Diagnosing the Market's Mood Swings

It was yet another week of high volatility for every corner of the world markets. For most days this week and the overall equity returns for the week, results were down. But that didn't stop the markets from rallying sharply on Friday. Bonds were also on a roller coaster of their own, with interest rates on the 10-year U.S. Treasury bond temporarily dropping below 2% on Wednesday before rebounding to a still-low 2.2% on Friday.

The market continues to oscillate between days when bad news is bad news and days when bad news is good news. Some days, bad news suggests that interest rates will stay lower and for longer than expected. On other days, markets seem to realize that bad news will eventually mean bad earnings. Speaking of earnings, it wasn't a stellar week.  Google (GOOG) surprised on the downside as did  eBay (EBAY),  Netflix (NFLX), and  Advanced Micro Devices (AMD). Still, banks' and  General Electric's (GE) numbers generally pleased markets, and overall growth rates in earnings still look OK. Whether those relatively optimistic growth rates fully include a slowing Europe, South America, and China, a stronger dollar, and potentially higher labor rates remains to be seen.

The news out of Europe continued to be worrisome, with more bad data and sentiment surveys out of Germany. The new worries caused rates to soar in Greece, Ireland, and Portugal as everyone began to question the European recovery and various budget-balancing maneuvers. I have never been an optimist on the European situation as structural issues--different in every country--continue to hold back any progress. The sniping among member countries, never a good thing, seems to be on the rise, too. Until the deeper structural issues are fixed, it would take a truly booming world economy to pull Europe out of its lethargy. A poorly performing Europe will hold back growth of Fortune 500 companies, but probably not U.S. economic data.

Unfortunately, a seemingly poor U.S. retail sales report on Wednesday suggested that the U.S., one of the few shining lights in the world economy, was slipping into the same lethargy. Although a fuller analysis of the retail sales data actually suggested a strong U.S. consumer, the headline decline in retail sales was enough to truly panic a market that was already running scared. Fortunately, better industrial production numbers on Thursday and modestly improved housing data on Friday seemed to suggest that the U.S. wasn't sliding down the same slippery slope as Europe.

Looking ahead to the final quarter of 2014, lower interest rates, moderating food prices, and falling gasoline prices combined with a relatively high savings rate suggest better days ahead for U.S. consumption. That's even before considering initial unemployment claims made a new recovery low this week. I am just a little surprised, and perhaps a little concerned, that some of these improvements have yet to show up in force in any of the spending-related reports. I suppose ebola and ISIS headlines aren't helping things in the short run, but those issues usually fade from consumer minds relatively quickly.

Retail Sales Aren't Slowing
The headline retail sales figures and some the newspaper articles accompanying the headlines were pretty scary. The surface analysis was frightening enough to contribute to Wednesday's stock market rout.

The report indicated that sales were down 0.3% between August and September, which would be the worst performance since January. It managed to underperform muted expectations of a 0.2% decline. Worse, the retail sales number was still down 0.1% when excluding automobiles and gasoline.

However, everyone tends to forget how volatile these numbers can be. Really good months are usually followed by really bad months that bring the overall average back to earth. The headline growth number for August was a stunning 0.6%. Therefore, one should not have been terribly surprised by the poor October number. In fact, the year-over-year, three-month average growth number was benign and well above the average of the past 12 months.

At 4.2% the year-over-year growth rate was well above the 3.5% average and was still very near the high for the past 12 months. The numbers were nearly as good when the effects of inflation were stripped out, which showed growth of 2.4% after our estimated inflation rate of 1.8% (which could prove to be a little high, based on recently falling commodity prices).

The month-to-month numbers in isolation are exceptionally volatile. I don't think this is necessarily the result of fickle consumers, but more likely statistical and calendar effects. I have written about difficulties in seasonally adjusting data for a company like  Target (TGT). Target used to sell primarily nongrocery items, but is now deriving substantial revenue from groceries. The seasonal factors for groceries versus general retail merchandise are just about opposites, confounding statisticians. Changing school start dates and various back-to-school tax holidays are another source of statistical volatility this time of year.

