Haver’s data services are provided through the DLX platform; delivering tools for data management, visualization, modeling and publishing. DLX is designed to increase productivity and integrates with Excel, PowerPoint, and major statistical software packages; MatLab, R, Python, SAS, E-views and Stata. DLX component application DLXVG3 offers precision graphics, publishing, analytic tools, and easy navigation for all Haver databases.

icon-chart.png

Chart and compare multiple data series with accompanying data table of values

Advanced functions (growth rates, moving averages, aggregation, seasonal adjustment)

Analytical tools (cycle charts, scatterplots, trend lines, and recession shading)

Complete graph customization (axis scaling, sizing, formatting, styles)

Excel and PowerPoint direct export and complete integration (dynamic spreadsheets and presentations update on a click)

DLXVG3

icon-db.png

DLX Add-ins for all versions of Excel provide instant data updates of entire workbooks

Use DLXVG3 to locate data series codes from within any file, easily creating new spreadsheets for analysis

Edit date ranges, archive data, apply functions

Excel Spreadsheet Add-ins

icon-time.png

Haver offers timely, accurate, and professionally supported data via an update service, meeting the strictest demands of the institutional research community

AutoUpdate provides fast and reliable data updates within seconds of source releases

Timer service runs continuously on an hourly basis when DLX is running

Manual updates offer clients the independence to update data retrievals on any desired schedule

All database updates retrieve only the data series that have changed and send only the data points that have been updated; including revisions

Timely Updates

DLX software tools to convert data into many different formats needed for diverse analytical frameworks

Haver supports a DLX API permitting clients to leverage and build processes with DLX data

Database Compatibility

products-dlx.svg

Data Link Express

Platform Independent site delivering DLX Databases across any workstation or device.

products-haverview.svg

HaverView

[Matlab](http://matlab.com) by The MathWorks lets you access, visualize, and analyze historic time-series data to identify patterns or uncover complex relationships. From within a single environment you can:

- Access data from multiple sources including files, spreadsheets, databases, data providers, and the Web
- Store data in financial time-series objects to simplify data management, data transformation, missing-data handling, and date-math calculations
- Perform technical analysis with moving averages, oscillators, stochastics, and indices
- Create custom analysis routines and visualizations or animations to capture, present, and share your analysis processes and discoveries
- Customize your analysis environment with advanced functionality from signal processing, statistics, or econometrics

products-matlab.svg

MATLAB

[Stata Data Analysis and Statistical Software](https://www.stata.com) by StataCorp LP is an integrated statistical package designed for researchers of all disciplines. It provides the ability to do data analysis, data management and graphics and has features such as multiple imputation, factor variables, GMM, competing-risks regression, state-space modeling, predictive margins, and more.

products-stata.svg

Stata

[SAS](https://www.sas.com) provides an integrated environment for predictive and descriptive modeling, data mining, text analytics, forecasting, optimization, simulation, experimental design and more. From predictive analytics to model deployment and process optimization, SAS provides a range of techniques and processes for the collection, classification, analysis and interpretation of data to reveal patterns, anomalies, key variables and relationships, leading ultimately to new insights and better answers faster.

[SASEHAVR](http://support.sas.com/documentation/cdl/en/etsug/67525/HTML/default/viewer.htm#etsug_sasehavr_gettingstarted01.htm) is a seamless interface between Haver and SAS data processing that enables SAS users to read economic and financial time series data that reside in a Haver Analytics DLX (Data Link Express) database. [SASEHAVR](http://support.sas.com/documentation/cdl/en/etsug/67525/HTML/default/viewer.htm#etsug_sasehavr_gettingstarted01.htm) requires both SAS/ETS and Base SAS Software.

products-sas.svg

[RATS (Regression Analysis of Time Series) Econometric Software](https://estima.com) by Estima is a leading econometrics and time-series analysis software package. It provides ability to do linear and non-linear least squares, forecasting, SUR and ARIMA models and in addition, supports advanced techniques such as GMM, ARCH and GARCH models, state space models, and more.

products-rats.svg

RATS

Session 1: Bootcamp Part 1

Visualization with GGPlot2 and data analysis

Session 2: Bootcamp Part 2

Session 3: Univariate Models Part 1

Session 4: Univariate Models Part 2

Session 5: Vector Autoregressions

Session 6: Cointegration and error-correction models

A popular open source option for statistical computing and graphics that is available on many UNIX platforms, Windows and MacOS

Estimating, evaluating, and forecasting ARIMA models

Understanding the concept of non-stationarity and testing for unit roots

Estimating autoregressions, granger causality tests, and impulse response functions

Understanding the concept of cointegration and testing using the Enger-Granger test

Estimating and forecasting with vector error correction models

products-r.svg

Teaches econometrics and forecasting at the University of Colorado Denver. He has over 15 years of econometric experience in both the private and public sector.

Daniel L. Jerrett, Ph.D.

Abdel M. Zellou, Ph.D.

Bootcamp and Econometric Tools 1 & 2

Econometrics Training: Applied Time Series in R

Session 5: Multivariate Models Part 1

Session 6: Multivariate Models Part 2

Haver partners with SmartMacro Labs to offer a two week immersive online course introducing participants to the power of the Python programming language. This applied hands-on training develops statistical economics skills with a concentration on economics and finance.

Gain proficiency in building modern time-series models in Python through real-world applications. Designed for professionals in both the public and private sector, the course does not require prior experience, assuming only that participants have a basic understanding of regression and statistical concepts.

