Goldman Sachs' latest research found that initial unemployment applications, Philadelphia Fed Manufacturing Index, ISM Services Index and unemployment rate are the best indicators to warn of an economic slowdown. These indicators usually signal just one month after the recession begins, while hard data such as GDP will take four months to show significant weakness.

When will the tug-of-war between Trump and the Fed push the United States into a recession trap?

Goldman Sachs' latest research, through analyzing the real-time data performance of 45 economic indicators, found that during the slowdown in economic growth, some indicators reflected changes in economic conditions earlier and more accurately than others. The research results show that initial unemployment applications, Philadelphia Fed Manufacturing Index, ISM Services Index and unexpectedly unemployment rate are indicators that provide the most timely signal.

These indicators outperform other data because they are published frequently, have small corrections, and can be published earlier. Initial unemployment applications are announced every Thursday, and unemployment rate data will be released next week.

In the study, Goldman Sachs used a daily dataset containing the latest data on 45 economic indicators, which evaluated real-time accuracy through their ability to predict economic slowdowns. The results show that business surveys and labor market data provide the most timely information at the economic turning point.

In the past, in recessions with clear catalysts—such as the oil shock in 1973, Volk’s rate hikes in 1979-1980, the surge in oil prices in Kuwait in 1990, and the collapse of the Internet bubble in 2001— hard economic data (such as real GDP) usually takes about four months to show clear signs of weakening in real-time data, and the expected portion of business surveys usually begins to decline about a month after the impact occurs.

The survey data has been alerted, but should the market believe it?

Goldman Sachs pointed out that the expected parts of the business survey have dropped significantly, and some have even dropped to their lowest levels except during the recession. But given that soft data has been sending pessimistic signals over the past few years but failing to accurately predict economic performance, this raises an important question: Should the market now believe the signals of survey data?

The research team believes that the current survey data may be more reliable than in the past for two reasons:

First, the recent deterioration was driven primarily by the actual decline in activity among more respondents, rather than simply the increase in respondents whose activity remains unchanged;

Second, some of the biased factors that influence business surveys in the post-epidemic period (such as the normalization of the pace of epidemic reopening, unprecedented supply chain disruptions, and the transformation from goods to services) are less important now.

The report reads:

Our analysis warns not to ignore the deterioration of current survey data, although these indicators have been too pessimistic in recent years. The pattern of deterioration in data in recent weeks is similar to previous “event-driven” slowdowns in growth.

Consumer spending will slow down with tariffs, and capital expenditure will suffer the biggest hit in the second half of the year

Goldman Sachs expects higher tariffs to push up consumer prices, thereby lowering actual disposable income and consumer spending. After analyzing the trade war between 2018 and 2019, the research team found that the transmission of tariffs to inflation was the most significant between March and July 2018 and January and April 2019.

The report notes that the inflation effect will usually appear within two or three months of the implementation of the tariffs, and consumer spending is expected to slow down soon after the price rises. Historically, core retail sales were the best hard indicator of consumer spending during the economic slowdown, providing a more timely real-time signal than monthly consumer spending and disposable income data.

The study further pointed out that tightening financial conditions and rising policy uncertainty will have a negative impact on capital expenditure this year. According to Goldman Sachs' capital expenditure pulse model, capital expenditure growth will suffer a drag of up to 5.5 percentage points in the second half of 2025.

In a capital expenditure-led economic slowdown in the past, it took about five months for real-time available hard data (such as capital product orders) to weaken and is noisier than soft data, which usually starts to worsen about a month after the slowdown begins.

Overall, Goldman Sachs expects survey data (soft data) to continue to soften, while hard data will not begin to weaken until mid-to-late summer, when higher prices, weaker spending and slower hiring may begin to appear in official statistics.

Goldman Sachs warned not to ignore the deterioration of current survey data, although these indicators have been too pessimistic in recent years.

As of now, the recent pattern of deterioration in data is similar to the previous “event-driven” slowdown, but it is still too early to draw strong conclusions from the current limited data.