April NFP Model Forecast Exceeds Consensus Estimate [View article]
Freddy,
Thanks. The NFP data does have a lot of variability from month to month, so it does help to have an objective model to help gain further insight into the strength of the employment market.
Earnings-Price Divergence Always Followed By Negative Equity Returns [View article]
Intuitively, the effect should show up in P/E ratios. However, the after-tax corporate profit data reported in the quarterly GDP release represents the data for the entire U.S. economy, adjusted for the inventory valuation and capital consumption adjustments. There are a lot of different definitions of earnings, which may or may not capture the same effects.
In addition, the P/E ratios for individual securities would need to be combined into an aggregate P/E ratio for all companies, using a consistent weighting methodology.
Finally, the above earnings-price divergence analysis looks at year-over-year changes, so the P/E ratio would also need to be evaluated over similar intervals.
I have not analyzed historical market P/E ratios for this purpose, so I do not know if the linkage is as strong as it is for extreme earnings-price divergences.
Earnings-Price Divergence Always Followed By Negative Equity Returns [View article]
Based on all quarterly data since 1960, the S&P 500 index reached its peak an average of 2.9 quarters after the extreme negative divergence reached its trough. In every case, the S&P index peaked on or after the date of the most extreme negative earnings-price divergence. If we apply the average 2.9 quarter lag to the most recent extreme divergence, the S&P index would reach its peak in July of 2013.
Obviously forecast the exact timing of market peaks is impossible, but the data make a strong case for seeing negative year-over-year equity returns in the next 12-18 months.
Trader Edge February Recession Model Forecast A Surprise [View article]
In some ways, it is easier to forecast recessions than it is to forecast GDP. The dependent variable is either zero or one, which keeps things pretty simple.
I have been working on a GDP model as well, but the effects of inventory changes and the variability in government spending make forecasting GDP a challenge. That said, the GDP model is currently predicting improving growth.
Trader Edge Non-Farm Payroll February Forecast Below Consensus [View article]
The household survey reported an increase of 170K jobs in February, with an additional 130K choosing to leave the labor force – hardly indicative of a strong economy. The decline in the labor force contributed to the drop in the unemployment rate.
The two month average NFP of +177.5K jobs is barely sufficient to accommodate new entrants into the labor force (when they graduate and seek full-time employment). That leaves millions of unemployed workers unable to find jobs or giving up entirely and terminating their job search.
I agree that even when the terminally unemployed find employment, the bulk of their paychecks would be required to pay down debt, which would not allow them to dramatically expand their standard of living via additional spending.
Income growth is a much better indicator of the economy than any of the employment numbers. Unfortunately, year-over-year income growth has been declining for many months. Spending growth cannot continue indefinitely without income growth.
Trader Edge Non-Farm Payroll February Forecast Below Consensus [View article]
The February NFP came in at +236, well above the model forecast of +151. However, keep in mind that the January NFP was revised from +157 to +119. The combined January and February forecast was +351 and the actual aggregate NFP for January and February was +355.
The decline in January NFP simply moved jobs from January to February.
U.S. Recession Risk Still Low In January [View article]
There is no question that economic stagnation can take place outside of a recession. I should also have emphasized that equities can suffer significant reversals outside of NBER recessions, especially when the market is significantly overbought and sentiment is overly bullish – as it is right now.
There is growing evidence to confirm an equity market trend reversal on the daily charts that is not yet supported on the weekly charts. If the trend reversal were confirmed, a material pullback would become increasingly likely.
Trader Edge Neural Network Model - Q4 2012 GDP Forecast [View article]
There were two major drags on GDP in the fourth quarter. Business inventories were drawn down and not replaced during the quarter. During periods of economic growth, this typically leads to increases in inventory in subsequent quarters, which spurs future GDP growth. As a result, this effect may be temporary, rather than the beginning of a new trend.
Government spending dropped by 6.6% during the quarter, which further dampened GDP growth. The decline in government spending could have been due to anticipation of the sequester. This effect may persist if the sequester goes into effect, which is looking increasingly likely.
Unfortunately, government spending and fluctuations in inventory levels are not captured effectively by the explanatory variables in the neural network model, which accounts for the atypically large error this quarter. Personal consumption expenditures (a major component of GDP) were up 2.2% during the quarter, which coincidentally was exactly equal to the model’s GDP growth forecast.
There is no question that P/E multiples could increase. However, as I mentioned in the article, according to FactSet, forward 12-month P/E ratios are now above their 5-year average.
In addition, higher P/E multiples can only be justified (in theory) by higher earnings growth rates and/or by lower interest (discount) rates. Treasury yields are approaching zero and have little room to decline further. Earnings growth rates could increase, but that would be challenging in the face of declining GDP growth expectations.
Unfortunately, I do not have access to FactSet’s historical data (beyond what is included in their monthly reports), but I suspect the period leading up to the 2001 recession would have demonstrated an extended period of equity prices growing far more rapidly than earnings expectations.
However, even in that period, it would have been unusual for earnings expectations to actually decline while prices were rising.