Proficiency working in Jupyter notebooks and the Python environment

products-python.svg

Remote Training: Applied Time Series in Python

Soaring AI-related investment in US underpinning strong contribution of business spending to overall growth

Michael J. Moran

260223F.jpg

260223G.jpg

The National Activity Index published by the Federal Reserve Bank of Chicago totaled 0.18 in January, a jump of .39 index points from the reading in December. The latest tally was favorable. Not only did the measure post a sharp month-to-month change, but it also signaled that the economy is growing faster than its long-run trend rate. 

(The index is constructed to equal zero when the economy is growing at its long-run trend rate. Thus, positive readings suggest brisk growth. Negative readings should not necessarily be viewed as disappointing. Researchers at the Chicago Fed note that the index needs to ease to -0.7 before recession becomes a risk. The measure would be deeper in negative territory if the economy was contracting.)

Although January brought encouraging news, an element of caution is warranted. The index often moves erratically, and thus a single monthly change could be dominated by random volatility. A three-month moving average squeezes out most of the volatility, leaving a more reliable perspective on the economy. Here, too, the news in January was favorable, as the 3-month average rose .23 index points to -.06, just a shade below a signal of trend growth and the best reading since April of last year. The eight months between April 2025 and January showed an average index value of -.24, with observations ranging from -.13 to -.38.

The National Activity Index is a composite of 85 economic indicators, grouped into four major categories: production and income; employment, unemployment, and hours worked; personal consumption and housing; sales, orders, and inventories. All four components improved month-to-month, with the production and income component providing most of the lift (up .22 index points to .19). Three of the four components were above zero in January, with only the sales, orders, and inventory in negative territory (and barely so at -0.02).

These figures are available in Haver's **SURVEYS** database.

- January results showed noticeable improvement from prior months.
- All four components contribute positively.

Chicago Fed National Activity Index: Suggesting Trend-Like Growth

Robert Brusca

260223E.jpg

Germany's IFO index shows improvement in February compared to January. The climate index improves, the current conditions index improves, and expectations improve. All three all-sector headlines improve in February compared to January. However, all three metrics continue to register net negative readings, indicating that the sectors are showing more contraction than expansion, but less contraction than it had been the case previously. We ranked these headlines on data back to 1993; the climate index has a 22.5 percentile standing, expectations have a 22.4 standing, and current conditions have a 16.9 percentile standing. All three of the headline metrics have standings in the lower quartile of their queue of values back to the early 1990s. This is clearly not a stellar report but a report from an economy that is struggling. Even though it's widely reported that the climate index is currently on a six-month high; it may be on a six-month high, but it's not really much different than it was six months ago and only up by 1.1% from its lowest mark in the intervening five months—so not impressive. This is a grueling, crawling, dig-out from difficult circumstances. The COVID pandemic was a sucker punch to the global economy, and it was followed by the Russian invasion of Ukraine that hit Europe especially hard because of the proximity.

Since then, other geopolitical tensions have risen. Europe's relationship with the United States has deteriorated, in part because Europe—Germany in particular—was reluctant to detach itself from its mercantile relationships with Russia and was also reluctant to increase its financial participation in NATO. The U.S. was carrying the burden of the protective umbrella for Europe at a time when Europe didn't really think it needed a protective umbrella and so it was unwilling to pay for anything more substantial. When the rainy day came and when the Russian invasion of Ukraine occurred, Europe was completely unprepared and the U.S. was its protector. While Germany and other European countries quickly made it clear that they were fully on board with NATO defense, the U.S. was much less enthusiastic about carrying the burden after Europe had put its financial probity ahead of its defense spending obligations. Now Europe does not see the U.S. and its perceived greater needs for security because of improved navigation from the Arctic Circle to the North Atlantic.

These troubles continue to dog the relationship between the United States and Europe, and the European Community. And Europe’s own economic struggles continue. In the climate readings, all four of five sectors improve from January to February. Current conditions improve in three of the five sectors, with retailing and wholesaling slipping month-to-month. Expectations do improve overall, but three of five sectors take a step back in February, with expectations improving sharply for services, and substantively for wholesaling. Still manufacturing, construction, and retailing expectations stepped back.

Across the three broad areas of assessment across five sectors in each, there are only two (two of fifteen) sector metrics that rank above their 50th percentile (above their respective medians) when ranked against readings back to 1993. Both of those elevated rankings are for construction, one for Climate, and the other for current conditions, but construction expectations, at a 44.5 percentile standing, are not far below their median.

Still, the lowest environmental standings are for services, except under expectations where the services reading is second-weakest to retailing. The graphic shows some tendency for an upturn in climate and current conditions, but they are countered by an expectations path that has topped out and has been gradually leaning lower—and doing that with a current ranking in only its 22.5 percentile. So, while there is some life in the IFO survey this month, we are still looking at a patient in intensive care, not one ready to join Olympic competition—at least not yet.

