A Happy Summer Ending

“Summer’s lease hath all too short a date.” William Shakespeare, Sonnets

It was an exciting week. Curiously, though, given the gravity of events, it was one almost without surprise – in general, a state of affairs that is more than agreeable for equity markets.

The major scheduled events did not disappoint: The Federal Reserve threw the doves the bone I expected, specifically by leaving in the sacred phrase “for a considerable time” (i.e., zero interest rates) in its monetary policy statement. Scotland did not secede from the United Kingdom. The Alibaba (BABA) initial public offering (IPO) on Friday packed a punch, opening up over 40% from the offering price and setting a record for the largest IPO ever in the process. Options expiration had its usual morning rally.

Yet it was all anti-climactic. The Fed’s hand had been tipped the day before the statement by a lengthy Tuesday webcast from Wall Street Journal reporter and unofficial central bank liaison Jon Hilsenrath. The depth and breadth of Hilsenrath’s message, staged as a question-and-answer session, displayed an extraordinary degree of access to Fed thinking and made for some serious inside baseball. I’ve no doubt his status must irritate quite a few competitors in the media, judging from the stony silence (apart from the Twittersphere) surrounding otherwise perfunctory reporting.

An understandable reaction, given the competitive nature of the business, though I’m sure nearly all of his critics and/or rivals wouldn’t hesitate to switch places with him (I wouldn’t, not that there’s any danger of being asked – I don’t think I could avoid starting a policy argument!). As informative as it was, though, I wonder if the whole affair was a good idea. Hilsenrath seems a decent enough fellow and one would think that the extensive picture he provided of Fed thinking serves a useful purpose in providing some additional clarity, not to mention dampening some of the endless speculation that swirls around practically every letter of Fed statements.

At the same time, however, it seemed slightly out of kilter. The webcast itself turned a mild stock market sell-off based on rumored fears that the statement would drop the phrase, “for a considerable time,” into a big reversal rally. I was on top of the webcast and so didn’t feel wronged, and I’m sure that much, if not most, of the professional side of the market was either prepared or had an ear cocked. Still, given the minimal prior visibility and a release during morning market hours, I couldn’t escape the feeling the playing field had been tilted.

I wrote a lengthy piece for Seeking Alpha on the Hilsenrath-FOMC double play, the gist of which might be inferred from the title that I had tweeted out the day before, in reaction to the webcast: The Fed: There is no bubble, there is no timeline, there is no exit strategy. I’m all for central banks being discreet about ever calling things a bubble, but I also expect them to refrain from saying something is not a bubble, and Hilsenrath’s webcast was quite plain that such is the Fed’s view of financial asset prices. I’m not the only one who wonders if the Fed has too much fear of market reaction, to the extent the bank goes too far in its efforts to keep the waters calm. Extended periods of forced calm tend to give birth to a damaging amount of subsequent volatility, as Nixon’s wage and price controls demonstrated back in the 1970s.

At any rate, the stock market welcomed the news by seeing it as dovish, while the bond market sold off on the news by seeing it as hawkish. The latter didn’t like the “blue dots” rate-estimates graphic (you can see the tabular form here) that seemed to imply members thought rates would be over 1% by the end of next year. Chair Yellen hastened to dilute the message in her post-meeting news conference by treating the dots as little more than casual dart throws without much conviction. She wasn’t that convincing, but time will tell.

The Scotland vote was also anti-climactic. The prior week’s poll result showing the secession movement in the lead roiled the markets but was never duplicated, and the pollsters themselves were unanimous in predicting a “no” victory on the eve of the vote. The tally wasn’t really close, with the secessionists quickly conceding the next morning while most of America was fast asleep. A pro-secession victory would have been ugly for markets, but fortunately we will never know exactly how ugly.

As for the AliBaba IPO, I admit to being sick of the whole thing by the time it had opened for trading. I can understand why the business networks and media would treat it as a big deal and possibly hope for big ratings therein, but by mid-day it had been laid on so thick I was wondering why all the newscasters on CNBC hadn’t already changed into BABA jumpsuits. The IPO didn’t as do much for the markets as it did for its investors. Stocks opened up on the strength of the Scotland vote and quarterly options and futures expiration, but turned listless in the afternoon and ended up about flat, with the S&P 500 slightly down and the Nasdaq (where BABA is not listed – it’s on the NYSE) turning its gains into losses.

The downside of these victories is the subsequent increase in complacency and the ever-perilous sense of assurance that nothing can harm the markets. Stocks finished up on the week, with new highs set on Thursday, but by day’s end on Friday signs of weakness were everywhere. The Russell 2000 small cap index continues to struggle, and risk bets like emerging markets and China were in decline as well. The period between quarterly rebalancing and the end of September, which is also the end of the third quarter, is typically a difficult one for equities. Stocks aren’t overbought on a short-term basis, but remain very seriously so on a long-term basis and the little relief rally provided by the Fed and Scotland could already be over. I was tempted to title this week’s essay as, “The Fall Approaches,” but it was just too sunny of a day and summer is too short as it is.

The Economic Beat

Last week highlighted what has become an intriguing divergence in the economy – the widening gap between survey responses and the related production data. I’ve made ongoing references to the drawbacks of surveys over the years, but recently the yawning gulf has me addressing the subject more head-on.

Surveys are useful, but most of the higher-profile economic ones fall into the category of diffusion-style affairs. The responder generally has three choices: better-same-worse, or perhaps good-fair-poor. The down choices are subtracted from the up choices, a little seasonal adjustment is thrown in, and voila! the sentiment score hits the wires.

As I’ve often said, the biggest drawback of such a survey is that they don’t tell you whether business is a dollar better or a million dollars better. Five companies grow their business a buck by each taking one from company six, and the net change in the economy is zero. In the diffusion survey, it looks like the group is hitting it out of the park.