Unfortunately, this renders some of the month-to-month category data nearly useless. Still, the sector data is worth a glance. There were a lot of down categories in September and not many that were up.

It is interesting to note that a lot of the really strong categories in August fell back in September, namely clothing and motor vehicles. In terms of losers, clothing was probably affected the most by school and tax holiday issues. However, there is also a background issue of people buying fewer but more expensive items of clothing, which I haven't fully gotten my arms around yet.

The relative lethargy of the housing industry is not helping the building material sector, another big loser this month, either. Two back-to-back disappointing nonstore retail (e-commerce companies such as  Amazon (AMZN)) numbers are a bit surprising, though our retail team notes that it has seen a similar pattern in past Septembers. On the positive side, electronics appear to have gotten a bit of a boost from the new iPhone, which may have taken cash and time away from shopping for other goods.

One category that convinced me that retail sales and the consumer are probably in good if not great shape was a very healthy 0.6% jump in restaurant sales. That is a very significant increase. It also follows a 0.6% increase in August. This is one of the key categories in evaluating how good consumers are feeling in the short term. Importantly, this category is not subject to as many seasonal swings and tax holiday issues as conventional retailers. This is yet another reason not to panic over the single-month, retail sales growth rate that threw everyone into such a tizzy this week.

​As a side note, the weekly retail sales data still looks strong, too, backing up my contention that retail sales are not in some type of free fall. The weekly sales data is still showing year-over-year growth of about 3.8%, down slightly from peak readings of just barely over 4% this summer. Looked at on a quarterly basis, which I haven't done in some time, shows a relatively powerful trend.

Housing Market Still Stuck in Neutral
The housing start and permits data looked better on a month-to-month basis in September, but overall starts and permits seemed to have stalled out at the 1 million or so units per year rate. The unaveraged unit permits numbers below show the pattern of limited growth.

​

Housing Still Performing Well Below Long-Term Norms
The good news is that housing starts are well off their recession low of about 500,000 units, but it has been exceptionally difficult to break that 1 million-unit barrier. Most students of the housing market, including me, generally believe that household formation trends, population growth, and normal wear and tear would point to needs of at least 1.4 million units per year. That doesn't even consider the underproduction of homes during the recession. Given an improving employment market and relatively low rates, we all hoped that advances in the housing markets would be consistent, and steady. That would get us back to normal in fairly short order. In fact, while no one believed that we would ever get back to the peak-production 2.2 million-unit level, the thought that we could do better than the 1.4 million-level at some point was nearly universal. But very tough lending conditions, high new home costs, and changing tastes have all conspired to keep a lid on new housing construction. Don't get me wrong: It's not a disaster. However, housing starts this year are likely to grow by only 6%, and the outlook for a higher growth rate next year is not great.

Affordability Issues Are Limiting Demand for New Space and the Increased Utilization of Already Built Homes
The high cost of fuel and energy as well as construction and land costs have conspired to make new homes, especially single-family homes, unaffordable. Anecdotal evidence suggests that the spread in price between a new and existing home has soared to six figures in many markets. This has in turn forced people to use existing space more creatively in already-built units. Young people are living with their parents longer, and most singles now live with roommates to reduce costs. Furthermore, two-bedroom units are being pushed into service as four-person facilities. And some typical single-family homes are now shared by unrelated roommates. These types of trends break down our usual housing market rules of thumb. I still believe that as young people marry and have children, that some of the more normal rules of thumb may come back into play. I remain hopeful that this will be the most likely path. However, there is some possibility that housing in the U.S. begins to look more like parts of Europe, where housing is exceptionally expensive. There, homes are often passed from generation to generation and a much higher percentage of the population rents.

All of that said, I think a lot of market participants need to rethink their theses on the housing market. Tract homes far from the city center are out and close-in rental units are in. Smaller tastefully decorated homes are in, McMansions are out. High-cost, massively featured homes are less popular while simpler, cleverly designed homes are making a comeback. And the key beneficiaries of this housing cycle might be remodelers and apartment builders and not tract homebuilders of the past.

Also, the high prices of homes in California and the Northeast will continue to force employers to consider plant and business locations where there is a bigger stock of land and lower-priced homes for employees.