260223.jpg

Germany

German IFO: Business Confidence Claws Its Way Out of a Hole

Tian Yong Woon

In our Letter this week, we delve deeper into AI, with a focus on Asia and, where relevant, comparisons with the US. The global economy remains firmly in the buildout phase of AI, where much of the near-term economic benefit is derived from investment in hardware and enabling infrastructure—such as AI chips, memory, and data centres. At the same time, though still at a relatively early stage, firms are beginning to adopt advanced AI capabilities and integrate them into workflows in pursuit of productivity gains. So far, labour productivity improvements have been evident in parts of Asia (chart 1), largely driven by stronger exports (chart 2) and capital deepening—similar to trends observed in the US (chart 3) and, within Asia, in economies such as Malaysia (chart 4). However, these gains have yet to translate into a clear and sustained acceleration in total factor productivity (TFP), which accounts for the combined use of labour, capital, and other inputs—though TFP is admittedly a challenging metric to estimate, as illustrated by the US case (chart 5). This raises an important question: what happens when the AI buildout phase begins to moderate, and the associated investment-led tailwinds fade? At that point, further AI-related gains will depend more heavily on the successful embedding and diffusion of AI across sectors. On this front, Asian economies remain at very different stages of readiness to adopt AI (chart 6), suggesting that the next phase of productivity gains may be uneven across the region.

**The productivity story**
Several Asian economies have already recorded substantial productivity gains in recent years, even before the narrative around AI-driven productivity boosts gained traction last year. As shown in chart 1, measured by labour productivity—simply defined as output per employed person—economies such as Taiwan and several in Southeast Asia stand out for their strong performance. However, these gains cannot be attributed solely to the promise of AI. By definition, labour productivity rises whenever output (the numerator) grows faster than employment (the denominator). Such an outcome can stem from a range of factors—cyclical recoveries, capital deepening, sectoral shifts, or efficiency improvements—and does not necessarily reflect widespread AI adoption.

260223T1.jpg

Digging deeper, however, it would be equally misleading to suggest that Asian economies have not benefited from AI-related productivity gains. As shown in chart 2, exports have been a key driver of growth across several economies in recent years. In particular, Taiwan and South Korea—both major producers of AI-related hardware—have benefited from strong global demand for advanced chips and memory used in AI infrastructure. Taiwan stands out: its Q4 real GDP growth surged to 12.7% y/y, with net exports accounting for nearly all of that expansion. Notably, integrated circuits—often comprising around a third of total exports—have played a central role in this performance. Thus, while many Asian economies may not yet be at the forefront of deploying cutting-edge AI models domestically, they are significant beneficiaries of the current AI infrastructure buildout. By supplying the semiconductors and components underpinning this expansion, parts of the region are already capturing meaningful gains from the AI cycle.

260223T2.jpg

**Capital deepening**
Another key channel through which labour productivity can be achieved—aside from reducing the denominator (employment)—is capital deepening, or an increase in capital per worker. On this front, some advanced economies, including the US, have already moved in this direction as firms ramp up investment in AI-related infrastructure and equipment. In recent years, growth in investment in information processing equipment and data centres has outpaced other demand components, such as consumption and exports, reflecting in part the ongoing AI-related buildout (chart 3).

260223T3.jpg

The growth-enhancing—and by extension productivity-supporting—effects of such AI-related capital deepening are also evident in parts of Asia. In Malaysia, for example, the rapid expansion of its data centre industry has spurred strong gains in investment and construction activity, alongside spillovers to related sectors (chart 4). Importantly, these AI-related, investment-led gains are not limited to data centre development. Capital outlays to expand production capacity for AI-related goods—such as semiconductor fabrication and advanced chip manufacturing—operate through the same channel, increasing the capital stock and, over time, lifting output per worker.

260223T4.jpg

**Beyond the AI buildout**
So far, the discussion has centred on productivity in terms of labour productivity—output per employed person. That metric can be lifted relatively easily by increasing the numerator, whether through stronger exports or higher investment, as outlined in the previous sections. But what about gains in total factor productivity (TFP), rather than just labour productivity? TFP takes into account inputs beyond labour, including capital and other factors. An increase in TFP therefore implies higher output generated from the same combination of labour, capital, and other inputs. It is arguably this type of efficiency gain that investors hope will materialise from the adoption of AI. That said, TFP is notoriously difficult to measure. In the US, for instance (see chart 5), estimates of TFP growth can vary depending on the data source and methodology used. Even so, higher-frequency measures of US TFP have yet to show an outsized acceleration, suggesting that AI-related optimism has not—at least for now—translated into a discernible surge in measured productivity.

260223T5.jpg

Another key dimension to monitor is how prepared Asian economies are to embrace AI and embed it effectively in order to capture its anticipated productivity gains. While many economies in the region are currently benefiting from labour productivity gains—driven by stronger manufacturing activity and exports of AI-related goods and infrastructure—such gains may eventually moderate, particularly once the initial buildout phase peaks. At that stage, the more durable benefits will need to come from a different channel: the total factor productivity gains associated with the actual adoption and diffusion of AI across industries. On this front, Asian economies differ markedly in their readiness. As shown in chart 6, an IMF index—incorporating factors such as digital infrastructure, innovation capacity, human capital, and regulatory frameworks—highlights this divergence. More advanced economies, including Singapore, South Korea, and Japan, rank highly, while others such as Vietnam, India, and the Philippines lag behind. That said, the concerted efforts by several developing Asian governments in recent years to strengthen digital infrastructure, upgrade skills, and refine regulatory frameworks suggest that this gap could narrow over time.

260223T6.jpg

Asia

Economic Letter from Asia: Asia and AI

260220N.jpg

260220O.jpg

Real Gross Domestic Product rose 1.4% in the fourth quarter of 2025, slower than the recent average but no doubt influenced by the government shutdown. Federal spending often moves erratically, but the tumble of 16.6% in the final months of last year was unusually large. Indeed, one has to go back to 1972 to find a sharper decline (and to the 1950s before that), and these results were only slightly weaker than the Q4 decline.

