Slowing global trade, weak economic figures and a sovereign debt crisis in the Eurozone (that just does not want to go away) have understandably led to volatile markets in recent times. Is it possible that the weakness in the U.S. has been overstated?
Abysmal 2Q GDP report
The theatrics on the part of Congress threatened to temporarily overshadow the batch of deteriorating economic reports culminating in the GDP report two weeks ago. In the report, the Bureau of Economic Analysis (BEA) estimated that the economy grew just 1.3 percent in the second quarter* (annualised) and what is more the BEA also made sweeping revisions to prior data points, which saw Q1 ’11 GDP growth be revised down to just 0.4 percent.
Chart 1: Gross Domestic Product
Given the weakness underlined by the GDP report and stressed in other reports such as ISM Manufacturing and Consumer Spending, it is constructive to keep an eye on what the forward-looking indicators tell us about the development in the coming quarters.
U.S. leading indicators
To investigate this we have gathered a collection of 25 time series believed to provide information about the economic growth in the coming quarters. We will offer a caveat up front, though, as the exercise forecasts GDP directly rather than the components. Hence unusual developments in government spending (roughly 20 percent of inflation-adjusted GDP) - for example elevated austerity - can alter the projected outcome.
From this basket of indicators we find the principal component representing most of the variance of the indicators one, two, three and four quarters ahead and regress this on GDP. In other words, by means of PCA we can represent the 25 indicators by a single series, which explains more than 50 percent of the evolution in the underlying indicators. This series can then be used to estimate GDP for all four forecasting horizons. To avoid being overly repetitive, we will only investigate the regressions for GDP two and four quarters out in detail.
Table 1: Regression information
| Coefficients* | ||
| Forecast horizon (quarters) | 2 | 4 |
| Constant | 6.27 | 6.41 |
| T-statistic | 13.02 | 14.26 |
| Principal Component | -1.12 | -0.94 |
| T-statistic | 8.78 | 7.54 |
| R-squared | 0.51 | 0.44 |
| R-squared, adjusted | 0.51 | 0.43 |
| F-statistic (df = 1, 73) | 77.14 | 56.82 |
| * multiplied by 1,000 | ||
The table above indicates that the parameters belonging to the constant and the PCA component are both significant at all reasonable levels with an adjusted R-squared statistic of 0.51 and 0.43, respectively. Moving on, let us have a look at the predictions visually:
Chart 2a: Fitted and predicted values two quarters ahead
Chart 2b: Fitted and predicted values four quarters ahead
Both charts more or less confirm our own outlook for the U.S. economy in the coming quarters. It will rebound from the weak performance in the first half of this year, but growth will not be impressive by any historical standards. If we dig a little deeper we get more of an understanding of why the models predict as they do. The 25 time series are represented in the first principal component two and four quarters out in the following way:
Chart 3: PCA loadings in first principal component
Unsurprisingly, several of the time series are well-known leading indicators often used when discussing the outlook of the economy such as ISM Manufacturing – New Orders and various regional Federal Reserve counterparts. We also note that the Conference Board’s (CB) own index of Leading Indicators enters into the principal component when we move out to four quarters and so does the ISM Manufacturing New Orders to Inventories differential. The differential is often used as a leading indicator, but it did fail in the summer of last year when it predicted more weakness than was actually the outcome.
What to make of it all?
Putting all the models together we get forecasts for the economy in the next four quarters, which suggest that even though there is definitely plenty of weakness the economy is not likely about to descend into recession again. Rather, the models predict that the U.S. will return to annualised quarterly growth rates of 2 percent in the near future with the yearly growth rate accelerating a bit to 2 percent as well.
Chart 4: Predicted Gross Domestic Product in the coming four quarters
We stress that the above prediction on the direction of the U.S. economy is only a qualified guess based on a basket of leading indicators; it is not resistant to neither renewed financial stress nor unusual government spending patterns. We also note that it does not take into account the coming revisions of GDP and other series.
Nevertheless, judging by the results shown above the economy is about to pick up a little bit of pace in the second half of the year, but not at all enough for us all to be running around with our arms above our heads.
source from: tradingfloor
No comments:
Post a Comment