Tim Graf (TG): This is Street Signals, a weekly conversation about markets and macro brought to you by State Street Markets. I'm your host, Tim Graf, head of macro strategy for Europe. Each week, we talk about the latest insights from our award-winning research, as well as the current thinking from our strategists, traders, business leaders, clients and other experts from financial markets. If you listen to us and like what you're hearing, please subscribe, leave us a good review, get in touch, it all helps us to improve what we offer. With that, here's what's on our minds this week.

The US government has shut down for the sixth time in the last 30 years. By the time we recorded this, the current shutdown was the fourth longest on record. And if prediction markets are to be believed, it's likely to be one of the three longest, or it could be all over by the time you actually hear this. So very little of this week's episode focuses on the causes, the shutdown, its current state of play, or its potential political fallout. Again, this might all be over in a couple of days. Instead, we spend our time thinking about the economic impacts, both in terms of a US economy that was already showing signs of slowdown, as well as in how to assess the state of the economy in the absence of public data. We have Noel Dixon from our team in Boston on hand to focus on some of those elements. But we also focus on the bigger picture questions of data quality, data availability and what to look for in alternative data series as complements to the official data when the official data are unavailable.

We talked a little bit about these data sets a few weeks ago with Simona Mocuta from State Street Investment Management. But these questions are even more poignant today with the release of an important US labor market data report now delayed and the timeliness of an upcoming CPI release next week potentially in question. The work we do with PriceStats takes on all the more importance when the public data cannot be seen. So we're thrilled to welcome back Alberto Cavallo, the founding partner of PriceStats from Harvard Business School, and Michael Metcalfe, our global head of macro strategy for a broader discussion of alternative data sets and the trends we currently see.

So Noel, starting with you, this shutdown has gone on now for about a week as of the time of recording.

How would you compare this to previous shutdown periods? I'm thinking here in terms of the political dynamics at play as well as the public response to it.

Noel Dixon (ND): Well, 750,000 employees got furloughed as a result of this. They're not really too concerned at the moment, but that probably will change this Friday when they missed their first paycheck. And I suspect that the pressure will increase even more when at the end of the month when mortgage payments, car payments, so on and so forth are due. So I think the pressure will start to mount as we get through this first paycheck cycle for the individuals.

I think from the markets perspective, it's not really impacting the economy overall that much. On average, every week that passes, it might impact the economy, maybe GDP 10 basis points annualized. The GDP NowCast, we're looking for 3.8 percent based on the latest reading for GDP for Q3. So given where we're at in the economy, I don't think that'll, at least from the, as far as the market's concerned is a dramatic impact, but on an individual basis, it's going to be a big deal.

TG: Well, Alberto, I wanted to bring you in at this point to talk a little bit about the data itself that has gone missing as a consequence of the government shutdown.

And I appreciate this may be a bit of a review for many of our listeners, but in the absence of robust public data, could you give people the context of how work like PriceStats can be useful to their process and complement the official data?

Alberto Cavallo (AC): Yes. So PriceStats has been measuring inflation in high frequency on a daily basis for the past 15 years in more than 25 countries. There's a lot of experience that essentially provides, I always tell people, three benefits. There's a benefit of the frequency with which we do this very quickly. Also you get daily inflation and gets updated with only a three-day lag.

There's a lot of granularity that allows us to look deeply into the data. For example, I'm doing some work now to try to understand how the tariffs are pushing prices up or not. That's something we can only do because of that granularity that we have in the PriceStats series. Obviously, the fact that this is an independent series and moments like these, we can fill in some gaps in information. We always emphasize we are a good complement to official statistics, but when there is trouble with the publication of the official data, that independence also provides value and we can show what's happening at that moment in time.

TG: Well, speaking of this moment in time, we have a CPI release next week that may or may not happen, depending on what happens with the shutdown. But, of course, PriceStats can give us some context on what we might expect for the number.

Can you talk a little bit about the trends you're seeing at the moment that might drive this particular data release?

AC: Sure. So, our main index, which is not seasonally adjusted and it does not include shelter, is showing an increase for the month of September of about 0.27 percent. So, it's slightly higher than what we had last month. On an annual basis, it's about 2.8 percent annual rate. And again, higher than what we saw last month.

You can augment our series with some rents, with some utilities and those numbers rise to about 0.36 percent on a monthly basis and the annual rate is slightly above 3 percent. So, it's pointing to a gradual increase, which is again consistent with one of my more academic analysis of the tariffs using PriceStats microdata as actually shown. And it's certainly suggesting that inflation is gradually going up. It's nothing sudden or extreme, but a worrisome increase nonetheless.

TG: Michael, I wanted to bring you in here at this point.

Can you put this most recent observation maybe in the context of the broader inflationary trends the last, say, six months or so? And specifically, I'm wondering, what has the US inflation experience overall looked like, both in the era of much higher tariffs and barriers to trade? And how does the message we get this September make you any more or less concerned about those trends?