Even allowing for the government shutdown, the latest results were a bit shy of expectations. Without the drop in federal outlays, which subtracted 1.2 percentage points from GDP growth, the economy would have expanded 2.6% in Q4, still shy of most surveys that showed consensus forecasts in the neighborhood of 3.0%. A downside surprise of sorts occurred in net exports. Monthly figures through November had suggested that international trade would provide strong support, but this sector added only 0.1 percentage point to GDP growth.

Other areas posted nondescript results. Consumer spending and business fixed investment were less vigorous than the best results of the past two years, but the latest growth rates were comfortably within recent ranges. Housing activity remained soft, making the sixth negative contribution to GDP growth in the past seven quarters. However, the constraints from housing over this span have generally been small.

The fourth quarter, and most of last year, were oddities because of the government shutdown in Q4 and the imposition of tariffs in April followed by marked adjustments in subsequent months. Such changes can lead to pronounced swings in economic activity and leave a muddled picture on underlying trends. In such environments, a different perspective on economic output can be helpful. Specifically, private domestic final demand (PDFD, GDP less government spending, net exports, and inventory investment) excludes areas that often move erratically, thus removing random shifts and reducing volatility.

PDFD was especially useful last year because tariffs generated pronounced shifts in net exports and inventory investment, and the government shutdown distorted activity in the fourth quarter. As shown in the chart below (left), quarterly variations in private domestic final demand were modest last year. The underlying trend decelerated, but the slowing was modest and growth in the fourth quarter was respectable at 2.4%.

The GDP price index rose at an annual rate of 3.6% in the fourth quarter of the year, one of the heftier advances of the past three years. A surge of 15.6 percent in the prices paid by the federal government had a strong influence on this result (shutdown related?), but readings of 3.9% for business equipment and 2.9% for consumer goods and services also were suggestive of an inflation impulse in the economy.

The increase in consumer prices was evident in the December report on personal income and consumption, which showed month-to-month increases of 0.4% for both headline and core price indexes for personal consumption expenditures. Year-over-year readings inched higher to 2.9% for all items and 3.0% for prices excluding food and energy. The latest results will no doubt leave Fed officials uncomfortable.

The GDP data can be found in Haver’s **USECON** and **USNA** databases. **USNA** contains virtually all of the Bureau of Economic Analysis detail in the national accounts. The Action Economics consensus estimates can be found in **AS1REPNA**.

260220P.jpg

260220Q.jpg

- Non-federal activity was lighter than expected, but still respectable.
- PCE inflation remains stubborn.

Q4 GDP: Constrained by Federal Spending

Winnie Tapasanun

260220F23.jpg

260220G23.jpg

New single-family home sales fell 1.7% m/m (+3.8% y/y) to a seasonally adjusted annual rate of 745,000 units in December—above expectations—following a 15.5% jump to 758,000 in November and a downwardly revised 8.8% drop to 656,000 in October (previously -0.1%, 737,000), data from the U.S. Census Bureau showed. The Action Economics Forecast Survey had expected sales of 733,000 for December. The December figure was the second m/m decline in three months, though sales remained 43.5% above the July 2022 low of 519,000. For all of 2025, new home sales declined 0.5% after gains of 3.0% in 2024 and 4.4% in 2023. The December m/m sales fall occurred alongside a decrease in the average 30-year fixed mortgage rate to 6.20%, the lowest since September 2024, down from 6.24% in November and 6.72% in December 2024; the rate had subsequently continued trending downward, according to Freddie Mac.

By region, December new home sales delivered a mixed performance. Sales in the Northeast plunged 37.3% (+12.1% y/y) to 37,000 in December following a 37.2% gain in November, registering the lowest level since September and the steepest m/m decline since February 2023. Sales in the South fell 6.7% (-1.1% y/y) to 430,000, the first m/m fall in three months, after a 10.0% November gain. To the upside, sales in the Midwest jumped 31.7% (30.1% y/y) to 108,000 after a 2.4% November decline, marking the largest m/m gain since January 2024 and the highest level since December 2020. Sales in the West, up for the fourth time in five months, rose 9.0% (1.8% y/y) to a three-month-high 170,000 on top of a 41.8% November surge. Notably, the South remained the dominant region, accounting for 57.7% of total U.S. new home sales.

The median sales price of a new home rose 4.2% (-2.0% y/y) to a three-month-high $414,400 in December, the first m/m rise in four months, following a 1.4% decline to $397,600 in November (the lowest since July 2025). The median sales price was 10.0% below its record high of $460,300 in October 2022. The average sales price of a new home, up for the third straight month, rose 0.5% (4.7% y/y) to $532,600, the highest since July 2022, after a 1.7% November increase to $530,200. The average price was 1.6% below a high of $541,200 in July 2022. These sales price data are not seasonally adjusted.

The number of unsold new homes on the market fell 2.7% (-3.5% y/y) to 472,000 in December after a 1.0% decline to 485,000 in November, representing the sixth m/m fall in seven months and the lowest level since August 2024. The latest figure was 10.8% above a low of 426,000 in July 2023. The seasonally adjusted months' supply of new homes for sale dipped to 7.6 months in December—the lowest reading since July 2023—after falling to 7.7 months in November. The latest reading, while above a low of 6.9 months in May 2023, remained below a high of 10.6 months in July 2022.

The median number of months a new home stayed on the market rose to 3.0 months in December, the highest since April 2022, after rising to 2.9 months in November. The latest number was above its record low of 1.5 months in September and October 2022 but well below a peak of 5.1 months in March 2021. These figures date back to January 1975.

