Standard and Poor's Statistical service has estimated future earnings for the S&P 500 that are wildly optimistic. As reported earnings are estimated to rise 21% from September 2013 to September 2014. Operating earnings are estimated to go up 15% in that time period.
Earnings are unlikely to rise anywhere close to the estimates shown in the chart below and could easily decline for the following 4 reasons. First, the improving budget deficit is now a headwind to earnings. In addition, earnings are far enough above trend that they are more likely to be flat or down than grow. Furthermore, economic growth does not support such rapid growth. And finally, profit margins will likely decline.
The chart above shows "as reported" and "operating" earnings since 1988, along with estimated future earnings.
Improving Deficits a Headwind
A couple of years or so after big budget deficits earnings tend to soar. When deficits improve earnings tend to decline after the lag. Below is a chart with as reported earnings and the Federal deficit as share of GDP.
The above chart shows the timing of the headwind, but a look at de-trended data probably gives a more accurate picture. The chart below shows the annual growth rate in operating earnings and the one year difference in the deficit as a share of GDP.
The last point on the red line, with a value of 1.82, shows that the deficit has improved from 6.84% of GDP to 5.02% during the last year. This point in the context of the correlation shown in the chart suggests operating earnings will decline about 6% in the year ending in the third quarter of 2015. In the quarter now being reported, Q4 2013 the deficit shifts to being a headwind for earnings. It was a slight tailwind inQ3.
Below is a look at the correlation between the deficit and annual real as reported earnings. While operating earnings have only been released since 1988, as reported earnings have a long history.
The correlation optimizing lead time for real as reported earnings is three months less than operating earnings, so the implied decline for as reported earnings is a bit more immanent.
Earnings Above Trend
Earnings are far enough above trend that they are likely to be flat or down in the next 12 months. The long term growth rate in as reported earnings is about 6.2%. Long term forecasts higher than that likely ignore inevitable downturns. The last quarter with complete earnings data is Q3 2013. For the year ending then the S&P 500 had earnings of 94.37. This is shown as the last point on the black line in the chart below.
I start the modern trend for earnings in 1937 when I believe the modern inflation trend started. The green best fit trend line is useful but it is subject to change. For example, when I fitted the data from 1937 to 2000 the best fit line sported a 6.5% growth rate. The last 13 years has pulled the trend down.
To get a more complete perspective I also use single exponential smoothing ("SES") shown in the blue line of the chart above. Since 1950 earnings shown in the black line have averaged being 226% above the smoothed earnings. The current earnings of 94.37 is 281% above the corresponding last blue point. This value is further above the smoothed earnings than in about 77% of history.
In the chart above the green bar shows when earnings percentage rank has been from 70% to 85%. The last point on the blue line at 77% is smack in the middle of the bar. From the range of this green bar earnings averaged growing 0.1% in the next 4 quarter.
You may have some questions about how SES works. If so read this paragraph, if not skip ahead. I use a high smoothing constant of 0.002 I multiple the latest data point by 0.002 and the smoothed data point from the month before by 0.998 (1 - 0.002). So for example to get September's value I multiplied 94.37 times 0.002 and added it to August's smoothed point of 24.61 times 0.998 to get 24.75 (94.37 x 0.002 + 24.75 x 0.998). To begin using SES you need an estimated beginning value. So back for the first point in 1871 I used 0.2. The initial estimated value makes a lot of difference in the first few years but virtually none after about a decade. I use monthly data. Earnings for the two months between quarterly data points are interpolated from the quarterly data points the same way Robert Shiller does to get monthly earnings data. With a smoothing constant of 0.002 the SES is about as smooth as a 50 year moving average, but since each monthly data point is weighted heavier than the one before it, the SES usually responds to a change in trend faster than a 1 year moving average.
For the last few years using real earnings probably gives a more valid comparison than nominal earnings. Inflation has annualized 1.7% in the last 6 years, but annualized 4.1% in the 50 years before that. So, real earnings are even more stretched than nominal earnings.
In the chart above real earnings shown in the black line average being 76% above the real SES earnings. September's real earnings are 133% above the smoothed earnings. This ranks at 95%. Real earnings are only this stretched about 5% of the time.
The blue bar in the chart above shows when earnings have a percentage rank above 90%. The last data point in the green line is right in the middle. From the above 90% level shown in the bar, real annual earnings average declining 6% during the following 4 quarters. The best they have ever done from that level is rise another 16% the worst was a decline of 54%. Achieving S&P's estimate that as reported earnings will rise 21% would require outdoing something that has only happened one time in the last 64 years.
