Introduction
.
Worries are mounting, but the market keeps rising and it is not just stocks, as other asset classes have charged ahead as well. We are entering the later stages of a cycle when risk rises; however, if earnings growth remains robust, the market normally ignores red flags and rises. We may be safe for the next six months as double-digit earnings growth is realized; although, a 5% to 10% correction is about due. The market normally discounts growth about six months ahead, and in 2H 2018, EPS comparisons become much more difficult.
.
The steps to the end of a cycle share some commonalities. The cycle begins with fundamental improvements. Then, people borrow to invest. Third, financial markets get carried away and this propels fundamentals further as people become excited, spend, and invest even more. Doubt eventually arises, but markets continue their run. Finally, markets burst and cycles end. The current cycle is at least in the third step and perhaps moving to the fourth. We are several years into the current recovery and earnings are rising (step 1). Leverage is up (step 2). Markets are excited (based on returns/valuation) and sentiment is high (step 3), and we are overdue for a correction and volatility is low (step 4).
.
Step 1: Fundamentals are Rising
.
Earnings Growth is a Key to Returns
.
When earnings growth is positive, the market has had a 73% chance of being positive since 1873 (figure 2). Since 1990, the probability was 92%. EPS growth is positive 63% of the time since 1990 and 78% of the trailing twelve months returns were positive as well. Positive revisions are also good for returns (figure 3) and current revisions are up, plus for the next year, earnings growth is expected to be double digits (figure 11). Thus, betting today against positive returns is a low win strategy based on history. However, one should be wary of signs of deterioration since the S&P only leads earnings growth by about six months (figure 4). Some people claim the current elevated P/E is a reason to worry. I agree since it implies high expectations, but an elevated P/E does not necessarily lead to reduced probability of positive returns (as long as growth is positive) (figure 2). Returns were positive 68% of the time since 1877 when growth was positive and the P/E was above its five year average, which is higher than its normal batting average (64%). Since 1873, P/E has risen only 34% of the time when earnings growth was positive (figure 2), so the rising P/E this year was unusual (figure 5).
.
.
Returns are also driven by expectations. Expectations can be quantified in the P/E ratio. A higher P/E implies expectations for rising earnings growth, lower risk, etc. Historically, changes in next twelve-month P/E explained nearly 70% of annual returns of the S&P 500, whereas EPS growth explained 32% (figures 6 and 7). Of course, a high P/E can fall, and if it does, as is normal, then this implies muted positive returns.
.
.
2016 and Projections
.
Figures 8 and 9 show how changes in P/E have been driving returns since 2012, and most of the return in 2016 as earnings growth was negative. 2017 started out the same, with P/E contributing more to returns than earnings growth; however, earnings growth became the main driver in Q2. 2Q 2017 is being driven by earnings, and healthy earnings that are supported with sales growth (figure 10).
.
.
While small cap margins are still off their prior peak, large cap margins are at a high (figure 12). Thus, sales growth is needed to continue propelling earnings forward. Earnings growth was up 13.8% in 1Q 2017, and consensus numbers imply it will be up 6.9% in 2Q (figure 11). The final numbers for 2Q are likely to rise about 10% given the strong beats so far in the quarter (figure 13). However, 2017 EPS through 3Q anniversaries weak numbers in 2016. Thus, comps have been easy this year. As we move into 2018, comps will become much more difficult as the market must grow off of strong 2017 earnings.
.
.
Drivers for Growth in 2018 and Beyond
.
Plus, 2018 earnings will benefit less from the improvement in energy. The energy sector is expected to be up 240.0% in 2017 (numbers are coming down as shown in figure 15) and contribute 2.6% to growth this year. Growth in 2018 from energy falls to 41.5% and only contributes 1.1% to total S&P 500 growth in 2018. This number is highly dependent on oil prices which lead earnings by about a year (figure 14).
.
