June 23, 2015

Dave Altig, Executive Vice President and Research Director, Federal Reserve Bank of Atlanta

Amy Hennessey: Good afternoon and welcome to the Federal Reserve Bank of Atlanta's ECONversations. We are pleased you could join us today. ECONversations is a series of lectures and webcasts that provide high-level updates of economic and policy issues affecting our region. I'm Amy Hennessey with the Atlanta Fed's Public Affairs Department, and I'll be your moderator. Now, it gives me pleasure to introduce Dave Altig, executive vice president and research director of the Atlanta Fed.

Dave Altig: Thanks, Amy. Thanks, everyone joining us out there. As we meet today, we are one week out from an FOMC [Federal Open Market Committee] meeting, which really didn't answer any questions that you might have really wanted answered, I don't think. I doubt it was really much of a surprise, but the key issue of when will the first increase in policy rates happen was not answered. Fortunately, we don't have to speculate too much about what comes next. It was a meeting where the chair had a press conference and provided us with this information—really, pretty much the definition of data dependence.

When is liftoff going to occur? Well, it's going to depend on incoming data and how that lines up with the assessment of an economic outlook. So that's what I'm going to talk about today. I'm going to talk about our economic outlook. This of course won't be the outlook of the Federal Reserve System or the Federal Open Market Committee. It represents how we're thinking about things here at the Atlanta Fed. There are some pretty interesting questions that remain and I want to spend the time today talking about those.

So, here are two pieces of information. Our current forecast, in green in this table—and for reference, the last time we had one of these ECONversations was last November, so I put up there what our forecast for 2015 was the last time we talked. I put up two columns there: one column is the total 2015 growth rate and I thought it would be interesting to also show you the third-quarter-over-third-quarter forecast, then and now. Third quarter might be informative because it includes September and I think that's probably one of the dates people are thinking about when they're thinking about maybe that incoming data will line up with the outlook. Maybe not, but that gives you some sense of the forecast and the data we expect to emerge over the course of the year.

The most notable thing about this table is that it looks like right now, 2015, will be a substantially slower year in growth than we were expecting in December, and that's really what I want to focus on. Despite the fact that we now expect 2015 to grow at a full percentage point slower than we were thinking at the beginning of the year, we're still pretty optimistic about the year and still assessing our view of the economy in very much the same terms we were thinking back in December. How can that be?

Well, here's the key picture. The key picture, of course, is that red bar there. That's the first-quarter GDP [gross domestic product] growth, which, as we all know, as of the last GDP report we had for the first quarter, was telling us the economy had shrunk by about seven-tenths of a percentage point. Now, note that we're on the eve of the next revision in the first-quarter report—the final revision, at least final for a while—and we are expecting that number to be revised to something near zero, rather than the -0.7 percent that we've got.

So that's better, and in fact that forecast that I showed you on the previous slide assumes zero growth in the first quarter will be the number we're actually dealing with. But that's still a far cry from what we were expecting at the beginning of the year, which is 3 percent as a whole.

So I'm going to deconstruct this first-quarter number a little bit for you and tell you about our thinking about why that number was as ugly looking as it was, and then kind of circle back and say why we're still pretty optimistic about the second half of the year, and in fact have written in a growth rate near that 3 percent we're looking for during the entire year to reemerge as we get to the latter half of this year.

So I'm going to give you four excuses for why the data was bad and our story is right. The first one has to do with what normally we think of as an unambiguous good thing, the fact that the United States is increasingly in the game as an energy producer. I think one of the things we've discovered over the course of the past year is that in the near term, in any event, of being a big producer as well as a big consumer is a bit of a mixed blessing.

Here's a chart that I think sort of indicates what it is that has changed in the environment, that is so important all of a sudden, when we're talking about the price of oil, for example. This is a picture basically of the percentage of total investment in structures and equipment as accounted for by investment in mining and energy-related capital. The key point of this graph is that over the past 15 years or so, essentially the importance of energy investment in the United States has tripled. Which is, over a very short period of time, an enormous change. That enormous change has had consequences, as we've seen, since basically the middle of last summer, near this time last year, the price of oil plummet so precipitously.