No Economic Support
Sustained earnings expansion generally requires GDP to grow above about 2.7%. This was not a problem prior to 2000 when GDP had annualized growing 3.6% for the previous 80 years. However in the last 7 years when GDP has only annualized growing 1.1% earnings have been volatile but are not likely sustainable. The chart below looks at seven year annual GDP growth shown in red and seven year growth of annual real as reported earnings in black.
Earnings growth has bounced more wildly above and below the growth rate indicated by GDP in the last 15 years. The unusually large plunge in 2009 and rebound in 2010 of the 7 year rate of earnings growth appears to be following the downward path suggested by GDP. The decline in the 7 year rate of earnings growth was slightly interrupted in Q2; it started back down in Q3. The weak seven year rate of GDP is a strong negative for earnings in each of the next several years. However, it is not a precise timing indicator that could be relied on to predict earnings will plunge this year. While volatility of earnings puts on a dramatic show, the underlying economic support is anemic.
Profit Margins Will Decline
Interest rates and the normal oscillation in profits should reduce margins going forward.
Profit margins have been running at a high and probably unsustainable level. As a proxy for this here is a chart of corporate profits relative to GDP.
Corporate Profits with Inventory Valuation Adjustment and Capital Consumption Adjustment are at 12.6% of GDP, a level only seen in two other quarters in the last 66 years. So 3 out of 266 quarters puts this at the 99% rank.
Interest rates apparently have a significant influence on how big the profit margins are. The correlation of the log of the interest rate is statistically more significant than the linear correlation. The chart below shows the correlation.
The 15 month lead time of the 10 year yield and its all time low yield of 1.53% in July 2012 suggest earnings hit a peak share of GDP in October 2013. It also implies the ratio will plummet through March 2015. Even if the decline in yields so far in January proves lasting the 15 month lead time suggests that won't help profits until April 2015.
In analyzing profits share of GDP there appears to be roughly a four year pattern where a big move one direction leads to a big move in the opposite direction four years later. To show this I looked at the four year difference.
The last point on the black line shows a 2.5 difference between the 12.6% share of GDP in Q3 2013 and the 10.1% share in Q3 2009. The red line shows this same point plotted using the red scales with an inverted vertical axis and a horizontal axis pushed forward 4 years. The red vertical axis is scaled according to the correlation the difference has with itself 4 years later.
It became clear the four year oscillating pattern became much stronger in about 1995.
The correlations in the two charts above both suggest the momentum behind profits rising as a share of GDP has peaked and started down and that this downward momentum will hit a low point at the end of 2016. While the momentum has turned down there is not a precise indicator when the margin itself will turn down. However, the likelihood of margins turning down increases with every quarter in the next three years and appears likely to start sometime in 2014.
The increasing volatility of earnings the last 20 years and the more defined 4 year oscillation may be an artifact of CEOs chasing performance pay. In 1993 corporations were limited in how much they could charge off for executive pay unless it qualified for an exemption as performance based compensation. Many CEOs only have to hold their options for three years. This creates a head I win tails you lose scenario where CEOs can take big leveraged bets. If the bets pay off well for a few years they can hit the pay bonanza. They get a sweet deal even if the bets eventually crash the company and the economy.
The game involves timing expenses and revenue so that there are multi-year periods of rapid growth interspersed with dismal results which set a low base for the next multi-year period of rapid growth. I suspect that if the performance compensation was required to be based on periods of 10 years or more that the volatility of earnings would go back to the level it had prior to 1993.
In the last 3 years the rate of earnings growth has fallen from triple digits to single digits. Real earnings are only this far above their long term trend about 5% of the time. Profits, historically, only have this big a share of GDP about 1% of the time. Improvements in the budget deficit suggest annual earnings now have a headwind and that the annual growth rate for earnings of the S&P 500 goes from positive to negative in the fourth quarter of 2013. The rising interest rates of the last 15 months suggest profit margins will fall from October 2013 to at least March of 2015. Historically, GDP needs to grow at 2.7% or above for sustainable earnings growth. The 1.1% GDP growth rate of the last 7 years implies the recent rapid earnings growth is the fruit of volatility rather than strength and that a volatile decline lies ahead. A regular pattern of oscillating between strong growth in earnings and rapid declines in earnings has developed in the last 15 to 20 years and now points toward decline.
While it is uncertain when earnings will begin a decline, the risk of decline increases with each passing quarter. The official Standard & Poor's double digit growth estimates for earnings in 2014 are extremely unlikely. Annual earnings growth will likely fade from the current single digits to negative sometime in 2014. Stock prices (SPY) will fall when earnings falter.
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.