Starting with today’s update of looking ahead to the next quarter’s GDP, I’ll present several forecasts drawn from different methodologies for contrast and comparison on a regular basis. With that in mind, let’s consider how several forecasts for third-quarter U.S. GDP stack up.

The first chart compares four in-house forecasts that will be updated as new data arrives. I’ll briefly explain below how each of The Capital Spectator’s GDP forecasts are calculated. But first, take note that the first chart also includes two widely cited surveys of economists. The average guesstimate for each survey is presented for additional perspective. One source is the Survey of Professional Forecasters (SPF) via the Philadelphia Fed. It’s updated relatively infrequently, however. The current average forecast of real GDP growth of 1.6% for Q3 was published on August 10. The other survey is the September outlook via The Wall Street Journal, which asks several dozen economists for their outlook each month. The Journal’s current forecast from dismal scientists reveals an average forecast of 1.9% for Q3 GDP, based on September 7-11 polling.

One rationale for keeping on eye on multiple forecasts and tracking how they change through time is that it gives us an additional layer of intelligence to consider. Point forecasts at one moment in time are fine, of course, but as conditions change, so too should the predictions. How they change can sometimes tell us as much, maybe more, as a specific forecast at one specific moment. Are the forecasts continually rising, falling, or going every which way?

With that in mind, here’s how our four in-house forecasts compare in recent weeks:

Next, here’s a brief profile of how each of The Capital Spectator’s forecasts are calculated:

**4-Factor Nowcast.** This prediction is based on a multiple regression of quarterly changes for four key economic indicators: real personal consumption expenditures, real personal income less government transfers, industrial production, and private non-farm payrolls. This model compares the data on a quarterly basis, looking for relationships with GDP within each quarter from the early 1970s to the present. The four monthly data series are updated monthly and so as new data arrives, the forecast is updated—thus the term “nowcast,” which can be thought of as a dynamic forecast that draws on frequently revised inputs.

**10-Factor Nowcast.** This model also uses a multiple regression framework that’s updated as new numbers arrive, but with two key differences relative to the 4-factor model above. First, this model uses more factors—10 in all. In addition to the data quartet used in the 4-Factor model, the 10-Factor nowcast also looks at the following series:

• ISM Manufacturing PMI Composite Index

• housing starts

• initial jobless claims

• the stock market (S&P 500)

• crude oil prices (spot price for West Texas Intermediate)

• the Treasury yield curve spread (10-year Note less 3-month T-bill)

The second difference with the 10-factor model is that it analyzes relationships across a longer span of time by considering the average of changes across the trailing one-, two-, three-, and four-quarter comparisons.

**ARIMA Forecast.**The engine here is known as an autoregressive integrated moving average. The technique is one of using real GDP’s history from the early 1970s onward as the raw material for predicting the next quarter’s change. As the most recent quarterly GDP number is revised, I re-run the ARIMA forecasts, which are calculated in R software, using Professor Rob Hyndman’s “forecast” package, which optimizes the prediction model based on the data set’s historical record.

**VAR Forecast.** Finally, I forecast GDP with a vector autoregression model, which looks to several data series in search of interdependent relationships to provide some guidance about the future. In particular, I use the four factors in the 4-factor model above to generate VAR forecasts based on the historical records dating back to the early 1970s, using the “vars” package for R.

Okay, so what do all these forecasts tell us? Clearly, there’s a mixed bag in terms of the recent trend. Two of our in-house forecasts turned higher, but two slipped modestly. Perhaps it’s just the usual noise that comes with updating forecasts. We’ll learn more in the days and weeks ahead as fresh numbers are published and we move closer to the government’s official release date for the first official estimate of Q3 GDP: October 26.

Meantime, there’s a reasonably strong case for expecting that the economy will continue to post slow growth for the July-through-September period this year. There’s nothing to contradict that expectation in our latest economic trend analysis, a separate econometric technique for assessing recession risk.

As usual, all this could change if the next batch of data is particularly ugly. Indeed, September’s economic news starts arriving this week. Based on what we know at the moment, however, cautious optimism isn’t easily dismissed for thinking that the trend will struggle forward.

Robert EubankI appreciate these updates. Thanks