What we've done is run this through a variety of models and analysis that we have to try to figure out what sort of impact we would have expected—the decline in the price of energy, the decline in the price of oil—to have had on the GDP numbers particularly as we finished out last year and entered this year. I'm going to show you a series of pictures like this. This is just like the first one that I showed you, with the first-quarter performance highlighted in red. You can barely see it there, but there's a little green bar there that indicates where we think first-quarter GDP growth would have been had the price of oil stayed at the levels it was at last summer. So the bottom line of this particular chart, this particular piece of analysis, is that according to our estimates, the decline in the price of oil that we've seen since the middle of last year basically took about six-tenths of a percentage point off of growth.

So, some of that weak performance of the first quarter was actually somewhat structural, having to do with the effect of lower prices on energy-related investment. At the end here, I'll talk about whether or not we should completely dismiss that as a one-off or whether this is going to be a more persistent issue going forward.

The second reason we can point to for first-quarter growth is along with that precipitous decline in the price of energy since last summer, we've seen a precipitous appreciation in the value of the dollar—none of which was really expected at this time last year, which would really have been the big part of the economic story over the course of the last 12 months.

There's a picture on your left of the appreciation of the dollar and, although we've seen some reversal of that recently, you can see that we're well above the levels of last summer. So these effects are still with us. On the right side, you can see one of the pictures that really sings to the impact in the near term that this has had on the economy by virtue of the effect on manufacturing exports. You can see an index of export orders, sort of a forward-looking measure of investment in export-related items. Along with that long run-up in the value of the dollar, you see the decline in that export order activity. We would look at various sorts of measures of manufacturing-related activities or data and see pretty much the same picture.

So, this clearly had some sort of impact and if we go back to our adding up of what effect we think that has had, this is a pretty big one. Our estimates suggest to us that the dollar influence accounts for about one-and-three-quarters percentage points off of growth in the first quarter. Again, the question is whether or not that is likely to be something that goes away and fades away going forward or whether this will be a persistent source of weakness during the course of the year.

Now we get to things that are a bit easier to dismiss than structural developments like changes in the price of oil or changes in the value of the dollar. That is, in fact, it was a pretty rough winter. It wasn't just a rough winter in Boston, Massachusetts. One of the ways that we gauge the magnitude of weather effects is we think about heating and cooling days, as measured by the National Oceanic Atmospheric Administration or whatever it is. It turns out that in seven of nine of their census districts that they measure such things in, they had a significantly cooler-than-normal winter. We're able to exploit that by sort of thinking about what normal weather looks like in an economic context. When we go through that exercise, we believe that the weather effects probably accounted for about seven-tenths of a percentage point off of growth in the first quarter as well.

So you can sort of see the game I'm playing here. The question is, when we were surprised by the negative first quarter, do we have explanations for what that surprise was all about, why we were surprised, and how much of the shortfall from what we are expecting can we account for? Then of course the question is, what does that mean?

You can begin to see, we're really beginning to add up a lot of effects here and there's one more that I want to talk about and mention—that is, "What we were seeing in the first quarter was all a big illusion, it was all a big measurement problem. It all had to do with the fact that we never see the data that actually exist in the world. When we see GDP reports, we of course always see seasonally adjusted data." Basically, everybody knows what that's about. We know that with some regularity, economic activity is going to be very robust in the fourth quarter, we know it's going to fall off in the first quarter. As audiences I talk to sometimes point out, "We know everyone's going to be broke after Christmas, so we're going to get less spending and less activity following the big Christmas season." So the Bureau of Economic Analysis just makes an automatic adjustment for that. They lop off a little bit of that growth in the fourth quarter and then they throw a little bit into the first quarter to sort of smooth out these regular and economically not-meaningful fluctuations in economic activity.

Basically here is a view that we have been doing that wrong. That view is supported by this picture, and what you should focus on are those orange-yellow bars, which give you over the course of the recovery, going back to 2010 through the end of last year, the average growth rate by quarter for the last five years. What you can see is the first quarter regularly has been quite a lot lower than the other three quarters of the year. In fact, the first quarter has averaged over this time frame just under 1 percent. We've been closer to that 3 percent that we always seem to be looking for in the last three quarters of the year.

