by Brian Pretti CFA, Contrary Investor. , 2008
Yeah, that's really the story. And I'm not so sure it’s even that, especially set against the context of an economy very heavily dependent on personal consumption expenditures in terms of GDP growth. I thought it a bit amusing this week to see the media pundit’s credit "better than expected retail sales" for the rally in equity market averages mid-week. C'mon, it had nothing to do with the retail sales number as, of course, it was an options expiration week. Not even a mention of that fact at all in the press, which would have been much closer to the truth than not. But since the number hit the Street and was supposedly better than it could have been, it’s time for a quick reality check in the land of retail sales trends. Why now? Because I'm seeing a few trends in subcomponents of the retail sales report we have not seen since the last recession. And this was what investors were apparently celebrating Wednesday? Any rationale for a frat party, right?
To the point, have a look at the as of January year over year trends in the components of the retail sales report. They tell one big story.
|Component of Retail Sales||Year/Year Change|
|Motor Vehicles and Parts||(0.1)|
|Furniture and Home Furnishings||(4.3)|
|Electronics and Appliances||(1.6)|
|Food and Beverage||5.7|
|Food Services and Dining Establishments||4.4|
The first and most simple observation is that strength was seen in the essentials, if you will, as opposed to the discretionary components of retail sales in total. This is not a picture of a US consumer spending wildly on discretionary items. Far from it. The year over year drop in furniture/home furnishings sales as of January, although not hard to understand at all in light of conditions in the residential housing sector, was the largest year over year decline seen since the last recession. Same deal goes for electronics and appliance sales. Building materials sales, although not too nice at all year over year, are simply self-explanatory. Finally, the retail clothing sales number was the worst year over year decline seen since April of 2003. If you ask me, this report speaks to a noticeably weakening US consumer. Hardly a picture of the character of consumption over which investors would cheer. And lest you think these are one off numbers or specific to the calendar rhythm of January results, that's not the case at all. Many of these components of retails sales have been in annual rate of change decline for many months now. These are trends, and they are deteriorating as of January. Of course the standout number in this report is the jump in year over year gasoline sales. If that's what's supporting the headline retail sales number (which it was), we've got a problem. The last time I checked, rising gasoline prices are inflation in action. So, at least in January, headline retail sales strength was driven by energy price inflation. That's great news, isn't it?
On a month over month basis in January, retail sales growth stripped of the influence of auto and gasoline sales was literally flat. I guess that's an improvement from being down, right? As usual, I personally like to look at longer term retail sales trends excluding what are often volatile monthly auto and gasoline sales. And what do we see? The following table documents the year over year growth rate in non-auto and gas retail sales going back to the end of the last recession. Is the trend of the last two years clear enough for you? Of course it is, unless you are employed as a media pundit.
As the history of these numbers shows us, it has been almost five years since we've seen a growth rate level as low as what was reported for January. In order to attempt to get some sense of rhythm in retail stock performance relative to actual retail sales trends, the bottom portion of the chart depicts the S&P retail index over the same period for which non-auto and gas retail sales trends are shown. From a cycle standpoint, retail stocks bottomed along with the year over year trend in non-auto and gas retail sales back in 2003. So as we look ahead, I believe it’s a decent bet to again look for some type of a bottoming in the rate of change in actual retail sales growth at some point to coincide with a trough in the equities that represent the sector. At least as of now, we have no defined bottom (although that will certainly be told in short term hindsight). Picking peaks in retail equities, at least over the prior cycle, was a bit tougher. It was clear that we were experiencing deterioration in the rate of change in actual retail trends long before the equities peaked as a group early last year. Maybe the best I can say is that actual retail trends gave us plenty of warning regarding a peak in retail stocks, although from the summer of 2006 through early 2007 the general equity market rally of that time swept the retailers along with it, regardless of clearly deteriorating fundamentals.
The last comment about the above chart is just a suggestion to keep an eye on the 50% retracement level for the S&P retail index. As you can see, the 391 level is the 50% price demarcation line from the cycle lows in early 2003 to the most recent peak in early 2007. For the retail stocks to at all suggest better days may lie ahead for both themselves and the consumption driven US economy, this index needs to stay above that perceptually important half way point. A very simple technical demarcation line? You bet. But I believe the inability of this sector index to remain above that level would not bode well for forward retail trends. Certainly the moment of truth in terms of where we rest surrounding this important level is at hand.
So what other corroborating evidence do we monitor as we form our opinions about where US retail sales trends will take us ahead? Really since the beginning of this decade, there has been a very close directional relationship between the year over year twelve month moving average of total retail sales and the year over year quarterly moving average of total US payroll employment. The following chart clearly shows us the very close directional correlation between these two moving averages. And at least at the last rate of change trough, it was the change in trend in payroll employment that led the ultimate reacceleration in retail sales growth. Will it be so again as we look ahead? We'll just have to see, but I believe it’s a very basic and common sense relationship. We watch trends in generic labor market strength.
Next is a very simple look at retail employment trends over the last three and one half decades. The chart depicts the year over year quarterly moving average (smoothes out a lot of bumpy "noise" in the data) of retail employment trends. I've marked prior period recessionary events for your viewing pleasure. As is clear, over this entire period, only once (brief dip in early 2006) have we been at current levels and not gone into official recession. Will we be so lucky in our current circumstances? According to comments from both Bernanke and Paulson on Thursday, they see no recession. Of course these two also told us over and over again that sub prime problems were contained, so their prognostication track records aren't exactly sterling, let alone even close to the mark. Wouldn't retailers be hiring if they were bullish about retail sales trends to come? You'd think so, wouldn't you?
One last chart and I'll call it a day. History tells us that alongside headline employment trends, wages are very important in terms of the forward outlook regarding retail sales and the equities that represent the sector. Personally, I usually focus primarily on service sector wage trends given that the service sector is the largest employer in the US. The following is the year over year change in average hourly earnings for the service sector going back a decade plus. I believe there is a certain directional correlation here, as well as a lead and lag relationship. In my own rough version, I've drawn in the lines that equate wage rhythm peaks and troughs with peaks and valleys in the S&P retail index. In very rough measure, I believe it’s fair to say directional changes in wages leads retail stock performance by maybe a year or so.
But what I cannot account for in the above relationship is the very important intermediary influence of credit, especially related to mortgage credit decade to date. Is credit the reason for the one glaring anomaly in the chart above that is the early 2004 decline in wage growth with no corresponding decline in retail stocks? Sure could be. If what you see above is even close to being correct in terms of defining a certain loose lead and lag relationship, I'd say it’s fair to suggest retail stocks will ultimately bottom after the year over year change in average hourly service sector wages reaches a bottom for the current cycle.
Before signing off, I'll leave you with one last observation I hope is helpful as we move forward that relates directly to the rhythm and tone of retail sales. In monetary easing cycles of the past, the two best equity sector performers once the easing cycle has gotten underway have been financials and consumer discretionary issues. As you know full well at this point, these have been two of the worst sector performers since the Fed began to work its magic last September. So just what is this telling us? Simple, this is no ordinary monetary easing cycle. Something else is happening here. The longer financials and consumer discretionary issues fail to respond to monetary easing, the more we need to question those still looking for some type of a typical monetary easing cycle outcome (improved economy, better tone to the financial markets, etc.). Let's just hope the equity markets in general don't catch on to the fact that something is different in the current easing cycle relative to our experience of the last few decades. Or has the equity market already caught on to the "it’s different this time" pattern currently playing out? Oh well, I guess fair retails can come true after all.
© 2008 Brian Pretti