U.S. Recession Risk Drops Sharply In December [View article]
When the economic and market data reflect increased recession risk: 1) The diffusion index (the percentage of variables indicating a recession) should increase. 2) The recession slack index (the median standardized deviation of the current value of the explanatory variables from their respective recession thresholds) should decrease. 3) The aggregate recession probability estimate (the probability that the U.S. is currently in a recession) should increase. 4) The aggregate peak-trough probability estimate (the probability that the S&P 500 is between the peak and trough associated with a U.S. recession) should also increase.
While the models generate monthly values, these values should be evaluated relative to warning thresholds, similar to the ones proposed in the article.
The relatively large changes in the past two months were probably due to the transitory influence of hurricane Sandy on the underlying data.
April NFP Model Forecast Exceeds Consensus Estimate [View article]
Thanks. The NFP data does have a lot of variability from month to month, so it does help to have an objective model to help gain further insight into the strength of the employment market.
Brian Johnson
http://bit.ly/TmxWlb
Earnings-Price Divergence Always Followed By Negative Equity Returns [View article]
In addition, the P/E ratios for individual securities would need to be combined into an aggregate P/E ratio for all companies, using a consistent weighting methodology.
Finally, the above earnings-price divergence analysis looks at year-over-year changes, so the P/E ratio would also need to be evaluated over similar intervals.
I have not analyzed historical market P/E ratios for this purpose, so I do not know if the linkage is as strong as it is for extreme earnings-price divergences.
Brian Johnson
http://bit.ly/TmxWlb
Earnings-Price Divergence Always Followed By Negative Equity Returns [View article]
Obviously forecast the exact timing of market peaks is impossible, but the data make a strong case for seeing negative year-over-year equity returns in the next 12-18 months.
Brian Johnson
http://bit.ly/TmxWlb
Earnings-Price Divergence Always Followed By Negative Equity Returns [View article]
-Silver futures down 10% overnight,
-Gold futures down 6% overnight, and
-Crude oil futures down 3% overnight.
Stunning.
Brian Johnson
http://bit.ly/TmxWlb
Trader Edge February Recession Model Forecast A Surprise [View article]
I have been working on a GDP model as well, but the effects of inventory changes and the variability in government spending make forecasting GDP a challenge. That said, the GDP model is currently predicting improving growth.
Brian Johnson
Trader Edge Non-Farm Payroll February Forecast Below Consensus [View article]
The two month average NFP of +177.5K jobs is barely sufficient to accommodate new entrants into the labor force (when they graduate and seek full-time employment). That leaves millions of unemployed workers unable to find jobs or giving up entirely and terminating their job search.
I agree that even when the terminally unemployed find employment, the bulk of their paychecks would be required to pay down debt, which would not allow them to dramatically expand their standard of living via additional spending.
Income growth is a much better indicator of the economy than any of the employment numbers. Unfortunately, year-over-year income growth has been declining for many months. Spending growth cannot continue indefinitely without income growth.
Trader Edge Non-Farm Payroll February Forecast Below Consensus [View article]
The decline in January NFP simply moved jobs from January to February.
U.S. Recession Risk Still Low In January [View article]
There is growing evidence to confirm an equity market trend reversal on the daily charts that is not yet supported on the weekly charts. If the trend reversal were confirmed, a material pullback would become increasingly likely.
Trader Edge Neural Network Model - Q4 2012 GDP Forecast [View article]
Government spending dropped by 6.6% during the quarter, which further dampened GDP growth. The decline in government spending could have been due to anticipation of the sequester. This effect may persist if the sequester goes into effect, which is looking increasingly likely.
Unfortunately, government spending and fluctuations in inventory levels are not captured effectively by the explanatory variables in the neural network model, which accounts for the atypically large error this quarter. Personal consumption expenditures (a major component of GDP) were up 2.2% during the quarter, which coincidentally was exactly equal to the model’s GDP growth forecast.
Trader Edge Neural Network Model - Q4 2012 GDP Forecast [View article]
There were some unusual events in Q4 (such as Sandy) that may or may not have been effectively captured by the forecasting models.
I only forecast GDP for the trailing three months, but the general consensus seems to be for improved GDP growth (and earnings growth) going forward.
Price-Earnings Divergence Suggests Market Pullback [View article]
In addition, higher P/E multiples can only be justified (in theory) by higher earnings growth rates and/or by lower interest (discount) rates. Treasury yields are approaching zero and have little room to decline further. Earnings growth rates could increase, but that would be challenging in the face of declining GDP growth expectations.
It will be interesting to see how it plays out.
Price-Earnings Divergence Suggests Market Pullback [View article]
Price-Earnings Divergence Suggests Market Pullback [View article]
However, even in that period, it would have been unusual for earnings expectations to actually decline while prices were rising.
U.S. Recession Risk Drops Sharply In December [View article]
1) The diffusion index (the percentage of variables indicating a recession) should increase.
2) The recession slack index (the median standardized deviation of the current value of the explanatory variables from their respective recession thresholds) should decrease.
3) The aggregate recession probability estimate (the probability that the U.S. is currently in a recession) should increase.
4) The aggregate peak-trough probability estimate (the probability that the S&P 500 is between the peak and trough associated with a U.S. recession) should also increase.
While the models generate monthly values, these values should be evaluated relative to warning thresholds, similar to the ones proposed in the article.
The relatively large changes in the past two months were probably due to the transitory influence of hurricane Sandy on the underlying data.
U.S. Recession Risk Drops Sharply In December [View article]