2018 earnings growth estimates may be tough to achieve, but we have a new tailwind as growth outside the US picks up. Europe and emerging markets are rebounding, and their earnings are well off their cyclical highs (unlike the S&P 500’s). 30% of S&P 500 sales are generated outside the US. PMIs for June are higher than their 12-month average for most of the major countries (figure 16). The dollar is flat over the last year, and it is negatively related to earnings growth (figure 17). A weaker dollar is a sign of world strength, and a synchronous economic recovery is good for sales and earnings. Of course, a strong dollar headwind in 2015-16 that turns to a tailwind will help FX translation. The dollar and commodity prices are inversely correlated, and improving commodity prices will aid commodity oriented emerging markets, boost earnings for the energy and materials sectors, and perhaps aid industrial stocks that are dependent on these sectors (e.g., CAT reported strong earnings in 2Q).
.
.
Fixed investment spending has been weak and has weighed on earnings and economic growth. It fell off substantially in 2016, partly in response to low oil prices (figure 18). It is unusual for the economy to avoid a recession when fixed investment retreats by the levels experienced last year. Overall, cumulative fixed investment has lagged prior recoveries as a percent of GDP (figure 19). Pundits complain that this lack of investment is a reason productivity growth has been so low. They also claim that instead of investing, corporations are focusing on the short-term benefits of share buybacks. This is only partially correct.
.
While they are correct that firms have invested quite a bit in share buybacks (and with debt), they are at least partially wrong about firms forgoing investment. First, share buybacks have been substantial. They made up about 2% of market value (figure 20). Management, unfortunately, is very good at buying high (figure 21), and this time they were likely doing so by raising debt (figure 22). However, share buybacks are now slowing (figure 21) and this is headwind to EPS growth. The claim that investment has waned is not entirely true. While capital expenditures are weak (figure 23), capital expenditures plus R&D is relatively robust (figure 24). It is almost 8.0% of sales, which is near its high (about 8.5%) and probably would be at the high if it were not for a decline in energy-related investments. Thus, firms have shifted their investments from capital, which is often associated with making more of the same product, to R&D which can result in new ways to make products and new products that can boost growth (or cut costs).
.
.
Where are We in the Fundamental Cycle?
.
This economic cycle is old, but it is not necessarily unhealthy. The body appears to still have years to go. The current expansion is 96 months old, longer than the last seven expansions which average 70 months (figure 25). However, real GDP has only expanded 17.5%, versus an average of 22.9% (figure 26). Plus, the recession was deeper so one would expect the expansion to be more robust. As discussed earlier, fixed investment is lower than past expansionary periods, but so is personal consumption (figure 27). While unemployment is down more than normal (figure 29), employment (non-farm payrolls) is up less than average (at the median) (figure 25). Thus, both the consumer and corporate sectors have room to expand further based on past cycles; although, they may be in the 7th inning.
.
.
The Feedback Loop
.
Moving from the 7th inning to the 9th may be dependent on confidence (see figure 30).
.
Both consumer confidence and business confidence are up (figures 31-2). Confidence, unemployment, wages, the stock market, and spending are correlated. While high-income consumers – with the most discretionary spending power – benefit from a rising stock market and spend more as a result (figure 34), as spending rises corporations hire workers which lowers unemployment and lifts wages for the remainder of the work force (figure 33). Wage growth is subdued overall, but wages for the younger work force are rising over 7% according to the Atlanta Fed. Higher wages and lower unemployment boosts spending further, so it is not necessarily a negative for margins (figure 35). Of course, higher consumption will be correlated with business confidence and this may drive up capital expenditures. Capital expenditures plans are up (figure 32). We may finally be at the point where we have a self-feeding cycle, which is also why the Federal Reserve may be comfortable with raising rates and starting to reduce the size of its balance sheet. Plus, the financial sector’s balance sheet is much improved. Total debt / total assets was 10.3% in 2016 versus 58.8% in 2007 (figure 39) and financial institutions just received a healthy CCAR scorecard. Lending growth, which has been subdued perhaps due to demand and supply, could rise (figure 36) (the current weakness is worrisome).
.
.