Michael Metcalfe (MM): Well, as Alberto said, it's kind of a continuation of the trend that we saw last month, but it is getting a little bit firmer. And obviously, the context when you're going into Q4 is that you seasonally expect prices to kind of to trail off a little bit and for them to be much softer. So that's something to watch very carefully as we go into Q4 as to whether we're going to get the normal seasonal discounting in this sort of slightly different inflationary era. But the one thing I would just say backing up a little bit is that PriceStats did a very nice job, and this obviously is when the official data was still coming out, of highlighting that earlier in the year, inflation was actually going to be a bit softer than expected. And so we caught quite a few of those downside surprises in core CPI.

But then something sort of shifted around June time, whereby the inflationary process just got a little bit firmer. And the way we've been characterizing it is it's sort of firm, but not yet frightening because of course, most forecasts, as the Fed included, expect the annual inflation rate to rise into year end. And I would say right now, we're still on track for inflation to go up towards where those forecasts are. There isn't really a need to change those forecasts based on the current trend. But just a kind of a warning is that the current trend doesn't start to look more like the usual seasonal pattern. That may yet change. So we're not relaxed about inflation yet by any means.

TG: Alberto, you mentioned the microdata when you were describing PriceStats. And one of the things I know clients love to hear about from you is what's actually driving inflation, kind of looking under the hood. And during the early days of this trade war we've mentioned, we got a lot of attention on the sector level data that PriceStats produces for sectors, especially with high import content. And the last time you were on, actually, we talked about the new Country of Origin import price series that you developed.

Can you take the trends that Mike has talked about and maybe add some detail as far as country level data, sector level data, imports versus domestic prices? What's driving PriceStats really in recent weeks?

AC: Yeah, so I've been doing a lot of work on understanding tariffs and the impact that the tariffs have been having. And they're definitely putting upward pressure on many goods. It's not true that, you know, the tariffs have not affected prices. They are simply doing it on a gradual basis.

We've seen it particularly in goods coming from China, which is obviously one of the countries that has been, you know, the goods from China are the ones that have gotten some of the highest tariff rates. And our estimates actually, from the academic work, suggest that these tariffs have added about 0.7 percent to the CPI since March. So it's really not trivial. You know, if we didn't have that effect, we would be much closer to the Fed's target otherwise. So I think it's important to bear that in mind.

And it has gradually, like, pretty much like we have seen in the aggregate data, it's gradually pushing prices up and putting some pressures. Our results actually suggest there's still some room to go in terms of pass-through. So we saw initially more pressure on imported goods, which are obviously the ones that are directly affected by the tariffs, but lately, in the analysis we've been doing, we're noticing some spillover effects on domestic goods prices. They are the ones actually, you know, that compete with imported goods, and they are the ones that now seem to be facing more upward pressure, which is something you normally expect to see, some of these spillover happening. It's just that at this particular moment in time, that's where most of the pressure seems to be happening.

TG: I wanted to switch gears a little bit, because inflation data isn't the only type of data that we're missing right now. We've actually already missed, because of the shutdown, the non-farm payroll release for September last week.

Michael, what sort of alternative data sets are out there to look at trends in labor market data, and what are you seeing in their message right now if you're looking at them?

MM: The reality is, there's a whole range of indicators on the labor market that don't necessarily come out from the government. And so, here, I was thinking about a variety of surveys, whether it's surveys of companies' hiring intentions, or in fact, consumers and their view of the labor market through some of the consumer confidence surveys. When we're trying to forecast payrolls, they're the kind of things that people use to do that.

And so now, you're just getting kind of imputed measures of payrolls growth from those indicators. There is, of course, this debate that we've quite often had, but we've had an awful lot this year about the differences in hard versus soft data and the idea that maybe some of the survey results are impacted by the political affiliation of the people filling them in and that kind of thing. So there is a debate about the accuracy of some of the soft data, perhaps. But then that leads you to kind of harder measures. And so that would be something like the ADP Employment Report, for instance, which obviously this month has gotten a lot of attention because of the lack of the official data. Our only concern there would be that fit with payrolls, there are big discrepancies between the series. And certainly recently, it's been a little bit softer than the actual official data. And then you end up with other datasets like Indeed, which has interesting data on job postings that fit very closely in a very kind of smoothed way against the official data on job openings. And actually, it turns out that the Indeed job postings, that it actually does a pretty good job of proxying for implied payroll growth.

There are alternatives and of course, the regional Feds themselves, and the Chicago Fed in particular has done a nice job, for instance, of gathering together all these different indicators and modeling what it would mean for the unemployment rate.

And you put all those things together, you take the picture from Indeed, which shows a continued steady decline in the number of job openings, you dig in a layer of the sector stuff, and you see that the job postings in health related industries, which would be a big part of employment growth this year, they're finally beginning to roll over. And so all of this suggests that the message that we got from payrolls in August and the back revisions continued in September, which is that the labor market continued to soften even before the government shut down, and the potential impact that might have on government jobs. But having said all that, the way in which the Chicago Fed's labor market indicators, which pull all of it together, was still consistent with actually an unemployment rate that may have been unchanged, with a slight bias to an increase of about one-tenth. So it's continued deterioration, but not a rapid deterioration from current levels.

Put that together with what we're seeing from PriceStats, which is firm inflation, but not yet frightening, and think about what this might mean for the Fed. It would be suggest that actually, they began their insurance cuts the last meeting. The alternative data would seem to suggest that they should continue doing that because labor market risks remain. They haven't had a need yet to upwardly revise their inflation view.

TG: Actually, I have one follow-up question to that. You mentioned some of the specific, perhaps idiosyncrasies about some of these alternative data sources.

But I wanted to ask you, kind of holistically, are there persistent gaps or, I don't want to call them discrepancies because these are complementary data sets, but things that we need to keep in mind about using alternative data to assess the overall picture of an economy vis-a-vis the quality of the publicly available data that we can't get our hands on right now.

MM: I think we're incredibly fortunate with PriceStats and having worked with Alberto for more than a decade now, that we have a long, rich history of understanding how the online inflation that PriceStats captures works together with the official data. Obviously, there are periods when it diverges, but generally, it does a very good job of tracking the overall trend. But then, of course, during the pandemic, we saw this rush of alternative data measures to try and measure the economy much faster. The reality is that quite a few series, therefore, don't have the history through the cycle to understand how some of the alternative data measures relate to the official data. So it's really just a question of trying to find the ones with the most history and the most out of sample history, as in fact, PriceStats has. But then it's also trying to figure out how this information all fits together. And I do again, going back to the Chicago Fed, or in fact, actually, the Kansas City Fed does something very similar, where what they're doing is rather than just rely on one labor market indicator, for instance, they're relying on several, and then trying to benchmark where all of them together.

So I think the other thing I would just note is it's important to rely maybe just not on one indicator, but a range of indicators to try and gauge where we're at. Since I mentioned the Kansas City Fed labor market indicator, what I should say, note what they're suggesting is that they're again, reinforcing the message from things like Indeed, which is that the labor market is softening and probably requires a bit of policy stimulus.

TG: Alberto, final question for you on this, and understanding that data quality issues were becoming a problem in the years before this shutdown, it has been an ongoing trend.

And as someone who in your capacity has advised and worked with central banks and government agencies on data collection and presentation, what solutions do you think are out there to ensure a timely or more accurate presentation of data? Really, what is the low-hanging fruit in improving data quality?

AC: It's true that a big issue that has been happening for over a decade is the low response rates in surveys. You know, that's something that's becoming for all statistical agencies around the world. Quite a problem. People are just not willing or able sometimes to respond to surveys as they did in the past. And, you know, there are certainly a lot of things that can be done to try to improve that and continue to improve the data collection. I've always advocated for statistical agencies to try to create their own innovation hubs. So, divisions that, you know, where they can experiment with new data collection methods. Pretty much like we do in the private sector as well. You know, we do it in PriceStats with online data collection, but you can think of alternative ways also to collect surveys.

It's often very hard to do for a statistical agency when they also have to take care of the regular production of, you know, the statistics that we all expect them to do well and to do on time. So, that's why I've always advocated these innovation hubs or innovation divisions that they could create. Unfortunately, also, that means they, you know, need resources to create them, but I do think it's a potential way to create experimental series that they try for a while and then they can decide to incorporate them into regular production or not. I've seen many statistical agencies starting, are starting to do things like this and I'm hopeful that if they get the support they need, we will see more of that happening in the future. I think that's very important to bring back that innovation into statistical agency.

TG: This is a markets-oriented podcast. I always like to tie what we discuss to what it actually means for traders and investors. So, Noel, I'm wondering if you can close us out. At the beginning, we talked about the public response to the shutdown, but the market response so far has been very, very relaxed. We're thinking about it, but volatility is still very low and risky assets are still at the highs.

But is there a point at which markets do start to get more concerned? Where do we look for that?

ND: Every shutdown is unique, so I think that's important to point out. But in 2016, I think after two weeks, I think the market corrected like 3 percent and then ultimately it recovered by the end of the month once things resolved itself. And in every instance, you got a weaker dollar, so I think that would start to see that start to drag on the dollar. You get a bit in gold and I think yields declined as well. So I think that's, those are kind of the key areas where you'll see some activity.

TG: Thanks for listening to Street Signals. Clients can find this podcast and all of our research at our web portal, Insights. There, you'll be able to find all of our latest thinking on markets where we leverage our deep experience in research on investor behavior, inflation, media sentiment and risk, all of which goes into building an award-winning strategy product. Again, if you like what you've heard, please subscribe wherever you get your podcasts and leave us a review. We'll see you next time.