New home sales are recorded when the sales contract is signed. New home sales activity and prices are available in Haver's **USECON** database. The consensus expectation figure from Action Economics is available in the **AS1REPNA** database.

260220H23.jpg

260220I23.jpg

- December sales -1.7% m/m (+3.8% y/y) to 745,000; still 43.5% above the July ’22 low.
- Sales m/m down the Northeast (-37.3%) and South (-6.7%), but up in the Midwest (+31.7%) and West (+9.0%).
- Median sales price up to a three-month-high $414,400; avg. price up to $532,600, highest since July ’22.
- Months' supply eases to 7.6 mths., lowest since July ’23.

U.S. New Home Sales Decline in December After November’s Jump

260220E.jpg

PMI data for February, issued by S&P on a flash basis, show relatively broad improvement PMI performance in February, with the exception of Australia, India, and the United States. The European Monetary Union shows improvement in its composite index as well as for manufacturing and services. This pushed to higher ratings, supported by Germany and France with the exception that French manufacturing took a step back in February. In the United Kingdom, the composite and manufacturing indexes moved ahead but services weakened. In Japan, the composite, manufacturing and services all improved on a month-to-month basis. However, Australia saw a step back in all three sectors after several months of continued strengthening. India also saw a step back in the composite and services after a recent string of improvements being reported. The United States shows uneven conditions, with a weakening across the board month-to-month in February vs. an across-the-board rebound in January, and then across-the-board weakening in December. This leaves the U.S. in a weakening and volatile situation

An unweighted average of the monthly composite, manufacturing, and services indexes shows a one-tick step back for the composite, a small step up for manufacturing, and a small step back in services for February compared to January.  January had made across the board improvements relative to December.

Sequential comparisons where data are lagged by one month and the calculations are made only off of hard data (not off of flash or preliminary data) show mixed economic conditions and unclear trends. EMU shows sequential strengthening in the composite as do France and Japan; Japan also shows persistent strengthening in manufacturing, as the yen has been dropping. And while there is a lot of weakening sequentially, there is no persistent weakening reported for 12-months to 6-months to 3-months.

Over three months, there's a split in what is reported with 12 of the 24 sectors weakening compared to six months ago and 12 of them strengthening. Over six months, only six of the sectors are weakening; the remainder strengthening over 12 months; eight of the sectors are weakening over 12 months compared to a year ago, with the remainder strengthening. While it's not a clear path to deterioration, there is a clear tendency to deterioration over three months compared to six months. However, an unweighted comparison of the three sectors shows that composite manufacturing and services still demonstrate a bit of muddy water with the composites weakening over three months only by a tick to 52.9 compared to 53 while manufacturing improves slightly and services weakens slightly compared to six months. Over six months, there's only a slight step back from the statistics reported over 12 months.

**Not a lot of shifting, but some good news and bad news**
The bottom line is there's not a whole lot of change going on according to the PMI data. On a breadth basis, there has been some improvement, but the magnitudes involved seem to wash out some of that improvement, neutralizing when expressing as average comparisons. However, one place where there does appear to be evidence of more solid performance is in the queue percentile standings. Among these 8 reporters involving 24 sectors, only four sectors report current standings below their median on data back to January 2022. One of those countries is India, with a weaker composite and services sector; the other is the United States, with a weaker composite and services sector. The U.S. composite has a queue percentile standing at its 42nd percentile, below its neutral position of 50%. The U.S. services sector is even weaker with a 36-percentile standing, close to the lower third of all the U.S. PMI values that have shown for services since 2022 on a monthly basis.

U.S. economy has been cranking out some reasonably strong economic numbers across the board including consumer spending and GDP figures, allowing for a disappointing fourth quarter that is being attributed to the government shutdown. However, what's been lacking on the U.S. side has been strong job market performance and, of course, because of mortgage rates being high, there has been a weak construction sector in the U.S. economy. Still, the queue standing across the board for other countries should be reassuring. The average composite ranking for this group of countries is in the 70th percentile, with manufacturing in the 75th percentile and services in their 61st percentile. These are relatively firm readings and should be reassuring. At the same time, central banks have made progress on inflation and tightening cycles have not been engaged or easing cycles are being extended for all central banks except for the Bank of Japan, which has been experiencing growth and inflation in different measures from the rest of the G7 countries. On balance, however, global growth appears to be moving forward and now we'll have to see what happens with the U.S. Supreme Court having overthrown the imposition of tariffs that the Trump administration put in place without Congressional backing. We could see some considerable shifting in the months ahead.

260220.jpg

S&P PMIs Are Mostly Stronger in February

Andrew Cates

**The US risk premium** 
Over the past three months, global equities excluding the US have outperformed the S&P 500, while the broad US dollar index has edged lower — a subtle but potentially meaningful shift in market dynamics. After several years of US exceptionalism driven heavily by a narrow group of AI-exposed mega-cap stocks, investors appear increasingly sensitive to elevated US valuations, geopolitical noise and questions about how much of the AI upside is already priced in. The combination of relative US equity underperformance and a softer dollar suggests a modest rise in the US risk premium and a gradual rotation of capital toward the rest of the world. It is far from a wholesale reversal, but it does hint that the automatic bid for US assets may no longer be quite as unquestioned as it once was.

