE index.

Week of September 25 to September 29

Ending with core prices on Friday, one of the busiest data weeks of the year gets rolling on Monday with the Dallas Fed manufacturing report which will offer specifics on Hurricane Harvey's effects. Tuesday follows with Case-Shiller home prices which are expected to firm in what might, however, not be a good result for home sales and especially first-time buyers. New home sales and consumer confidence for September are released later on Tuesday with neither expected to show significant hurricane effects. Janet Yellen speaks about the economy midday Tuesday and could take the opportunity to address any loose ends from the prior week's FOMC decision to begin balance-sheet unwinding. Wednesday begins with durable goods orders where continued strength in core capital goods will be the desired result with pending home sales to follow in a report that, like the housing sector itself, has not shown much strength at all. The third estimate of second-quarter GDP is out on Thursday morning and a small upgrade from the second estimate's 3.0 percent is the call. Jobless claims are the week's wildcard and may show continuing labor-market effects from Hurricane Irma's landfall on Florida. The week winds up on Friday with personal income and spending where expectations are mixed with, however, a respectable showing the call for core PCE prices.

Monday

National Activity Index for August

Consensus Forecast: 0.11

Consensus Range: -0.25 to 0.13

August is difficult to assess given Hurricane Harvey's late month hit on Houston which hurt industrial production and construction. August payroll data were also soft but were sampled before the Harvey's landfall. Forecasters see the national activity index rising slightly to 0.11 vs minus 0.01 in July.

Dallas Fed General Activity Index for September

Consensus Forecast: 12.0

Consensus Range: 8.0 to 14.3 

The Dallas Fed manufacturing report will get close attention as its results will offer a direct assessment of Hurricane Harvey's impact on the state. This report had been very strong going into the hurricane with orders, production, and hiring all showing gains. Econoday's consensus for September's general activity index is 12.0 and down from 17.0 in August.

Tuesday

Case-Shiller, 20-City Adjusted Index for July

Consensus Forecast, Month-to-Month Change: 0.4%

Consensus Range: 0.1% to 0.4%

Case-Shiller, 20-City Unadjusted Index

Consensus Forecast, Year-on-Year Change: 5.9%

Consensus Range: 5.8% to 5.9%

Case-Shiller home prices have been firm but slowing held down by weakness centered in Chicago, Cleveland and New York. In contrast, western cities led by Seattle and Portland have been strong along with Miami and Tampa in Florida. Speaking of Florida, Case-Shiller will gain special attention in coming reports for August and September that will include data on cities hit by Hurricanes Harvey and Irma. For July, the consensus for the 20-city adjusted index is plus 0.4 percent vs a 0.1 percent gain in June. Year-on-year, the unadjusted index is expected to come in at 5.9 percent in what would be a 2 tenth increase from 5.7 percent in June.

New Home Sales for August

Consensus Forecast, Annualized Rate: 583,000

Consensus Range:  555,000 to 620,000

New home sales are expected to accelerate in August to a consensus 583,000 annualized rate vs what was a disappointing 571,000 rate in July. Existing home sales in August were held lower by significant disruption in Houston in the wake of Hurricane Harvey and a similar effect is a risk for sales on the new home side. Another effect may be seen in supply as housing completions, due to Harvey, fell back in August.

Consumer Confidence Index for September

Consensus Forecast: 120.2

Consensus Range: 118.0 to 123.0

The consumer confidence index, underpinned by unusually strong assessments of the labor market and income prospects, has well surpassed the consensus over the last three months, rising 2.3 points to 122.9 in August which next to 124.9 in March this year is the highest score since the dotcom days of December 2000. But the August results did not include the Texas effects of Hurricane Harvey at month-end nor Hurricane Irma's devastating Florida landfall in September. Yet forecasters aren't expecting much hurricane effect, calling for a modest retreat to 120.2 for September. The only negative in the report has been inflation expectations which have been extremely weak.

Richmond Fed Manufacturing Index for September

Consensus Forecast: 13

Consensus Range: 6 to 14

Robust hiring has only been one of the strengths of the Richmond Fed's manufacturing index which like other regional reports has been running at the hottest levels on record. The index in August came in at 14 with forecasters calling for 13 in September.

Wednesday

Durable Goods Orders for August

Consensus Forecast, Month-to-Month Change: 1.5%

Consensus Range: 0.4% to 3.5%

Durable Goods Orders, Ex-Transportation

Consensus Forecast: 0.4%

Consensus Range: -0.2% to 0.7%

Durable Goods Orders, Core Capital Goods (Nondefense Ex-Aircraft)

Consensus Forecast: 0.3%

Consensus Range: 0.3% to 0.5%

Strength for core capital goods, both in orders and shipments, has been the key highlight of recent durable goods reports where aircraft-distorted headlines have otherwise made the report hard to read. A negative in the report has been weakness in vehicle readings, one consistent with weakness in underlying consumer sales and very apparent in the separate monthly report on industrial production. August durable goods orders are seen bouncing back from July's 6.8 percent plunge with a 1.5 percent gain in August. Ex-transportation orders are expected to climb 0.4 percent on top of July's 0.5 percent gain while the call for core capital goods orders (nondefense ex-aircraft) is a gain of 0.3 percent following a 0.4 percent increase in July.

