In January, I introduced a new aggregate neural network model that I developed to forecast the seasonally-adjusted, annualized, real rate of change in U.S. GDP. The GDP growth rate is only reported quarterly, but the model provides a new rolling 3-month GDP growth rate forecast every month (with a one month lag). As a result, the model generates more timely information about the growth of the U.S. economy than the quarterly GDP data.
Aggregate GDP Model
The Trader Edge aggregate GDP model represents the average of two neural network model forecasts. When constructing the two neural network GDP models, I started with the same data that I used to develop the recession models. For the first model, I identified the optimal look-back period for each of the independent variables and discarded variables that had limited explanatory power. I then estimated several neural network models with different architectures and retained the one with the best overall performance. For the second neural network model, I started with the same data, but let the neural network model choose the best combination of independent variables using a greedy algorithm.
The Trader Edge aggregate GDP model performed well historically. The standard error of the estimate was a respectable 0.75%. However, the forecast for the Q4 2012 GDP exceeded the reported GDP by over 2% and the latest rolling 3-month forecast is well above the consensus estimate for Q1 2013.
As explained in previous posts, there were two major drags on GDP in the fourth quarter. Business inventories were drawn down and not replaced during the quarter and government spending dropped by 6.6%. The sequester is now in effect, which could lead to further declines in government spending.
The model forecasts from 2000 to 3/1/2013 are included in Figure 1 below (blue). The purple line illustrates the actual quarterly GDP data. Both lines use the left vertical axis. The most recent "actual" GDP observation represents an interpolation of last quarter's actual data and next quarter's consensus forecast of +1.7%. The gray shaded regions represent past U.S. recessions as defined by the National Bureau of Economic Research (NBER).
Rolling GDP Forecast: Three Months Ending 03-01-2013
The interpolated, annualized real GDP growth rate for the three-month period ending 03-01-2013 was only 1.20%, which represents an increase of 0.41% from last month's interpolated consensus/reported value. The model estimate for the three months ending 03-01-2013 was a surprisingly strong 3.31%, up 0.29% from the revised estimate for the three months ending 02-01-2013. As was the case in the fourth quarter, the latest model forecast was significantly above the actual interpolated GDP data.
The GDP neural network model is still in the experimental stages. Unfortunately, it does not currently use any explanatory variables that measure the effects of inventory changes or reductions in government expenditures on GDP. If further austerity measures are implemented at the state and federal level, this may limit the ability of the model to accurately forecast the reported growth rate of GDP. Time permitting, I would like to research new variables that could capture these effects on a timely basis.
In the interim, the trend in the model forecasts should still provide insight into the growth rate of the private sector in the U.S. economy. The month-to-month increase in the rolling three month forecast continues to be a positive sign.
Your comments, feedback, and questions are always welcome and appreciated. Please use the comment section at the bottom of this page or send me an email.
If you found the information on www.TraderEdge.Net helpful, please pass along the link to your friends and colleagues or share the link with your social or professional networks.
The "Share / Save" button below contains links to all major social and professional networks. If you do not see your network listed, use the down-arrow to access the entire list of networking sites.
Thank you for your support.
Copyright 2013 - Trading Insights, LLC - All Rights Reserved.