Soft economy is not all about the weather

Things started going south in November—before the elements did, writes Morningstar's Bob Johnson.

Robert Johnson, CFA 19 February, 2014 | 3:03AM

U.S. equity markets had a good week, with the S&P 500 moving up 2.4% while the 10-year U.S. Treasury bond yield moved up from 2.68% to 2.74% as the market shrugged off another set of poor economic indicators.

Retail sales in January were down 0.4%, and the previous two months were revised sharply downward. Industrial production fell 0.8%, and unemployment claims moved up a little. On the positive side of the ledger, single-week shopping center growth popped above 2% again after last week's disaster. New U.S. Federal Reserve Chairwoman Janet Yellen's testimony on Capitol Hill also seemed to reassure markets (more of the same policies, tapering to continue absent an economic train wreck). Finally, a quick agreement on the lifting of the debt ceiling until 2015 seemed to calm markets considerably.

European GDP growth looked a little better, with growth of about 1.1% annualized for the fourth quarter. My colleague in the fixed-income department, David Sekera, offered that the European growth was good enough to keep the credit situation from getting worse, but not enough to pull them out of the ditch, either.

FactSet estimates that fourth-quarter corporate earnings still look set to grow about 8.3% year over year. However, this week's reports were lackluster, with   Cisco  CSCO offering this week's negative surprise. Cisco, along with   Nestle  NSRGY and several others, reported that weak emerging-markets activity affected results. Despite more tempting emerging-markets valuations, I continue to worry about emerging-markets fundamentals in a world of soft commodity prices and continuing adjustments in China.

The economic indicators I track have been trending poorly lately, yet equity markets continue ever higher. Markets have been initially selling off on the bad news, before resuming their upward trend an hour or two later. I think the market is ignoring all the economic data right now, assuming that the only reason the data looks terrible is the unrelenting flow of bad weather. I am not so sure it is all about the weather, as some of the negative trends started in November when weather was not as much of an issue. And in this week's data set, online retail sales, which should have been boosted by bad weather, fell instead.

I have always believed that consumers are a key driver in the short-term economy, and their worries continue to mount. It now looks like extended unemployment payments are off the table, with Congress in a two-week recess. Those benefits expired way back in December. The food stamp program was automatically cut in November, and now the recently passed farm bill will further trim the program. The cold weather has also meant mounting utility bills for consumers, limiting spending on other categories. Higher interest rates resulting from the Fed's tapering program won't be helpful, either. And with equity price turning more volatile and home price increases decelerating, those sectors cannot be counted on to boost the economy in 2014. Employment data, which has been at least a little affected by the weather, also suggests slower income growth, especially in the short run. Higher health insurance deductibles and premiums will also be another source of stress in 2014, though more widely available coverage may offset some of that weakness.

As discussed below, the GDP estimate for the fourth quarter is likely to be revised down to the 2.0%-2.5% range from 3.2%, with the government overestimating retail sales, inventory growth, exports, and residential and public building spending. Interestingly, the large change in the single quarter will still likely leave the year-over-year growth rate at 1.9%. Paradoxically, a lower fourth-quarter 2013 GDP figure would make it easier to hit my 2.0%-2.5% growth rate for 2014.

Retail sales slip more than expected
Retail sales is one the key metrics in determining the overall strength of the economy, along with industrial production, employment, and income. And as I have stated before businesses will not produce things and expand capacity unless consumers are buying. Businesses do not produce goods for the fun of it. That is why this week's retail sales report was such a big disappointment for the overall economy. I do caution that weather may have played an important role in holding back the data. Still, it was not all about the weather, as even online sales, which should have spiked with the bad weather, were down month to month. And some categories, particularly building materials, benefited from the bad weather.

First, examining the headline data, overall retail sales declined 0.4% between December and January, led by sharply declining auto sales. Without the poor auto sales, retail sales were unchanged for the month, even before adjusting for the effects of inflation. The previously reported data just looked too strong versus reports from individual retail stores and weekly shopping center data. November was revised down to a 0.2% decline from a small gain, and December's reported, eye-popping 0.7% gain was lowered to a more down-to-earth gain of 0.3%.

Those large reductions are very meaningful for a couple of reasons. Most notably, the large revision will likely mean another 0.1%-0.2% reduction in the fourth quarter's reported 3.2% growth rate. Given other data released last week, that 3.2% estimate is increasingly looking like a 2.2% GDP growth rate (in a range of 2.0%-2.5%) with all of the newly available data. The retail sales revisions also seem to be indicating that the softness in retail sales is not entirely weather-related, as even November, which was not nearly as affected by adverse weather, was one of the very weakest months for retail sales.

