There is good and bad news in today's economic data release: on one hand retail sales in September beat expectations at 1.1%, on expectations of 0.7%, and up from an upward revised 0.3% in August. Retail sales less autos was a modest beat at 0.6% on expectations of 0.3%, although the previous number was revised substantially higher from 0.1% to 0.5%. Yet confirming that the bulk of the "beat" was in auto and associated gas sales, was that Retail Sales ex Autos and Gas (duh) came at 0.5% on expectations of 0.4%. Basically, surging subprime loans to autopurchasers and the resulting increase in gasoline sales was the reason for this "surprise" beat. And as for the bad news, import prices jumped to 0.3% in September, on expectations of -0.4%, a surge from August's revised -0.2%. And while fuel imports had dropped in August -1.4%, in September these jumped to a positive 0.1%, showing just how big the monthly sensitivity to any moves in the energy complex are. In other words, should inflation persist, don't expect for retail sales, which we expect to decline to recent deleveraging at the consumer level, to persist.
For those surprised by how resilient car sales have been here is Goldman Sachs with an explanation:
- Vehicles sales have increased at a healthy pace over the last two years despite a weak household sector and a tepid recovery overall. What explains the steady growth in vehicle sales since the recession ended?
- Based on the composition of sales, the main factor appears to be business investment spending. Vehicle sales are often thought of as an indicator of consumer demand, but companies account for about half of the dollar value of new purchases. Since vehicle sales bottomed, firms have accounted for about 70% of the growth in purchases. The outlook for business vehicle purchases arguably remains bright, mostly due to pent-up demand.
- Over the short term, consumer purchases of vehicles may also remain strong as recent supply-chain disruptions are fully resolved. However, the medium-term outlook appears less favorable than for business sales. Most importantly, unlike in the business segment of the market, there is no obvious pent-up demand for consumer sales—our modeling suggests consumer vehicle holdings are close to equilibrium. Secular trends (population growth, depreciation, etc.) should support consumer sales over time, but cyclical factors (unemployment, credit conditions, and oil prices) may be a persistent headwind.
After falling off during the spring and summer, vehicle sales have picked up recently. Total light weight vehicle sales rose to a seasonally adjusted annualized rate (saar) of 13.1 million units in September, up from 12.1m in August. Part of the improvement reflects a greater availability of Japanese-brand vehicles following supply-chain disruptions earlier this year. However, vehicle sales have also increased at a healthy pace over the last two years—at an annualized rate of 18%—despite a weak household sector and a tepid recovery overall. Given the depth of the recession in the vehicle sector, some bounce back should have been expected. But what fundamentally explains the strong recovery in vehicle purchases?
Based on the composition of sales, the main factor appears to be business investment spending. Vehicle sales are often thought of as an indicator of consumer demand, but companies account for a large share of new purchases—as of Q2, business purchases accounted for 49% of the dollar value of new sales, according to the Department of Commerce (DOC). Each month the DOC calculates the share of unit sales to consumers and to businesses, using state-level data on vehicle registrations. As shown in the exhibit below, since vehicle sales bottomed in February 2009, sales to firms have accounted for about 70% of the growth in new purchases (the breakdown of sales is not yet available for September, so for last month we have assumed the shares were unchanged from August; also note that sales to businesses are broader than “fleet sales”, and include purchases by small business at retail vehicle dealers).
Most of Sales Growth Coming from Business Sales
The outlook for business vehicle purchases arguably remains bright, mostly due to pent-up demand in this segment of the market. Real investment in transportation equipment fell by 68% peak-to-trough during the recession, a far larger drop than for other types of business investment. And despite a strong recovery to date, year-over-year growth in the stock of transportation equipment is still negative, meaning that investment spending over the last year has run below the rate of depreciation (Exhibit 2). Modeling business demand for transportation equipment is challenging (as with most types of nonresidential investment), but a simple model relating the stock of business transportation equipment to real GDP and relative prices suggests transportation capex is probably well below “equilibrium”. Short-term indicators are mixed, with strong heavy truck sales and durable goods orders, but weakness in the timelier business surveys. If the economy enters another recession, business vehicle sales would likely decline too. However, barring that, they should be poised for continued strong growth.
Low Capital Stock Growth Suggests Pent-up Demand
Over the short term, consumer purchases of vehicles may also remain strong as supply-chain-related issues are fully resolved. However, the medium-term outlook appears less favorable than for business sales. We model the stock of vehicles owned by households with four main factors (model details appear at the end of this note):
1. Demographics. Population growth is the single most important determinant of vehicle stocks—more people equals more cars on the road. Gradual growth in the population should continue to support vehicle sales in the future. The age structure of the population also matters: states restrict driving among the young, and driving miles and vehicle ownership eventually fall off for older age groups (according to data from the Federal Highway Administration and the Federal Reserve’s Survey of Consumer Finances, respectively). The age structure of the population is currently a neutral factor for the vehicle sales outlook, but will likely become a negative in 5-10 years as the baby boomers age.
2. Income levels. Income levels also help determine vehicle ownership. For example, in a large sample of countries, GDP per capita explains about 80% of the variation in the number of vehicles per person (World Bank data for 2008). Rising income levels have also supported greater household ownership of vehicles in the US over time, and today 87% of US households own a car or truck (not including leased or employer-provided vehicles).
3. Cost of vehicles and financing. Not surprisingly, car and truck sales are also sensitive to the cost of vehicles. In our modeling, we incorporate four specific variables: 1) the relative price of motor vehicles; 2) the relative price of gasoline; 3) real interest rates; and 4) credit conditions. Lower costs help explain the strong growth in consumer auto purchases in recent years: until recently, new car and truck prices had fallen steadily relative to overall consumer prices, and real interest rates have generally declined as well.
4. Labor market conditions. Lastly, labor market conditions—and particularly the unemployment rate—appear to matter for vehicle sales. These variables may proxy for income expectations, or perhaps for consumer confidence in general.
Given these fundamentals, we see a relatively weak outlook for vehicle sales to consumers over the next 1-2 years. Most importantly, unlike in the business segment of the market, there is no obvious pent-up demand for consumer sales: our model suggests that the stock of consumer vehicles is close to equilibrium at present. Therefore, growth in consumer vehicle sales will require improvement in the underlying fundamentals. Some factors will remain supportive of consumer auto sales, such as population growth, depreciation, gradual per capita GDP growth, and perhaps lower relative prices for cars and trucks. However, cyclical factors—especially unemployment, but also possibly credit conditions and oil prices—may be a persistent headwind.
Our equity research analysts forecast total vehicle sales of about 14.0m by the end of 2012. Based on our broader macroeconomic forecasts, this estimate looks achievable. But we suspect that most of this growth will come from sales to businesses, with the consumer segment of the market growing at a much slower pace.