260219Z1.jpg

**Bank lending growth**
A key pillar underpinning the still-upbeat tone in global equity markets is the turn in the monetary cycle. As this chart shows, bank lending growth has begun to firm again across several major economies, most notably in the US and Japan, while the euro area has stabilised after last year’s contraction. Recent and prospective rate cuts — alongside earlier liquidity support — are now feeding through into credit creation. Stronger bank lending flows suggest that financial conditions are no longer tightening and may, in fact, be becoming more supportive of activity. For equity investors, this matters: improving credit momentum reinforces the view that the global economy can avoid a hard landing and that policy easing will cushion growth, helping to justify still-elevated valuations.

260219Z2.jpg

**US productivity growth and AI**
Turning to productivity, the recent improvement in US labour productivity is striking — but its composition matters. As the chart shows, much of the acceleration has coincided with a sharp rise in capital input, reflecting the surge in AI-related infrastructure spending: data centres, semiconductors and high-performance computing. In other words, what we are seeing is primarily a capital-deepening effect — the production of AI — rather than clear evidence that the widespread use of AI is yet transforming efficiency across the broader economy. Total factor productivity, the cleaner measure of underlying technological progress, has not exhibited a sustained breakout. That does not invalidate the AI story — diffusion lags are typical with general-purpose technologies — but it does suggest that markets may be extrapolating productivity gains that have yet to show up convincingly in the data.

260219Z3.jpg

**Industrial production in Advanced Asian economies versus the RoW**
If the productivity story in the US is being driven by the production of AI infrastructure, the industrial production data show where much of the early stages of that production process is physically taking place. Advanced Asia excluding Japan has clearly outperformed, with industrial output rising markedly relative to both the US and the euro area. This divergence neatly reflects the region’s central role in the semiconductor and electronics supply chain — from advanced chip fabrication in Taiwan and Korea to high-end components across the broader region. As AI-related capital expenditure has surged, demand for chips, servers and related hardware has provided a powerful cyclical boost to manufacturing activity in these economies. In that sense, Asia is not merely a passive beneficiary of the AI boom; it sits at its industrial core, supplying the physical backbone of the global AI build-out.

260219Z4.jpg

**US manufacturing orders of high-tech products**
The strength in advanced Asia’s industrial production is mirrored by developments in the United States, where new orders for computers and electronic products have continued to climb steadily. As the chart shows, US manufacturers’ orders for high-tech goods — a category that includes semiconductors, communications equipment and other AI-related hardware — have risen to fresh highs in recent months. This reinforces the view that the AI investment cycle remains firmly intact, with strong upstream demand for chips and electronic components feeding through both sides of the Pacific supply chain.

260219Z5.jpg

**Portfolio flows in Emerging Markets**
Capital flows are beginning to reflect that same industrial divergence. Emerging market equity inflows have turned more supportive in recent months, with Emerging Asia accounting for a meaningful share of the rebound. This aligns closely with the strength in industrial production across advanced Asia ex Japan, where AI-related demand for semiconductors and hardware has lifted manufacturing momentum. Investors appear increasingly willing to allocate capital toward economies most directly embedded in the AI supply chain. In that sense, the AI infrastructure cycle is not only reshaping trade and production patterns — it is also influencing global portfolio flows, reinforcing Asia’s central role in the current technology-driven upswing.

260219Z6.jpg

Recent weeks have seen renewed market swings — a softer US dollar, rotation within equities and shifting rate expectations — as investors reassess valuations, policy risks and the durability of US exceptionalism. Yet a consistent macro thread runs through our charts this week. Global equities continued to outperform the US, hinting at a modest rise in the US risk premium alongside AI-valuation concerns (chart 1). At the same time, firmer bank lending growth across major economies reflects looser monetary policy and arguably helps explain why risk assets remain broadly supported despite volatility (chart 2). On productivity, a US pickup is evident but likely reflects capital deepening from AI infrastructure rather than a decisive surge in total factor productivity (chart 3). Industrial production in advanced Asia reinforces where that build-out is concentrated — at the heart of the semiconductor supply chain (chart 4) — while strong US orders for computers and electronic products underline the intensity of domestic demand for high-tech and AI-related equipment (chart 5). Finally, renewed equity inflows into Emerging Asia show investors positioning around this AI-centric growth dynamic (chart 6). In short, easier policy and AI-driven investment are sustaining momentum — but the breadth and durability of the productivity dividend remain a key open question.

<iframe data-testid="embed-iframe" style="border-radius:12px" src="https://open.spotify.com/embed/episode/6HdtU5cBvoC3e7yqIV9oWN?utm_source=generator" width="100%" height="152" frameBorder="0" allowfullscreen="" allow="autoplay; clipboard-write; encrypted-media; fullscreen; picture-in-picture" loading="lazy"></iframe>

Charts of the Week: AI, Policy Easing and a Shifting Risk Premium

Sandy Batten

260219J20.jpg

260219K20.jpg

The U.S. trade deficit in goods and services (BOP basis) widened to $70.3 billion in December from a downwardly revised $53.0 billion in November (previously $56.8 billion), according to the U.S. Census Bureau. The goods deficit increased markedly to $99.3 billion from a downwardly revised $83.6 billion in November (previously $86.9 billion) while the services surplus narrowed slightly to $29.0 billion from $30.6 billion (previously $30.1 billion). The Action Economics Forecast Survey looked for a deficit of $55.5 billion.

The real (inflation-adjusted) trade deficit in goods widened to $97.1 billion (2017$) in December from $84.5 billion in November (previously $87.1 billion). However, for all of the fourth quarter, the real deficit narrowed to $245.0 billion from $261.6 billion in Q3. This suggests that trade contributed around 0.5%-point to overall GDP growth in Q4. Q4 GDP will be released tomorrow, December 20.