Year-over-year retail sales trends deteriorating
The year-over-year comparisons show similar disappointing results, though on a more muted basis. The table below shows retail sales, excluding autos and gasoline, on a year-over-year, three-month moving average basis. The trend in raw sales is not good, and adjusted for inflation, it is even worse.

Noninflation-adjusted retail sales have been slipping on a relatively steady-state basis since August while the inflation-adjusted data has been slipping just since November. But that downturn has been much more abrupt, with the rate of growth being cut almost in half in a very short period of time. With sales growth slipping and inventories increasing, the outlook for manufacturing should be a little softer in the months ahead. Some of that is already evident in the January report on Industrial Production discussed below.

The retail sector data shows across-the-board weakness
Though auto sales weighed heavily on the headline decline figures, the weakness was pervasive across sectors. The sector weaknesses were also more persistent than usual. With retail sales data, a good month is almost always followed by a bad month, which moves things closer to average growth rates. This time, data in both December and January looked soft.

The softness in sporting goods, hobbies, and books was at least a little surprising to me as I would have expected that cold and snowy weather might have helped the sales of winter sporting gear. Or I would have thought more people would have had time to work on hobbies or read a book, with all the cold weather.

Clothing was one category that did reverse, with bad weather likely accelerating the sale of winter goods, but then, there wasn't much left to sell in January, and there certainly wasn't a lot of interest in spring merchandise. Furniture weakness probably reflects softer existing-home sales that we have talked about for some time. On the plus side, sales of snowblowers, shovels, and salt probably helped move sales at building material stores. Unfortunately, that probably won't be repeated in the months ahead, and the spring planting season is likely to get off to a slow start.

Gas sales mainly reflected an increase in price and not in volumes. The large swings and the really poor number for electronics remain a mystery to me, but likely some of that is due to softer prices. Groceries continued to benefit from people stuck at home and needing to eat in. However, the stores didn't have the massive stock-up before the storm rush that pushed up December sales so dramatically.

One category that didn't make the big movers list was the sharp decline in nonstore retailers, which includes online retailers such as Amazon. Those sales declined 0.6% in January, though that followed a relatively strong 1.3% gain in December and even stronger gains in November.

Restaurant sales take it on the chin
One category I always watch to gauge consumer confidence is restaurant sales. They were down 0.6% in January following a 0.7% decline in December following a relatively stronger fall. The data was probably affected by weather (sales here are difficult to make up at a later date) as well as ongoing price wars in some restaurant sectors. So the news here is not great, but also not the disaster it could have been. Still, poor weather may keep a lid on restaurant hiring, which had been one of the few bright spots in 2013 employment.

Industrial production takes a tumble
Recently, some production data had appeared stronger than consumer demand would have seemed to suggest. Some purchasing data was even stronger than the reported production results. As a result, inventories were on an upswing, as production increased faster than sales. This was particularly true in the auto industry, which has sharply ramped up production because of new models, abnormally low inventories after Hurricane Sandy, and aggressive forecasts of long-term auto demand. I have been harping about an overly aggressive auto industry for some time, but this week the Wall Street Journal ran a long story on the sharp buildup in auto inventories.

The graphic showing a huge jump in days of inventory on hand is particularly illustrative. I caution, though, that poor weather depressed the sales part of the calculation, and these data do not appear to be seasonally adjusted. (Inventory is generally built in the winter months to support spring sales.) However, the general pattern is still quite evident.

In fact, just after we went to press last week,   GM  GM announced some incentives in an attempt to move more vehicles. So far Detroit is fine-tuning price schemes, but not changing production by a lot, just yet.

This is all a very long windup to the fact that industrial production for manufactured goods fell by an unusually large 0.8% from December to January (that's 9.4% annualized). Yes, weather can take a lot of the blame, but rising inventories and slow demand also began to extract a heavier toll. Both monthly and year-over-year production figures (on a three-month averaged basis) have been falling since October, so the problem isn't entirely about weather.

I have been sharply critical of the ISM Purchasing Managers statistics that showed near-boom conditions through November, a month after sequential industrial production statistics peaked. Nevertheless, I must admit the huge falloff in the ISM January data did correctly point to this week's poor industrial production numbers.

The month-to-month data is highly volatile, as the table above shows, with many negative and positive numbers. The trend is a lot easier to see in the year-over-year averaged data that shows production growth steady in the 2.0%-3.3% range for the last year. The annual data is even more stable. After a big post-recession bounce-back, annual industrial production growth has averaged just over 3%, hardly a boom.

