Wednesday, May 14, 2008

Understanding the Inflation Report

The Bureau of Labor Statistics (BLS) reported today that inflation was contained in April, with prices rising only .2% (this would be about 2.4% on an annualized basis). This was welcome news for the equity markets that responded positively to the report. Sounds good, except we have the usual adjustments and assumptions built into how this number is calculated. The biggest standout in this report is Transportation, which is down by .7 percent in April. Lets take a look at this category to try and understand how these numbers work.

Here is the report’s discussion of why Transportation prices fell by .7% in April:

The transportation index declined 0.7 percent in April, reflecting a 2.0 percent decrease in the index for gasoline. The index for new vehicles declined 0.2 percent and was 1.3 percent lower than in April 2007. The index for used cars and trucks declined 0.3 percent in April, but was 1.8 percent higher than a year ago. The index for public transportation declined 0.4 percent in April, reflecting a 0.5 percent decrease in the index for airline fares. (Prior to seasonal adjustment, airline fares rose 0.9 percent and were 10.1 percent higher than a year ago.) (Gasoline prices rose 5.6 percent in April. Compared to a year ago, these prices were up 20.9 percent. Gasoline prices increase seasonally during the first five months of the year, with the largest increases occurring in March and April and decline seasonally for the remainder of the year).

Taking a closer look at this part of the report, it states that the gasoline price index is down 2 percent in April. This would probably surprise most drivers. In fact, gasoline prices were up 5.6 percent in the month, so how can they be down 2 percent? Well, the government expects gasoline prices to go up 7 percent in April as people drive more, and then go down later in the year as people drive less. Because prices only rose 5 percent, the seasonally adjusted price index fell 2 percent. This assumes that gas prices will, as the report states, decline seasonally from June through December. We know, however, that oil prices have risen for future delivery and gasoline prices are expected to go up, not down. Based on the seasonal adjustment figures, prices should go up about 3.5 percent in May, then drop for the rest of the year. So should the gasoline number be seasonally adjusted downward by 5.7 % even though oil futures are up? I suppose what that means is that future inflation numbers will be worse.

The next subcategory is new vehicles. According to the report new car prices fell .2 percent (the seasonal adjustment for this category is .22 percent, so prices must have fallen by approximately .4 percent). This is interesting, especially when Toyota just announced price increases for their new vehicles in the U.S. These increases will reportedly become effective later in May, so this price decline will likely also reverse next month or in June. That’s bad news because based on the seasonal adjustments, car prices are supposed to go down .32 percent in May and .33 percent in June, so once adjusted this category could look even worse.

The drop in the price of used cars and trucks is not surprising given that we are in the midst of a shift to more fuel efficient vehicles. Used vehicles, which would tend to be high fuel consumption vehicles, are declining in value. If a person goes to buy a new car, don’t they usually trade in the old one? So if the value of the old one declined, the actual cost of a new car, on a net basis, has not declined by .2 percent because the buyer will get less credit for the trade-in. So this is good news for used car buyers but not as good for people who buy new cars.

Finally, airline fares declined! That’s news to me. Before seasonal adjustments they were up .9 percent. So if the reported number fell .4 percent, we know the expected seasonal rise must have been 1.3 percent (there are rounding issues – the actual adjustment is 1.44 percent). Because they rose only .9 percent, on a seasonally adjusted basis they were down .4 percent! What about the fact that the Easter holiday was in March this year but April last year? The seasonal adjustment is down a bit from last year, from 1.95 percent to 1.44 percent, but so is the March adjustment, from 1.23 percent to 0.97 percent. These changes do not reflect adjustment for this point as far as I can tell. Airline prices may reflect this increasing a seasonally adjusted 3% in March and decreasing a seasonally adjusted .4 percent in April.

Seasonal adjustments do make sense to me on a normal ongoing basis, but when we have better information (such as futures prices for oil) perhaps we should take that information into account. Then again, there is a lot to be said for consistency, so making adjustments to adjustments may not be such a great idea. Finding the real story, however, requires work.

In the unlikely event that you would like to look at other categories in this way you can find a lot of information at the BLS. The way to do it is look at the seasonal adjustment table for 2008. The expected change is the percentage change in any given category from one month to the next. If you get a 1 percent increase, then an actual increase of 1% will be reported as a zero percent increase. If you are a statistician and want to get into the calculation of the seasonal adjustments you can start with this report.

Sphere: Related Content

No comments: