Forecasting NFP with the Fed’s next move in focus

The Federal Reserve’s monetary policy balances a dual mandate, aiming for a 2 percent inflation target while supporting sustainable employment. The Fed’s steep tightening cycle was extended at the July policy meeting, which unveiled a 25-basis-point rate hike to contain inflation risks on the back of a resilient labour market. 

The next key data point for Fed watchers is the US non-farm payrolls (NFP) report, which will be released on Friday. This employment statistic is closely studied by economists, strategists and market participants and will likely influence the Fed’s next policy decision in September. 

To generate this prediction for the Friday NFP figure, we deployed univariate and multivariate forecasting techniques, using our partnership with Indicio Technologies.  

Considerations for developing the univariate and multivariate forecasts

Historic NFP data: This serves as the base data for the univariate model. The persistence of recent data points in the time series – the recency bias -- suggests an estimate close to 145,000. It also reflects a sequential decline from the print of 904,000 in February, 2022. 

This lower-than-consensus number does not reflect the economic resilience seen recently, especially in the last couple of quarters. (As the right-most columns of the next table show, the last three years have been far stronger on average than the past decade.) 

When we construct a multivariate model, we incorporate factors influencing underlying economic shifts and driving employment trends. So, to create a better gauge of NFP, several other strongly correlated factors are considered:

Employment-related indicators: the unemployment rate, jobless claims, the labour force participation rate and private sector-employment data. These indicators provide a broader context for labour-market health. In particular, the high-frequency weekly initial jobless claims have a clear inverse relation with the NFP job creation, as the next chart shows.

Economic indicators: GDP growth, inflation, consumer spending and business investment all influence the labour market. Higher GDP often correlates with positive non-farm payroll numbers:

Industry-specific data: This includes manufacturing versus services trends, as well as specific industries or sectors that have a significant impact on the overall job market.

As the next chart shows, post-Covid, services sectors have driven payrolls growth much more than their counterparts in goods sectors. This trend was helped along by government transfers that stoked consumer spending on discretionary services.

Next, breaking down the services segment, key contributions were made by professional and business services, transportation and warehousing, and the education and health sector.

‍Monetary policy: Global and domestic economic shifts influence the Fed, and a restrictive or expansive Fed stance influences business sentiment and hiring intentions.

Monetary transmission from steep rate-tightening cycles adversely impacts NFP, typically with a two- to three-quarter lag, as this chart shows.

The consumer confidence index: This indicator can provide insights into potential changes in consumer spending and hiring.

Leading Economic Indicators: These include metrics like the purchasing managers' indices (PMI) and housing starts, which can help predict economic trends.

The deceleration seen in Services PMI (the employment index) suggests softening is ahead for NFP.

Seasonal adjustments: Non-farm payrolls can exhibit seasonal patterns due to factors like holiday hiring or school-year cycles. As a result, seasonal adjustments are applied to help remove these effects and reveal underlying trends.

Sentiment analysis: Keeping a close tab on media reports, expert opinion and surveys helps gauge market sentiment and the expectations that impact job-market dynamics.

Statistical forecasting methods: Indicio utilises various statistical techniques including time-series analysis, regression and ARIMA. It considers 16 statistical models, such as VARX Lasso, HVAR, MIDAS , ANN and others (if a statistic fit is found) to arrive at a weighted forecast for non-farm payrolls.

The result of our forecast

As the next chart shows, we are predicting that non-farm payrolls will come in at 193,000 for July. That’s a tad lower than the consensus of 200,000 – and, in our view, not a deterrent for a Fed “pause” in September if CPI figures continue to show slowing inflation.

The multiple explanatory variables analysed in this blog all make different contributions to the overall result. Our last chart shows how these various factors are a support or a drag when our prediction is compared to the previous NFP print (209,000).

In conclusion, our analysis suggests that the cyclical downtrend for NFP will continue. And if inflation figures continue to ease, we could well have reached the peak of this Fed rate cycle. This would be consistent with market expectations – which are currently pricing in only a 16 percent chance of another interest-rate hike in September.