There could be lots of explanations for that, but one explanation is when they do this seasonal adjustment, they just don't do it very well. I would feel a bit embarrassed about relying on this explanation were it not for the fact that the Bureau of Economic Analysis [BEA] and the Commerce Department, under which the BEA resides, seem to think the same thing. So, actually when we get into July, what we're going to see is a revision in some of the seasonal-adjusted methodology that they've been deploying—on how they treat government spending, on how they treat investment, and how they treat some forms of consumer spending, which would presumably begin to fix some of that "regular irregularity" in the quarterly data. No one really quite knows exactly what that's going to do to the reported numbers when we see them but we have a guess for that, too. And our guess for that is about a half percentage point of the slow growth in the first quarter might be attributed to this mismeasurement problem, the fact that these seasonal adjustment factors have been kind of messed up for a while.

Where did we get the half a percent, and how confident am I that that's going to be the number? Well, [to answer the] second question—not confident at all. This is basically the lower end of most estimates of what that adjustment is going to look like. The San Francisco Fed has been notable lately for doing some work on this, and they have a number that is actually quite a bit higher than this—closer to 1.5 percent rather than a half-percentage percent. This is a conservative guess, and I think one that's pretty reasonable.

So if you take all those excuses and you think they are okay to invoke, that basically suggests that we'd be adding a substantial part to growth, had-not mismeasurement problems: bad weather, energy price developments, exchange rate movements, had they not occurred, we would have been exactly right in our forecast.

Now the question of course is, how easy is it to rely on those explanations as a reason to be somewhat sanguine or optimistic—the other way of thinking about it—about the balance of the year? So here are the key questions. If you're talking about weather effects, that's clearly temporary. Those are not things you can predict. The effects of them tend to get paid back in later quarters—that is, although we had some transitory softness in the first quarter from weather, [we have] every reason to believe we'll get at least some of that back in the second quarter, maybe into the third quarter. So, easy enough to take that 0.6 percentage point that we got off of growth from weather and say we're going to add that back in for the rest of the year.

Seasonal adjustment problems are a little trickier. What seasonal adjustment does is, of course, it shifts measured activity from one quarter to another, so it's not all that obvious that it affects your annual growth rate. So I wouldn't say that that is going to really alter our view of the year as a whole, but what it does is sort of leaven the observation that the first quarter was so low because we think when all is said and done, the first quarter, because of these measurement problems, will have looked a little stronger than it looks in the data we have in hand right now. And just by the way, one thing people have noticed about these seasonal adjustment problems is that when growth is low in the first quarter, you really don't tend to see it get shifted into the second quarter, you really see it in the third quarter. So it will be a while before we know how big of an impact that's going to be. In any event, the Bureau of Economic Analysis is going to tell us something about that in a month or so.

A little trickier and certainly much more important are these issues that have to do with the dollar and energy prices. The dollar and energy prices are real, structural effects that you can't just explain away. They had a real impact on the economy. The question is, will those impacts continue? The key to our outlook at this point is that the exchange rate appears to have more or less stabilized. Oil prices appear to have more or less stabilized. Those headwinds that really were hitting in a fairly substantial way—so if you think back at our numbers, we were thinking about over 2 percent off of growth—they will eventually disappear if these prices remain stabilized. And so, we expect maybe a little bit of this will last into the second quarter, but by the time we get to the third and fourth quarter, our story is, once again—we're going to see this 3 percent growth, things are going to be looking good as we get into fall, and we'll see what's going to happen next. I think I'll stop there and turn it back over to you.

Hennessy: Thanks, Dave. We really appreciate it. We do have some questions from the audience. Laura Lampron wants to know why bad weather negatively impacts growth.

Altig: Well, there could be lots of reasons. One reason is that people would normally go out, they would go to restaurants or they would take more shopping trips. It's cold, it's snowy—you stay inside.

Hennessy: You kind of hibernate.

Altig: Yeah, you kind of hibernate. Exactly! Another reason might be that transportation is disrupted in a substantial way. We really saw it last winter, when we had a similar story in the first quarter. So those are sort of the mechanisms by which you get the slowdown, and it's another reason why you think some of it is going to come back. It won't be all lost. If all you did was differ your purchase of a new pair of shoes from January into March, the spending still eventually takes place.

Hennessy: Okay, thank you. This next audience member asks, "Since last week's meeting, both Williams and Powell have come out and stated they are in the two-hike camp. Where is President Lockhart and the Atlanta Fed?"

Altig: Well, you would have to ask President Lockhart that question. In our forecast, we don't have such a precise calibration of how, whether it's one move or two moves. We have a very gradual path out of the year. Probably in the models it doesn't line up exactly in a discrete manner—here's a meeting, there's a meeting, meeting move, etc. If you were to look at the summary of economic projections and look at the median path from those, what we have built into our forecast is pretty much right on that.

