TRANSCRIPT
Jeremy Glaser: For Morningstar, I'm Jeremy Glaser. I'm joined today by Bob Johnson--he's our director of economic analysis. We're going to answer some reader questions on GDP, retail sales, and on the housing market.
Bob, thanks for joining me.
Bob Johnson: Great to be here today.
Glaser: Bob, I know you get a lot of comments and a lot of emails from readers asking questions, and you want to take some time to answer some of those. One that you've gotten a lot recently is on the Atlanta Fed's GDP forecasting tool. Is that a good way to think about what GDP is going to look like? Do you think it has good predictive power?
Johnson: Well, it's a fascinating tool, and anybody can go look at it. They post the numbers every time there is a new change to the data. It's called GDPNow, and it's done by the Atlanta Fed. They try to estimate what the quarterly GDP is going to be on a year-over-year basis, and they give you all the supporting details of all the different categories as well. So, it's a very interesting document, but keep in mind that this is not a long-term forecasting tool. For the fourth quarter of 2015, they put out their first forecast at the very end of October when the data came out.
So, they gather all the data like the retail-sales report and some of the housing reports and so forth that generally have very good correlations with the GDP calculation and run those. They run them on very big computers, and they can do some very complex things with them. Based on the data that's already there, they can be very accurate.
Now, the problem is they have to take the other months that they don't have and do either some kind of trend line or reversion to mean or some type of methodology that they use to get a forecast for the out months. And as we move through and get more actual data and less of the trend-line forecasting, usually the closer they get to the estimated number. This particular quarter, they started out with estimating 2.5%; at the moment, they are at 0.7% for the fourth-quarter growth--that's December versus the September quarter. Our forecast happens to be a little higher at 1.6%, but certainly their number seems to be in the ballpark to us.
Glaser: So, if you did see a number that was pretty low on Friday when we do get that first estimate of GDP, that wouldn't surprise you.
Johnson: No--because the biggest component in there that makes their number low--and a lot of other people feel the same way--is the inventory adjustment, and that's a very complicated adjustment. We can all look at the reported inventory levels and say, "They're down and, therefore, it's going to be a problem," but they also have to adjust it for something called the "last in, first out" inventory adjustment. And they also have to adjust it, yet again, for prices--and prices are widely varying between groups these days. It isn't as if one number works for all, like it used to be.
So, the calculations by category are very complex, and they've got something that models it pretty well. They are thinking it will take 1% off the GDP forecast. So, the [0.7%], if you add the inventories back, would be more like 1.7%, which is a much more reasonable estimate of where the underlying strength of the economy is. We don't worry too much about inventories. They go up and down by quarter, but at the end of the year, they usually kind of cancel each other out. So, worrying about where those are each quarter is kind of silliness.
Glaser: Let's turn to the housing market. There are some questions about what the strength of the market looks like right now. People are worried that housing is not going to be there to support the economy. What are your thoughts on where housing is today?
Johnson: I think we are doing better now. I think we had a great spring and summer, and then we had a bunch of odd factors creep in during the back half of the year. Certainly, in November, we had some issues with paperwork. New laws went into place where you had to disclose things to the borrower. Of course, everybody up the food chain--the mortgage processor and then the bank itself and then whomever they sold it to--all had to sign off on this process. So, some closings got delayed in this process, and it certainly made November sales look not so hot.
Now, in a lot of the data, we've come roaring back in December, and [yesterday, Jan. 27] we got the report on new-home sales--more than 5.4 million units. It was one of the best two or three reports of the whole recovery. So, that was really great news there, and inventories were up a little bit, so there is more supply to sell in future months. The prices were a little bit more reasonable than they had been so that more people can afford it. All of those were nice positives in that report, and we were very glad to see. I think a lot of people were caught off guard last month because realtors jumped up and down on existing-home sales about this paperwork snafu. But it kind of caught the new-home sales data, too, and we didn't realize it in our thinking. But certainly, this month's report is stellar--as was the existing-homes report. Starts were kind of mediocre, but the permits, which are more forward looking, also looked very, very good.
So, overall, I'd say the housing market continues to move along. I don't expect it to grow any faster or to add any more to GDP than it did in 2015. It's hard--there isn't enough land out there. Prices and affordability are issues. So, it's not going to wildly add more and cause this huge swing but it's going to be a continued adder to GDP growth in 2016.
Glaser: Moving to retail sales, we had a report that looked not so good before it was adjusted for inflation but much better after adjusting for inflation. I know you got a ton of questions about why it was important to adjust retail sales for inflation. Can you address why you need to make that adjustment there?
Johnson: Sure. In GDP, the "P" is production, and you really want to know what inputs in the economy are going into it. If the price of something is falling but you're actually still making more of it, that's really good because GDP measures how many people you use and what resources you are using. So, you really want to be focused on counting unit sales. But retail sales are in nominal dollars. So, we have to go in and adjust those retail sales. They may look like they are down, but they may actually turn out to be up in units. And again, why are those units important? Well, if you are selling a lot of units and they just happen to be at a lower price, you're still going to need just as many clerks, if not more, than you did before; you are going to need more stock people in your back room to move the stuff around; you are going to need more truckers to haul the stuff in; and you are going to need more manufacturers--many of whom may be overseas--to actually produce the goods.
So, there are a lot of components that [inflation] drives in GDP, and there are a lot of reasons why you need to [adjust for it]. We can sit here and argue all day along about substitutions and some of the other effects in the inflation reports and how you average the numbers together and I-know-my-inflation-rate-is-higher arguments or whatever; but when you are actually saying how many TV sets did you sell and how many people did you need to move those around, the inflation adjustment is pretty darn good. It's effective. And again, when considering how it impacts incomes and the overall flow of the economy, you've got to be a little bit more careful maybe; but on a retail-sales basis, it is extremely helpful to adjust it. And again, it's one of those things where, over the years, some people become a little careless when adjusting the retail-sales data and have just said, "Let's adjust it for CPI." But now you've got goods prices going down and services going up, so it may not make sense to use the broad basket CPI data to adjust it anymore. So, that's also an interesting factor.
Glaser: Bob, thanks for addressing these reader questions today.
Johnson: Thank you so much.
Glaser: For Morningstar, I'm Jeremy Glaser. Thanks for watching.
