Tom Webb's Blog (Manheim Consulting)

Vehicle Miles of Travel: no snow means we go

Monthly vehicle miles of travel (VMT) rose by 1.3% in December versus a year ago.  Attribute much of the rise to mild winter weather as evidenced by the biggest increases being reported in the North Central and Northeast regions. (http://www.fhwa.dot.gov/policyinformation/travel_monitoring/tvt.cfm).

Even with December’s gain, VMT for the full year was down 1.2%, and the twelve month rolling average was still 2.5% off the peak reached in November 2007.

A look at annual changes since 1973 shows that something more than a recession and gas prices is going on.  It’s a structural shift. (Additonal comments in 1/24/12 posting.)

The Yield Curve – A forever useless tool?

For more than half a century, the yield curve has been one of the more useful items in the economist’s toolbox.  An inverted yield curve (long-term yields lower than short-term yields) has predicted seven of the last eight recessions, with only one false positive.  A steep yield curve has often been correctly associated with forthcoming inflationary growth, but it has less predictive power than an inverted curve.

The Federal Reserve’s unprecedented policy actions, along with changes in communications strategy, have rendered the yield curve useless in providing insights into the current economic cycle.  For one, the tool’s calibration has been set askew by historically low interest rates.  Indeed, as noted in an earlier post, there were times recently when one couldn’t legitimately say whether the yield curve was becoming steeper or flatter.

To keep interest rates low and to flatten the yield curve the Federal Reserve did things such as changing its communication strategy (e.g., associating a specific time to what was previously known only as an “extended period of time”) and shifting holdings to longer-term maturities (e.g., “operation twist”).  But did the yield curve actually flatten?  It depends how you measure it.  Traditionally, analysts have looked at percentage point differences.  By that measure, the spread fell from a high of 2.8 points in February to 1.7 points today (which is still high).  But, with interest rates so low, the actual ratio between the yield on a 10-year note and a two-year note rose from 4.7 to 1 in February to 8.2 to 1 today.  And, more fundamentally, as succinctly described (along with an example) on the Econompicdata blog site (http://econompicdata.blogspot.com/2012/02/bond-101-duration-risk-at-zero-boundary.html) investors take on duration risk not only to achieve higher yields, but also for possible bond appreciation if rates fall.  Low rates today mean less room to fall, thus, less potential appreciation even though the yield curve is steep.

The Fed has clearly telegraphed that interest rates will remain very low for a very long time.  The most likely scenario is that the economy will enter a recession (even it is a long time from now) well before the yield curve inverts.  Thus, to answer the title question – yes, it has become a useless tool which will not give a heads up as to the next recession.

We may at some point long for the days when Fed policy was opaque and its communications enigmatic.  True, functioning markets – or at least ones you can make money in – require that participants have differing opinions as to the future.  The Federal Reserve policy actions have been removing uncertainty regarding interest rates, but increasing uncertainty as to the broader economy.   

   

                  

Real GDP & Auto Sales

With Friday’s release of real GDP we can now add another year to this interesting graph.  Each dot represents one of the 57 years between 1955 and 2011.  The x-axis shows the annual percent change in real GDP for that year and the y-axis is the percent change in new vehicle sales.  Naturally, there is strong correlation.

Two outliers are noted in the chart above.  At the far right, close to the horizontal axis, we see that in 1966 new vehicle sales declined 2.1% even though real GDP grew by a very strong 6.5%.  Why’s that?  It’s because new vehicle sales had already increased dramatically in each of the prior four years.  (New vehicle sales first exceeded the ten million mark in 1965.)

As an opposite case note that in 1982 new vehicle sales declined by only 1.9% even though real GDP dropped by 1.9%.  (Up until our recent recession, that was the largest decline in real GDP in post World War II history.)  The disconnect is because new vehicle sale had already fallen significantly in 1979, 1980, and 1981, producing a peak to trough swing of 31%.

By just ignoring those two outliers, the R-square increases from .63 to .71 for the remaining 55 years plotted.  If we also ignore 2011, the R-square rises to .74.  As noted in the chart, 2011 was only time that an increase in real GDP of less than 2% was associated with a rise in new vehicle sales.

Strong cycles produce outliers, and after an outlier there is a movement closer to the regression line.  That will occur again this year, although I fully expect that the plot for 2012 will remain well above the regression line.

Also note that, in this comparison, 2009’s plunge in new vehicle sales (-21%) was not abnormal given the 3.5% decline in real GDP.  Indeed, it was almost right on the regression line.

   

Vehicle Miles of Travel: The long road back (updated)

This is an update to a post first made 12/5/11.  As expected, vehicle miles of travel (VMT) declined again in November.  The importance of this statistic to the automotive industry is simple - "vehilce miles of travel" is another way of saying "vehicle consumption".  As noted in the post on the aging of the vehicle stock, vehicles today simply last longer.  They will last a lot longer if they just sit in the garage.

Report details show that the smallest declines in VMT occurred on urban interstate roads.  The biggest declines were for rural roads.

Total vehicle miles of travel (VMT) are still be below the peak reached four years earlier.  Current statistics (updated through November) show a 2.6% decline from the November 2007 record.  And, in recent months, we have been moving further away from the peak; not towards it.

The decline in VMT is greater (and its recovery slower) than either the past recession or higher gas prices would suggest.  The pattern in VMT is, however, in tune with the falloff in employment, which peaked approximately the same time as employment (January 2008 versus November 2007).  With respect to employment, we are even further away from the previous peak (-4.4%).

As with employment, we suspect there are secular, as well as cyclical forces, preventing a quick return to the previous peak.  Shifts in demographics, housing patterns, and working environments have all reduced needed miles of travel.  And, of course, the weak economy continues to restrain discretionary travel.

And consider the recent report that teenagers are less apt to run out and get their driver license the first day they are legally able.  In 1985, 87% of all 19 year olds had a driver license.  By 2008, that percentage had a fallen to 75% - and that was before the recession, which so decimated teenage employment.  In a reversal of phrase, Michael Sivak, of the University of Michigan Transportation Research Institute, said, “…some young people feel that driving interferes with texting…”   

Does an Aging Fleet Equate to Pent-up Demand?

In 2011, the average age of cars and light duty trucks in operation reached a record 10.8 years, according to Polk.  It’s an interesting statistic with implications for new vehicle sales, but it is wrong to consider this simple aging of the fleet as some sort of quantification of pent-up demand.  There is no relationship between the average age of vehicles in operation and future new vehicle sales – either statistically, or theoretically, speaking.

As for the statistics, note that the vehicle parc has aged steadily for decades.  The average age was 6.5 years in 1975, 7.7 years in 1985, 8.4 years in 1995, and 9.5 in 2005.  Those earlier numbers would now be considered a “young” fleet, but that did not stop new vehicle sales from reaching new highs each decade.  Conversely, the average age of the fleet first hit the double-digit mark on July 1^st, 2008.  That did not even mitigate, much less stop, the subsequent plunge in new vehicle sales.

The aging of the fleet is as much engineering-driven, as economics-driven.  Vehicles today simply last longer.  And, with multi-vehicle households, individual vehicles do not rack up miles as fast as they use to.  For example, yours truly owns a 13-year old Chevrolet pickup.  It is not the primary vehicle in the household, but it is more than capable of serving its specific purpose.  And I expect it will continue to do so for many more years.  In no way does this truck represent pent-up demand for a new one, as it will be replaced with a used one.

And that is where we also find the underlying theory wanting.  Much of the increase in the average age of the vehicle parc is a result of the growing share accounted for by very old units.  (In other words, the average age has risen more than the median age)  After 15 years of use, less than 29% of the 1970 model-year vehicles sold were still on the road.  For 1990 model-year vehicles, that percentage rose to 60%.

