Our effort is to find the seven factors that work well together and give the most accurate look at the initial reading on GDP. The model does not include data points that become available after the initial announcement which make more accurate after-the-fact estimates possible.
Since we last wrote we have replaced two of the factors in the model. As reported earnings of the S&P 500 replaced stock prices and the quarterly difference in oil prices leading 21 months replaced the GDP's auto correlation.
Below is a chart that shows quarterly GDP, the model, the 7 factors in the model and the monthly data points used in each factor.
The model suggests GDP will grow 0.1% in the fourth quarter. The standard deviation of 1.3 suggests better than even odds growth will be between -1.2 and 1.4. This is a big gap from mainstream forecasts of around 2.5% to 3%.
Housing starts (shown in brown in the chart above) may account for a large part of the gap between our model and other forecasts. Using the nine month lead time, which has the best correlation, starts suggest GDP will have its weakest quarter since 2009. Housing starts were also calling for a weak Q1 2014 GDP which came in at -1.9%, while mainstream predictions for the initial estimate were around 3%.
The brown housing starts axis is calibrated for its weight in the model, as are the axes for the other six indicators. So if you took a ruler and visually added up the net change in each of the seven indicators it would equal the change for the model shown in red.
Monthly data points are shown as dots. Each segment in a line is the average of three monthly points that correspond with a quarter of GDP growth. Each monthly point is based on the annualized growth rate or change from the prior quarter's line segment. For example, the last brown dot on the chart is for November housing starts at 1028 thousand. November starts were 0.16% lower than the line segment for Q2 2015 growth (average of July, August and September starts). At an annualized rate that becomes -0.65%, which is where the last brown dot is plotted. When December starts are available the brown line can be extended to show housing starts' influence on growth for Q3 2015.
There are now two factors based on oil. The black line shows the concurrent change in price which is an indicator of worldwide GDP. The factors that make the world economy speed up or slow down affect the price of oil in the same way. In about 3 out of 4 quarters, the quarterly change in the price of oil and the change in the quarterly US GDP growth rate move in the same direction. The rate of decline suggests Q4 will be the weakest since 2009.
So housing starts and the price of oil suggest the weakest GDP since 2009.
The gold line shows the positive impact falling oil prices have on GDP. Note the gold axis for oil is inverted. Surprisingly the strongest correlation is with a 21 month lead time. So the recent fall in price should benefit growth in 2016. Oil prices were rising 21 months ago and the correlation suggests Q4 GDP will weaken.
The negative GDP impact of the falling oil price on a concurrent basis is almost three times the positive influence falling prices will have on future growth. Note the difference in the scales for the gold and black axes. The meme that low oil prices will give consumers increased disposable income to drive faster growth in 2015 is likely a logical sounding hypothetical without a basis in reality.
Industrial production (Green line and dots) suggests Q4 will be the weakest in 5 quarters. Although the November increase (last green dot), with the two month lead time, suggests Q1 2015 will get off to a good start.
S&P 500 earnings and retail sales suggest growth in Q4 will be stronger than in Q3. The other five indicators all point to weaker growth.
The difference in the yield of 10 year Treasury Bonds (blue line and dots) suggest Q4 will be the weakest since 2012 and that growth will weaken slightly further in Q1 2015.
Collectively these seven indicators suggest a very weak end to the year.
Manufacturers' new orders of durable goods may help explain why growth was so strong in Q3 and will likely be weak in Q4. New orders are not in the model because the correlation with industrial production was a little too high. Durable goods lead GDP by one month, so the November data released last week completes the estimate for Q4.
July was a blowout month for orders; at an annual rate it was up 147% from the prior quarter. This made for the strongest growth of durable goods for any quarter since the data series began in 1992. The three months corresponding with Q4 GDP growth all were about 27% below Q3. This makes for the weakest estimate since Q1 2009. Q3 was strong, but it likely robbed growth from Q4.
2014 will probably finish with a growth rate near 2.2%; it is at 2.6% through three quarters. It will likely be the 4th year in a row growth has been weaker than 2.4% and the ninth year it is under 2.7%. The last 9 years have annualized 1.4%. Quarterly growth fluctuates. The 5% Q3 growth does not mean the economy is strong or that it is accelerating.
Weak Q4 GDP growth should have feedthrough effects. Corporate earnings announcements starting in January should reflect the economic weakness. Job growth which lags GDP should weaken sometime in early 2015. The shattered perception of an accelerating economy will make the trade weighted dollar index and stock market (NYSEARCA:SPY) more vulnerable to sharp corrections.
Early indications are that 1st quarter 2015 GDP will return to modest growth. However, with growth running at a low 2% rate, the economy could dip into recession at any time.
Disclosure: The author is short SPY.
The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: There is no guarantee analysis of historical data, their trends and correlations enable accurate forecasts. The data presented is from sources believed to be reliable, but its accuracy cannot be guaranteed. Past performance does not indicate future results. This is not a recommendation to buy or sell specific securities. This is not an offer to manage money.