So while manufacturing gets a lot of attention and makes a lot of headlines with its breathtaking monthly swings, the annual data shows that manufacturing growth is moving at a slow but steady pace that is never much off a 3% trend line. Given the continuing extreme weather and now the poor retail data, the monthly data is likely to be on the low end of the recent averages for at least a few months. Then we can probably count on a few better months to bring the manufacturing sector back up to its longer-term averages.

Autos push the industrial production figures down sharply
Autos were the big losers in industrial production this month, falling 5.0% from the previous month. Higher inventories and some production slowdowns at   F  F, as well as weather, affected the January data. That said, just about every category was down in January except home electronics, clothing, energy, and information processing. Anything tied to housing seemed to be particularly soft, including wood products, appliances, and furniture. Again, this probably relates to the softer sales of new and existing homes over the past several months, a trend that may not improve much over the short term.

I was pleased to see several technology-related categories acting well. I suppose some of the cloud-computing-related sales are helping overall industry sales. Morningstar's technology team continues to believe that cloud computing is a game changer in 2014.

Perennial winner aerospace was down for the third month in a row, as   Boeing  BA massive production ramp-up nears completion and fall defense sales begin to dominate the category. Boeing has now moved to produce 10,787 aircraft per month (at $200 million or more apiece) and is expected to stay at that level through 2016 before beginning another medium-size increase. The 787 wasn't in meaningful production as recently as 2010, and even by the end of 2011 Boeing was producing just 2.5 aircraft per month of the pricey airliner, according to our manufacturing analyst Neal Dihora.

Business inventories continue to push ahead, though not as fast as the government expected
For the just-reported December data, sales were up a meager 0.1% while inventories grew a more robust 0.5%. The inventory follows a gain of 0.4% in November and 0.8% in October, suggesting that inventories are a bit higher than they should be, and that inventory growth is unlikely to be a contributor to GDP growth in the first quarter of 2014.

I usually don't talk about inventories because they usually don't do that much to affect the economy except at key turning points. Also, I am not a huge fan of inventory data because it's not one of the more accurate indicators, prone to volatility and myriad potential errors. I learned this the hard way more than 30 years ago as an internal auditor. One of my jobs was affirming inventory balances on the books against actual goods in the warehouse. Besides the physical labor, the job was difficult because the counts were often off. We continuously had study groups to investigate inventory shrinkage. And outside of a few cycle counts, the big adjustments and physical inventories happened just once a year. While inventory technologies and plant security have surely gotten better, I would still hate to base my economic forecasts off this particular piece of data, especially in the short run.

Recall that inventories are used to construct the GDP report, with production determined by adding or subtracting changes inventory to sales levels to determine what was produced (the P in GDP). The general trend inventories have been up lately, aiding the GDP calculation in both the third and fourth quarter. Producers made a lot more than what was sold. If that was on purpose because business prospects look great, that is a good thing. However, if inventories built up because demand flagged compared with what was expected, that is a bad thing. It means production will need to be cut in the future.

Though inventories showed healthy growth in December, the BEA thought inventories would grow even faster. My preliminary analysis suggests that the inventory contribution to GDP in the fourth quarter will be slightly less, perhaps around 0.1% than the initial read of a 0.4% contribution. Not far off the mark, but it adds to the already long list of overestimated variables that the BEA used in its initial read of fourth-quarter GDP.

Housing data and inflation data on tap for next week
Builder sentiment, housing starts and permits, and existing-home sales are all due next week. Builder sentiment is expected to stick at 56, near its recovery high as mortgage rates remain low, even in face of Fed tapering and more land becomes available for building. Starts and permits are expected look a little weaker, with starts expected to fall to 963,000 units on an annualized basis from 999,000 units in December--well below the recovery high of more than a million units. Poor weather and poor permit data for the previous month are the reasons for the pessimism. Existing-home same sales are expected to continue in a funk for the same reason. Those sales are expected to fall to a 4.7-million rate from 4.87 million in the previous month, based on the weather and poor permit data. That is well south of the recovery high of 5.4 million units. Low existing-home sales was the main reason investment in homes was down in the fourth-quarter GDP report.

It's hard to believe that we all used to live and die by the inflation report, and it was just a year ago when investors were clamoring for special inflation reports and commodity/inflation protected investment products. The Consumer Price Index is expected to be up 0.2%, down from December's quick spurt to 0.3%. If correct, that would move the year-to-year rate up to 1.5%, a little closer to the Fed's target of 2%. Higher utility prices and gasoline prices have moved this indicator.

About Author

Robert Johnson, CFA

Robert Johnson, CFA  Robert Johnson, CFA, is director of economic analysis for Morningstar.