Pent-up demand for new vehicles comes from vehicles 3 to 6-years old, not those over 10 years old.  And the number of 3 to 6-year-old vehicles in operation will hit a cyclical trough this year.  The older vehicles only have second-order derivative effects in creating new vehicle demand as, for example, the 10-year-old vehicle is traded for a 6-year-old one, and that owner then trades for a 3-year-old one, and that owner then trades for a new one.  We see that effect through used vehicle valuations, and there is no doubt that the run up in wholesale used vehicle prices has been, and will continue to be, beneficial to new vehicle sales.  

But for now, let’s just stick with the mathematically unchallengeable: “new vehicle sales influence the age of vehicles in operation.”  The reverse statement is much less certain.

Future VIO Growth: Slower and More Predictable

The growth rate of vehicles in operation (VIO) declined precipitously during the recession.  In fact, in 2009, vehicle scrappage exceeded new vehicle sales and, thus, there was a decline in the vehicle parc – a post World War II first.  With the improvement in the economy and new vehicle sales, the VIO count has once again begun to rise, but at an annualized pace (below two million) that is far short of the six-plus million annual growth rate posted in 2004, 2005, and 2006.

To assess what part of the slowing in VIO growth is a result of current economic conditions as opposed to longer term changes in demographics and lifestyles, it is useful to segment the growth into root causes.  In the simplest of terms, increases in VIO can be the result of either:

• growth in the number of households,
• growth in the number of households owning a vehicle, or
• growth in the number of vehicles per household.

Given time series for the number of U.S. households, the percentage of households owing at least one vehicle, and the average number of vehicles per household, we can mathematically segment VIO growth into one of the three root causes.  (Technically, there is also the synergistic effect between individual factors, but that can be simply prorated back to one of the three.)

This is what the analysis looks like for the past six decades:

 

Naturally, the share accounted for by an increase in the percentage of households owning a vehicle became miniscule after vehicle ownership became almost ubiquitous.  We also see a general decline in the share accounted for by an increase in the number of vehicles per household.  This was also a reflection of saturation as the number of vehicles per household rapidly approached the number of persons per household - and often exceeded the number of driver licenses at that residence.

In the last decade, there was a slight decline in the share accounted for by simple growth in the number of households.  This was a result of slowing household formation rates as economic conditions produced more “doubling up”.  (When kids move back in with parents, or unrelated persons double up, they bring their vehicles with them.)

It is logically to assume that in coming decades VIO growth will be driven primarily by household formation rates alone.  This will make future increases in the vehicle population both more predictable and less rapid.  Slower VIO growth, coupled with the long-term decline in the scrappage rate, suggests that the underlying rate of demand for new vehicles is far below the 17 million annual sales rate that was achieved during the seven year period between 1999 and 2005.

 

End-of-Term Residuals: From Big Losses to Big Gains

The following statistics are from the Ford Credit Auto Lease Trust 2011-B prospectus filed with the SEC on October, 26, 2011.  These trends are reflective of the leasing industry as a whole, although Ford certainly had some swings which were greater in magnitude than the industry as a whole.  (The data reflects all Ford Credit leases written since 2002.)

End-of-term residuals: from big losses to big gains.  In 2008, Ford Credit experienced an average loss of $3,390 per end-of-term vehicle returned and sold.  In the first six months of this year, that flipped to its mirrored image – a gain of $3,445 per unit.

And, of course, the percentage of end-of-term units that were being returned was highest when the losses were the worst.  Specifically:

	Return Rate
2007	77.4%
2008	82.2%
2009	75.7%
2010	66.0%
6 mos. 2011	57.5%

The change in the return rate meant that Ford Credit sold 7% fewer end-of-term units in 2010 compared to 2008, even though actual terminations rose by 15%.

Losses were a result of the market, not subvention.  The spread between contract residuals and ALG projected residuals remained fairly constant throughout this period.  It was the ALG projected residuals that proved overly optimistic in 2008 and overly pessimistic in 2011.