Hennessy: Thanks. So how has your outlook for inflation changed, if at all, and what kind of PCE [personal consumption expenditures] inflation do you expect?

Altig: There's a sense in which our outlook for inflation hasn't changed. We are still expecting something a little bit north of 1.5 percent for 2015. I think what has changed is our view of the risk to realizing numbers that are softer than that. I think probably that the last time we talked, there was some concern that the trend in inflation may be softening, that we were really not on a track to persistently move in the direction of that 2 percent objective, and of course energy developments were contributing in a large way to those concerns. I think we're at the point now where we believe—particularly with prices of commodities like gasoline, like oil—stabilizing, we begin to see a reemergence of numbers that are consistent with our baseline forecast that over a series of years we will be able to continuously move towards the 2 percent. So our base benchmark forecast is about the same, but I think we have more confidence in it than we did.

Hennessy: How will increased rates help to improve GDP?

Altig: Well, here you always have to be careful, so I really am going to talk the way we think about it on staff here. We're not really thinking of it as increased rates improving GDP, we're thinking about it as a normalization process, that we're trying to do is align policy rates with the underlying state of the economy. Many people have said this in many different venues and I think I would agree with it. Again, speaking for myself, zero is a rate you associate with very low growth, with emergency measures that require a lot of stimulative monetary policy to support growth. I think our forecast is one where we're moving away from that situation. So, normalization in my mind is all about getting back to normal—that is, not so much boosting GDP but rather aligning policy with the realities of GDP and of employment growth, inflation, etc.

Hennessy: Thanks. Do you believe that the downside in oil prices has already been reached? You had mentioned that it has stabilized.

Altig: Yeah. In our way of thinking, we've run most of the negative effects out of us. We're expecting stronger—and we're seeing in the incoming data that we've got—some strengthening of the investment picture. As of right now, I sort of skipped over our GDPnow chart at the end, but if you disentangle our "nowcast" exercise, we still have a negative effect on structures growth, but it's much less negative than it was in the first quarter. And if these prices remain stabilized, I think by the time we get to the third and into the fourth quarter, we will have worked through the effect that those energy price declines have had.

Hennessy: So, should there be a Greek exit, and we know that that's not decided...

Altig: You know I'm not going answer that [chuckles].

Hennessy: ....but from an impact on the European Union, would there be any way, or how would it find its way into our economy?

Altig: Yeah, so that's the trillions and trillions of dollars question. I know nothing more than what everyone else can see when they turn on the news every morning. It's an ongoing saga, it's going to be an ongoing saga, even as they get through the current round of negotiations. There's going to be more as we move into the fall. I'll make two simple observations. One is that the European Central Bank and all the parties involved appear to have been managing their way through it to this point. Our benchmark assumption at this point is they will continue to do that. But it remains—and this is a second point—a downside risk to the U.S. economy. If things turn negative very quickly, we will feel the effects.

Hennessy: So, we've talked a lot about a stronger dollar, and this audience member wants to know why it has such a negative effect in the U.S., and they are implying that it might not have as much, or any at all, in China or developing economies. But is that necessarily so?

Altig: No, that's really not necessarily so. You often hear and may hear again a lot of consternation among emerging economies as the Fed begins the process of normalization and that will have an impact on their currency and it will have an impact on capital—the related issue of capital flows in and out of their countries. I think we've gotten to a state in global economic history where no one is really immune from these sorts of currency fluctuations and the underlying phenomena that create them. The mechanism by which it matters is pretty straightforward in many ways. It affects, again, how much money in financial flows is coming in and out of your country, and affects whether your exports are relatively cheap or relatively expensive. For many, many years, as a budding economist, it was the sort of thing U.S. economists didn't think too much about. Those days are over.

Hennessy: We have to stay on top of it.

At this point, we will be happy to provide answers via e-mail to the additional questions that we'll be unable to answer because we are basically near the end of our session. I thank you, Dave, for joining us, and I thank all of you for joining us today. We hope that you've found Dave's information very valuable. I know that I have found his information very valuable. And we look forward to you joining us for our next ECONVersation in September.

Remember, in the meantime, that you can always find banking, economic, and policy information on our website at frbatlanta.org. Thank you for joining us and have a good afternoon.