Builder Sentiment Dips
After four straight months of improvement in builder sentiment, this key housing metric dipped in September. Overall, the index dropped from 59 to 54 with current sales, projected sales, and current traffic all falling. That said, any reading over 50 suggests that more builders are seeing better conditions than worse. Again, the numbers suggest neither boom nor bust. Coincidentally, the reading was the same both this October and last October, suggesting little is changing in housing growth rates in the near to intermediate term.

Industrial Production Suggests Manufacturing Still Healthy
The headline industrial production growth rate surged to 1% in September, more than recouping the 0.2% decline registered in August.

Unfortunately, this report is the opposite of some of the other data this week that was better than the headline represented. This month's number was inflated by a 0.9% jump in utilities, which is usually more related to weather than any changes in the economy. The mining sector grew a surprisingly strong 1.8%, which comes as a bit of a surprise given the dramatic decline in energy prices. Without these two factors, the manufacturing sector grew at 0.5%, still very healthy but not nearly as the dramatic 1% headline increase.

The year-over-year data remains healthy and well above averages and the purchasing manager data suggest that things will be good in the months ahead, but perhaps not quite as robust as in certain points this summer. My only lament here is that I wish the manufacturing sector were bigger.

Sector data wasn't a lot of help this month. It is a huge surprise, though, that overall production grew at 0.5% even though the auto sector declined 1.4%. Autos have almost always been in the vanguard of improving production numbers this recovery. Taking over the leading role this month was aerospace, where production grew 1.7%. There were a lot of bounce-back categories this month that suffered large losses last month and now look much better. Furniture went from a negative 1.2% to 2.4%, and apparel went from negative 3.4% to a positive 1.4%. These huge swings suggest measurement or seasonal issues and not massive real-world changes. Chemicals, plastics, and rubber products also had great months, potentially benefiting from lower energy prices at last.

One last caution on the manufacturing data: The data here seems to be showing sharp acceleration that really doesn't fit some of the other data. Consumption looked at the right way is increasing, but not at an accelerating rate. If that is the case, why are we building more? And housing construction doesn't seem to be adding to the story recently, either. Likewise, exports would seem to be slowing not accelerating, which also doesn't square with some very meaningful increases in production. That is, unless inventory is stacking up somewhere, which doesn't seem to be the case.

Will the U.S. Show Deflation Next Week?
Next week's calendar is relatively thin, with a couple of housing reports, the consumer price report, and the flash Markit data from world purchasing managers. Given near-deflationary results from Europe (just 0.3%--that is year over year and not month to month) and this week's decline in U.S. producer prices, expectations of 0.1% price growth in September look downright inflationary, and perhaps a little on the high side. At 0.1% growth for September, year-over-year inflation would drop back to 1.8% (which should approximate the Social Security Cost of Living Adjustment for 2015, which is based on a very similar average to this one) on a three-month average basis, down from 2.1% this summer and likely to drop even lower in the months ahead. While energy will drive most of this move, other categories might show slowing price appreciation, too. These low rates of inflation might explain some of the lower nominal growth rates in some reports recently, including retail sales. For now, lower inflation is very good news for U.S. consumers, at least if things don't get out of hand. That said, it may make it increasingly difficult for corporations to raise prices.

Based on small changes in the pending homes sales report, expectations are low for the existing-home sales report. Unit sales are expected to decline from 5.05 million units in August to 5.0 million in September. Given lower rates and an improving job market, I believe there is still a decent chance that existing-home sales will improve and beat expectations.

New home sales spurted higher in August even in the face of other weak housing data. This seems to imply either a big revision to next week's new home report for August or a huge decline in the September data, which would be consistent with this week's modest housing starts report. For now, new home sales are expected to fall from 504,000 units to 455,000 units.

The Markit data on the manufacturing economy is always market-moving, and I don't expect next week to be any exception. The U.S. data has been relatively strong, China has been weak but improving, and Europe had been falling for some time. Any reversal in these trends could get the market truly worked up. We won't talk about what might happen if the numbers are a disappointment. Let's just say that this week might look like a walk in the park if both Europe and China were to drop under 50, the level separating growth from decline.