Outlook: In 2012 and 2013 we expect that the average dollar gains at end-of-term will be smaller (as contract and ALG residuals move up), but we expect return rates will continue to fall.   

 

Narrowing Gross Margins: Should We Worry?

Now that Penske has reported we can update last week’s posting on the sales-weighted gain in used vehicle retail sales achieved by the seven publicly-traded dealership groups.  It was 4%, not the 3% originally noted, since Penske had a better than expected gain.

Nine consecutive quarterly gains is certainly good news, but some are troubled by the erosion in gross margins.  Depicted here is the trend in the sales-weighted retail gross margin for these groups.  A note of caution in interpreting - since this is a sales-weighted average, a shift in sales volume between the groups will distort the true trend since each company has historically different margins that are dependent on sales mix, geographic mix, and even accounting.

Nevertheless, there is no doubt that industry-wide across all franchised dealers that average retail used vehicle gross margins in the third quarter of 2011 were less than in the second quarter and less than a year-ago.  The question is: how worrisome is this trend?

That depends on your theory as to the cause.  My belief is that most of the margin reduction is the result of rational business decisions – not desperation moves to move metal.

For example, there is no denying that more dealers are willing to retail out at a wholesale price rather than actually wholesale the unit.  The rationale is logical – earn doc fees, generate F&I income, allow salespeople at least some commission, create potential service revenue, and possibly garner a customer for life.  Likewise, more dealers are utilizing auction inventory as their own virtual inventory.  If you have already pre-sold a unit, a lower margin is acceptable since you have eliminated both holding costs and risk.

There is one major downside to this trend.  And dealers should be careful.  As the saying goes “gross is a state of mind”, thus, once you start doing low gross deals, you run the risk of all your deals becoming narrower.

 

Same Store Used Retail Unit Sales Continue to Rise

Six of the seven publicly-traded dealership groups have now released third quarter earnings.  (Note: Penske will release on November 2^nd and we have shifted CarMax’s quarterly filing since they report on a fiscal quarter that end 8/31/11).

Based on those filings, these dealers will post a sales-weighted 3% increase in same-store used vehicle retail unit sales for the third quarter.  This is down from a 7.1% gain in the second quarter and double-digit gains in the five quarters preceding that - but after nine consecutive quarterly gains, the comps are getting tough!

These dealerships are also in acquisition mode again.  As such their absolute used unit retail sales volume in 2011 will get close to one million units.  

The Disconnect Between Consumer Confidence and Consumer Spending

Consumer confidence numbers released today by the Conference Board showed a significant decline in October.  Indeed, it moved the Index back to recession lows.  Couple this with European and U.S. debt issues, and we are setting up for a possible bleak holiday selling season.

Economists, however, generally discount consumer confidence measures as simply coincident indicators that add no predictive or explanatory power beyond that provided by hard statistics like jobs and income.  But when you see a disconnect like that shown below, something must be going on.  In this instance, the popular explanation is that high-income households, which account for the bulk of retail sales, have kept on spending and now account for an even larger share of total sales.

That explanation is partly true, and is confirmed by the performance of high-end retailers like Neiman Marcus and card issuers like American Express.  But there are two simpler explanations that are just as important.  First, the retail sales tally is indifferent as to whether the purchase is made with a paycheck, a welfare check, or food stamps.  In the latter two cases (which are a growing share) the buyer was probably not exuding confidence.

Additionally, there is probably more inflation hidden in the total retail sales than many expect.  Food consumed at home has risen 6.3% in cost over the past year, and gasoline is still up 33%.  There is also inflation in all types of medical services (another area where the consumer is often not the direct payer).  

All told, this was a concerning release.  It may not tell us much about future retail sales, but is does point to a level of consumer anxiety that politicians of all stripes will attempt to exploit – and, by doing so, will only instill more uncertainty.