A second, newer problem is survivor bias. A lot of smaller and/or marginal companies have gone out of business or been swallowed up in the last two recessions, and their weaker answers may have disappeared indefinitely. Manufacturing is a much smaller part of the economy than it was in the 1990s. The result can be that the total output in one year may have had a lower score than the same output today, or put another way, higher scores today might be representing fewer dollars of output
A softer issue is the reliability of answers. All economic surveys tend to be influenced by headlines about the stock market and economy. Many answers are gut responses made under hurried circumstances, perhaps by subordinates whose answers might be mainly guesswork.

The two biggest examples of a gap this week were the August industrial production report from the Federal Reserve, and the housing starts and permits report that follows the Housing Market Index (homebuilder sentiment) from the National Association of Home Builders (NAHB). They were both at sharp contrast with their related survey reports. The production reports are only initial estimates and subject to occasionally large revisions, so it would be rash to start making lapidary pronouncements. Nevertheless, the pattern is a recurring one.

The purchasing manager index for manufacturing from the ISM is the granddaddy of them all, and the August number stood at a very high 59, with new orders shooting to their highest reading – I don’t say “level,” because we don’t actually know what the level of orders were – in ten years. The July readings were also well above-average, at 57.1 for the broad index and 63.4 for new orders.

The Fed, however, reported that industrial production actually fell by 0.1% in August, with manufacturing itself falling a much larger 0.4%. As noted above, that’s only an initial estimate, and a revision upward to say, +0.1% in place of (-0.4%), while unusually large, would not be completely unheard of. Even with such an unlikely event, however, the end result of only 0.1% would still be well below what the market was inferring from the survey results.

A quite similar outcome ensued with housing. The builder sentiment index posted a score of 59 (50 is neutral), the highest score since 2005. In the past, the index has been a very good clue to the new housing starts data usually released the next day, but recently that reliability has been slipping, with the August report being a glaring miss indeed: Starts fell over 11% in August (unadjusted – the adjusted decline was even larger), and permits fell 5.6%. Now, the starts figures are subject to bigger revisions than industrial production, and the total was still up 8% from August 2013. But the improvement was all in multi-family – in the initial estimate, single family starts are down a few percent from a year ago. The broader, less-volatile year-to-date figures tell a similar story: single-family up 2.1%, total up 8%.

The story in housing is similar to manufacturing – a lot of smaller firms have disappeared, having either folded up or been acquired. It ought to be said as well that multi-family business is nothing to sneeze at. No matter which data you look at, though, the pace of new home construction and sales is growing modestly at best and at a rate well below the post-WWII average, let alone the bubble years. As I’ve often written, housing just isn’t going to boom this cycle, even though some of the homebuilders will still make some good money. Next week brings most of the rest of the housing picture, with existing home sales Monday, the Federal agency price index Tuesday and new home sales Wednesday.

That housing data weighed upon the leading indicators figure, a statistic I don’t mention too often because I believe its value has declined. The largest component is the yield curve, which is artificially positive due to Fed policy. I’m not sure the drop to +0.2% in August (from 0.7% in July) means very much either.

Two of the most widely-followed regional manufacturing surveys are the New York and Philadelphia Federal Reserve surveys, and their results dovetailed with the national survey and subsequent industrial production miss. The New York reading soared from an already-high (for New York) 14.7 to 27.5, the highest in about five years. Philadelphia reported a result of 22.5, not as high as the 28 reading in August (a multi-year high), but still quite elevated. Neither regional survey had any correlation with the national production numbers. That might only represent regional differences, of course, and such divergences aren’t new – but not on that scale. I’ll repeat an observation from last month – the six-month outlook in last month’s Philadelphia survey hit a peak, and it has a history as an outstanding contrarian indicator. I would look for the Philadelphia current-activity responses to keep declining through the rest of the year. What that would mean for actual production isn’t clear, but it could be a drag on equity market sentiment.

One area that strengthens the Fed’s no-action stance was the latest inflation data. It was weak, weak enough at the consumer level to alarm some bond veterans. That may partly be a reflection of softer oil prices, whose effects tend to show up everywhere. Producer price inflation was unchanged for the month and is at 1.8% over the past year, but consumer inflation slipped by two-tenths (-0.2%) and the year-on-year rate tumbled from 2.0% to 1.7%. The “core” rate, which excludes food and energy, was unchanged for the month, the first time in four years.

The Federal Reserve staff lowered its economic forecast yet again, this time for 2014 and 2015. I’m not sure whether that’s good or bad. Here’s a chart (via Deutsche Bank (DB) and Twitter (TWTR)) that presents a nice graphical picture of the staff’s persistently over-optimistic record.

Maybe this time the forecast is being cut to “can’t miss” levels. Or maybe they can’t miss forever. Perhaps I’m being too snide, but the staff’s record has been abysmal. I can’t help but wonder if the numbers aren’t at least partly being directed from above, a frequent accusation in the world of government data, though one I almost never pay heed.

Weekly jobless claims appeared to have declined sharply, but it’s really a seasonal-adjustment effect. Claims during the two-week period used for the monthly jobs report were exactly in line with the year-to-date trend.

Chinese monthly data on industrial production (+6.9% year-on-year) and retail sales (+11.9%) came in below estimates, in particular the production data. I don’t take Chinese data at face value except on a directional basis, but that production drop is trying to tell us something.

Besides the housing data, next week’s highlights are the durable goods report for August on Thursday, followed by the final stand-alone revision to second-quarter GDP on Friday. It’ll be accompanied by the latest NIPA estimate of corporate profits, which declined in the first quarter. Fed regional reports include Chicago on Monday (national activity), the Richmond survey on Tuesday and the Kansas City edition on Thursday.