Exports of goods and services fell 1.7% m/m (+6.3% y/y) in December on top of a 3.4% monthly decline in November. Goods exports (Census basis) dropped 3.0% m/m in December following a 5.2% m/m decline in November. The December decline was led by a 14.5% m/m fall in exports of “other goods,” the fourth monthly decline in five months. Exports of industrial supplies and materials slumped 12.3% m/m on top of an 8.0% m/m decline in November. The $8.7-billion December drop in industrial supplies exports was almost completely accounted for by a $7.1 billion decline in nonmonetary gold exports. By contrast, exports of nonfood consumer goods excluding autos jumped 8.4% m/m in December and capital goods exports gained 4.2% m/m. Exports of services increased 0.4% m/m in December after a 0.2% monthly decline in November. The December gain was led by a 2.5% m/m increase in travel exports and a 1.6% m/m rise in transport exports.

Imports of goods and services increased 3.6% m/m (-2.6% y/y) in December on top of a 4.2% monthly jump in November. Goods imports (Census basis) rose 3.8% m/m on top of a 5.8% monthly jump in November. Imports of industrial supplies surged 15.9% m/m in December, more than reversing a 5.6% m/m decline in November. Capital goods imports increased 5.5% m/m, their third consecutive solid monthly gain. Auto imports rose 3.9% m/m in December, the first monthly increase in four months. By contrast, imports of nonfood consumer goods excluding autos fell 5.9% m/m, but this was after a 15.6% m/m surge in November. Imports of services rebounded in December, rising 2.7% m/m following a 1.3% m/m decline in November. Transport imports jumped 12.1% m/m, their first monthly gain in three months. Travel imports rose 2.4% m/m following a 1.6% monthly decline in November.

By country, the goods trade deficit with China narrowed to $12.4 billion in December from $14.4 billion in November. Exports to China rebounded in December, surging 26.5% m/m after slumping 20.1% m/m in November. Imports from China declined 2.0% m/m in December on top of 4.1% monthly decline in November. This was their fourth monthly decline in the past five months. The trade deficit with the European Union narrowed to $11.1 billion from $14.3 billion in November. The deficit with Japan widened to $5.3 billion from $4.7 billion in November.

The international trade data can be found in Haver’s **USECON** database. Detailed figures on international trade are available in the **USINT** and **USTRADE** databases. The expectations figures are from the Action Economics Forecast Survey in **AS1REPNA**.

260219L20.jpg

260219M20.jpg

- The overall deficit widened to $70.3 billion in December from $53.0 billion in November.
- The goods deficit widened markedly to $99.3 billion from $83.6 billion in November.
- The services surplus narrowed slightly to $29.0 billion from $30.6 billion in November.
- Exports fell 1.7% m/m on top of a 3.4% monthly decline in November while imports rose 3.6% m/m following a 4.2% gain in November.
- For all of Q4, the real goods trade balance narrowed, suggesting that trade contributed to overall GDP growth.

U.S. Trade Deficit Widened Much More Than Expected in December

Kathleen Stephansen, CBE

large_260225A1.jpg

medium_260225A1.jpg

small_260225A1.jpg

thumbnail_260225A1.jpg

260225A1.jpg

large_260225B1.jpg

medium_260225B1.jpg

small_260225B1.jpg

thumbnail_260225B1.jpg

260225B1.jpg

Mortgage applications rose 0.4% w/w (69.1% y/y) in the week ending February 20 after rising 2.8% w/w (57.7% y/y) in the week ending February 13, according to the Mortgage Bankers Association’s (MBA) Weekly Mortgage Applications Survey. Applications for loans to purchase a house declined 4.7% w/w (+13.1% y/y) in the latest week, following a decline of 2.7% w/w (+9.1% y/y) in the prior week.  Applications for loan refinancing rose 4.1% w/w (150.3% y/y) in the week ending February 20, after rising 7.1% w/w (131.8% y/y) in the February 13 week.

The effective interest rate on a 30-year fixed-rate loan dropped 9bps to 6.24% in the week ending February 20 from 6.33% in the week ending February 13. The rate on 15-year fixed-rate mortgages edged down 2bps to 5.66% in the latest week from 5.68% in the prior week. The 30-year Jumbo rate rose 3bps to 6.32% in the February 20 week from 6.29% in the February 13 week. The rate on a 5-year ARM dropped 14bps to 5.38% in the week ending February 20 from 5.52% in the week ending February 13, reaching the lowest level since September 2022.

The share of applications for refinancing an existing loan rose to 58.6% of total applications in the February 20 week from 57.4% in the February 13 week. The adjustable-rate mortgage (ARM) share of activity was unchanged at 8.2% in the February 20 week from the prior week.

The average size of a mortgage loan edged up 0.4% w/w (8.3% y/y) to $416,800 in the week ending February 20, after rising 2.3% w/w (8.6% y/y) to $415,100 in the week ending February 13. The average size of a purchase loan declined 1.1% w/w (+0.4% y/y) to $442,900 in the latest week, after rising 0.7% w/w (2.0% y/y) to $448,000 in the week ending February 13. The average size of a loan to refinance a mortgage rose 2.0% w/w (34.3% y/y) to $398,400 in the week ending February 20, after rising 4.1% w/w (34.1% y/y) to $390,700 in the week ending February 13.