Pending Home Sales Index for August

Consensus Forecast, Month-to-Month Change: -0.1%

Consensus Range: -0.6% to 1.2%

Pending sales have been accurately forecasting this year's steady weakness in the existing home sales report. The South, by far the largest housing region, was very weak in July and probably won't be getting any help in August from Hurricane Harvey. After July's 0.8 percent dip, the Econoday consensus for the August pending sales index is a dip of 0.1 percent.

Real GDP: 2nd Quarter, 3rd Estimate, Annualized Rate

Consensus Forecast: 3.1%

Consensus Range: 2.8% to 3.2%

Real Consumer Spending, Annualized Rate

Consensus Forecast: 3.3%

Consensus Range: 3.1% to 3.3%

GDP Price Index

Consensus Forecast: 1.0%

Consensus Range: 1.0% to 1.0%

Boosted by upward revisions for inventories and construction, the third estimate for second-quarter GDP is expected to firm slightly, to a consensus 3.1 percent annualized rate vs the headline catching 3.0 percent rate of the second estimate. Consumer spending is expected to hold at the second estimate's 3.3 percent rate. The GDP price index is seen unchanged at 1.0 percent.

International Trade In Goods for August

Consensus Forecast, Month-to-Month Change: -$65.7 billion

Consensus Range: -$66.1 to -$64.0 billion

The goods deficit in August is expected to widen to a consensus $65.7 vs $63.9 billion in July ($65.1 billion initially reported). Exports of consumer goods and autos weakened in July while imports also dipped on declines for autos and industrial supplies as well as petroleum where an August gain is likely given the month's hurricane-related increase in oil prices. Advance data for both wholesale inventories and retail inventories in July, which are also GDP inputs and which have been mixed, will also be released with this report.

Initial Jobless Claims for September 23 week

Consensus Forecast: 270,000

Consensus Range: 260,000 to 300,000

Hurricane volatility makes initial jobless claims difficult to call but the consensus for the September 23 week is 270,000 vs an average of 280,000 in the prior three weeks which all had hurricane impacts. Claims in Texas rose sharply at first but have since eased while the initial impact of Hurricane Irma on Florida was heavy.

Friday

Personal Income for August

Consensus Forecast, Month-to-Month Change: 0.3%

Consensus Range: 0.1% to 0.4%

Consumer Spending

Consensus Forecast, Month-to-Month Change: 0.1%

Consensus Range: -0.2% to 0.3%

PCE Price Index

Consensus Forecast, Month-to-Month Change: 0.3%

Consensus Range: 0.2% to 0.4%

PCE Price Index

Consensus Forecast, Year-on-Year Change: 1.5%

Consensus Range: 1.4% to 1.6%

Core PCE Price Index

Consensus Forecast, Month-to-Month Change: 0.2%

Consensus Range: 0.1% to 0.3%

Core PCE Price Index

Consensus Forecast, Year-on-Year Change: 1.4%

Consensus Range: 1.4% to 1.7%

Both income and spending rebounded in July but not the Federal Reserve's PCE price indexes which remained lifeless. Personal income is seen is rising 0.3 percent in August vs July's 0.4 percent gain though consumer spending, reflecting August weakness in retail sales, is expected to slow to a 0.1 percent increase vs 0.3 percent in July. PCE prices, both overall and for the core, managed only 0.1 percent gains in July with the year-on-year rates both at a very soft 1.4 percent. Overall prices, boosted by higher energy prices that in part reflect gains for energy, are expected to rise 0.3 percent with the yearly rate seen at 1.5 percent. Core prices, which exclude both energy and food, are seen rising 0.2 percent in the month for a yearly rate of 1.4 percent in specific results that would not further heighten expectations for a December rate hike from the Federal Reserve.

Chicago PMI for September

Consensus Forecast: 58.6

Consensus Range: 55.0 to 60.0

The Chicago PMI has been one of the highest flyers of any regional reports this year, posting consistent readings near or over 60. For September, forecasters are calling for 58.6 vs August's 58.9. This report tracks both the non-manufacturing and manufacturing sectors of the Chicago-area economy.

Consumer Sentiment Index, Final September

Consensus Forecast: 95.3

Consensus Range: 94.5 to 96.0

The consumer sentiment index showed only marginal impact from Hurricanes Harvey and Irma in the preliminary September report where the index slipped only 1.3 points to what is a still very strong 95.3. The current conditions component in fact hit the highest level in 17 years driven by extraordinarily strong confidence in personal finances. But inflation expectations, like those of other reports, remained unusually low despite hurricane-related increases in energy prices. Econoday's consensus for the final September, at 95.3, is unchanged from the preliminary reading.