The Mortgage Bankers Association Survey covers 75% of all U.S. retail residential mortgage applications and has been conducted weekly since 1990. Respondents include mortgage bankers, commercial banks and thrifts. The base period and value for all indexes is March 16, 1990=100. The figures for weekly mortgage applications and interest rates are available in Haver’s **SURVEYS** database.

- Applications for loans to purchase declined in the latest week while refinancing loan applications rose.
- Effective interest rate on 30-year fixed loans declined 9bp to 6.24%.
- Average loan size edged up.

U.S. Mortgage Applications Edged Up in the February 20

Our global data offering is unparalleled in terms of coverage, quality, and organization. We offer 290+ databases from 2500 government and private sources, a wealth of key survey and forecast data covering the world’s economies, and a robust offering of archive databases that drive machine learning for big data applications.

Optimized for precision analytics, our data is seamlessly integrated with Microsoft Office and all leading statistical packages through our DLX® software and platform independent HaverView™ website. We further offer direct database feeds, client server or cloud-based delivery, and RESTful API.


Privacy Policy

Disclaimer

Contact us

Service Health

A premier time series data platform for the global strategy, research, and quant communities. 

London

Singapore

Tokyo

New York

Economy in Brief

Viewpoints

Contact Us

Our Data

Our Products

What We Do

Our Authors

About Us

Data News

Featured Data Additions

Resources

Release Calendars

Training Center

Webinars

Newsletters

Download Software

For Clients

Live Updates

#### World Economy Tracked for You 
Haver's worldwide economic data coverage provides a productive resource for analysts, strategists, and risk managers alike. With 45+ years as the industry leader, our breadth of data is unmatched.


#### Diverse Data Sources
From international banks to the farthest reaches of emerging markets, our coverage meets the needs of the global research community.

#### Driven By Quality
Committed to quality and transparency, all data sourced and delivered is continuously monitored and validated to ensure the highest level of accuracy. 

#### Global Data 
From archives to forecasts, primary-sourced from around the globe, we offer 290+ time-series databases.

#### Speed Matters
Delivering precise data as quickly as possible is our priority, which we accomplish in about one minute of release.

Data

#### Comprehensive Solutions
Effortlessly navigate, analyze, and visualize data with our broad array of tools and seamless integration with industry leading statistical and econometric packages. Gain insight and share your perspectives with ease.

#### Secure Data Access 
Effortlessly manage and work with data — delivered to your network or via our cloud services. The choices are easy, and the results are the same, fully secure and consistent access.

#### HaverView™
With a commitment to making data accessible for any work environment, our HaverView web application offers a flexible and productive resource for the dynamic research markets we serve. 

#### Data Exports and Workflow Management
Our DLX software integrates with MS Office allowing seamless publication in PowerPoint and Word as well as intuitive data management in Excel. Flexible data feeds are also supported by a variety of methods. 

Software

#### Econometric Software Support
Use our various integration tools to access our data in popular econometric software such as SAS, Stata and MatLab.

#### Designed for Programming
Access our data using multiple programming languages including Python, R, C and C++.

#### RESTful API
Restful API access to all DLX databases, supporting quant and machine learning applications. Big data rejoice!

Connectivity

#### Commentaries, Charts and Webcasts
Our team of noted economists produce a daily offering of commentaries supported by our data and analytics. Our charts and webcasts offer a fresh perspective on the state of world economy.

#### Value Added Calculations
World and regional aggregates and many other value added calculations for individual countries including seasonal and holiday adjustments. We love math!

#### Blue Chip Consensus Forecasts
Survey of America's leading business economists for their forecasts of economic and financial data. Results in two monthly publications, the Blue Chip Economic Indicators and the Blue Chip Financial Forecasts, published by Wolters Kluwer.

#### Curated Navigation
Intuitive menus designed and reviewed by our economists for easy navigation of Haver's expanding ocean of data.

Research

#### Data for Colleges and Universities
Customized data available through HaverView providing access to over 120,000 economic and financial time series for the US and UK.

#### Academic Data Scope
National Account (GDP), Production, Labour Market, Inflation, Monetary/Fiscal Policy & International Trade. High-frequency data covering financial variables and drivers of the real economy, and key ESG environmental indicators. Plus, time series from the IMF on 190+ countries to aid cross comparison.

University

#### Move Back through Time
Terabytes of time series from primary sourced databases available in monthly snapshots as early as 2000, in weekly snapshots as early as 2007 or the point of inception of the database. 

#### True DLX Database Snapshots
The data are true snapshots of DLX databases unrevised from the point in time of update.  Easily view our archive data in HaverView with thousands of monthly archives currently available for over 60 unique databases. 

Discover Haver!

Data

Software

Connectivity

Research

University

Archive

World Economy Tracked for You

Diverse Data Sources

Driven By Quality

Global Data

Speed Matters

Economic Commentary

U.S. Mortgage Applications Edged Up in the February 20

Recent Commentaries

German Confidence: GfK Consumer Climate Erodes

Featured Data Additions, February 25, 2026

Consumer Confidence: Improvement in February from Upward Revised Levels in January, but Individuals Are Still Unhappy

U.S. FHFA House Price Growth Slows in December; Q4’25 Momentum Picks Up

French Manufacturing and Services Surveys: Two Sectors, Two Different Trends, and Lingering Weakness

U.S. Factory Orders Fall in December on Weak Nondefense Aircraft Orders

Latest Releases

Graph of the Week

Soaring AI-related investment in US underpinning strong contribution of business spending to overall growth

Economy in Brief

Viewpoints