This self-feeding cycle could boost 2018 earnings; however, keep in mind that it will lead the economy to its last innings, and as noted earlier, the S&P 500 returns lead earnings growth by about six months. This feedback loop could be magnified by policy moves such as tax cuts (a 5% tax cut could boost EPS by over 7%), reduced regulations, and infrastructure spending; however, policy actions are looking less and less likely given the failures in 2017. Thus, policy missteps has the potential to be more likely to reduce confidence and growth than policy achievements can boost growth. Still, policy probably takes a back seat to unemployment, wages, and the stock market on confidence.
.
.
Step 2. Debt is Building
.
Cycles are normally accompanied by a build-up of debt. This cycle is no different. Overall nonfinancial debt to GDP is higher than ever (figures 37-8). It is interesting and alarming that to solve the financial crisis the Federal Reserve made it easier for people to borrow ever increasing amounts of debt at lower rates. Boosting the economy by encouraging borrowing and investment when times are difficult is prudent to jump start the economy. However, later in the economy, additional debt just borrows from future consumption which will be lower as a result. It also reduces an important policy tool during recessions, namely the ability to borrow to spend.
.
.
Let’s start with the good news. The financial sector is much improved. Figure 39 shows that since 2007, profitability (ROA) is back to near a high. The reason ROE is not likewise high is because debt ratios have fallen significantly. Thus, the ROE is a higher quality (lower risk) ROE. Also, regulatory costs and low interest rates are pressuring margins and financial institutions stand to benefit if these regulations are lessened and as interest rates rise. Already, the recent solid CCAR scorecard is allowing banks to boost returns to shareholders (dividends). Delinquencies are also low, except for student loans (figure 40). While low, delinquencies are beginning to rise for autos as sales decline from a peak. The auto industry is important to the economy, so this is something to be watched closely.
.
.
.
.
While the financial sector has cleaned up its act, all other sectors of the economy – business, government, and consumers – have issues.
.
While consumer plus mortgage debt to GDP is better than before the financial crisis, it is still high (figure 43). Plus, the financial obligation ratio is still elevated despite lower rates (figure 44). This FOR is the ratio of payments of debt (mortgage, credit cards, property taxes, rents, insurance, and auto lease payments) to total disposable income. It currently stands at 15.5%. While down from a high of 18.1%, imagine (e.g., 18% or higher) what it would be if interest rates were 2% higher. One can see that consumer debt to GDP normally peaks before recessions, and it is at a peak (figure 42).
.
Business debt to GDP is at a high, and it also tends to peak before recessions (figure 46). While it is not necessarily high versus profits, profits are elevated (figure 48).
.
Of course, Federal debt to GDP is high and we continue to run a healthy deficit late in an economic cycle (figure 49). What will it be during a recession?
.
Step 3: Reflexivity and Extrapolation Well Underway
.
.
Markets have become excited. Many people believe equity is extended as the market has been strong (figure 50), the market cap/GDP ratio is high (figure 51), the P/E is high (figure 52), and the Shiller CAPE ratio is high (figure 53).
.
.
While it is true that these rising ratios suggest the market is more excited than in the past, it does not necessarily mean it is overvalued. Viewing these ratios in isolation without considering underlying fundamentals is flawed. Plus, as shown earlier, as long as earnings growth is positive, the probability of positive returns is high. Including fundamentals with these ratios makes them look somewhat better – still not cheap, but also not as extreme.
.
Figure 54 shows that price rises with earnings. Figure 55 shows that a high P/S ratio (i.e., figure 51) is justified by a high profit margin. Although, margins are elevated (see figure 12) which may not be sustainable. A high P/E can be justified by low inflation and low “perceived risk” (risk is higher now late in the cycle, but most perceive it as lower) (figures 56-7). Plus, if rates rise, P/E can expand even further as it could indicate subsiding risk. However, P/E numbers above the current level were generally associated with recessions, and in a recession the market could fall substantially as earnings decline so this is not much of a comfort.
.
.
While Shiller adjusts for abnormally high earnings with his CAPE (Cyclically Adjusted P/E) ratio (figure 53), this ratio is still incomplete. He divides price by a moving average of earnings over 10 years to normalize earnings. However, as shown above, more than earnings influences P/E. It is also driven by inflation, the equity risk premium, and the payout ratio. The Cyclically Adjusted Implied Growth Rate (CAIGR) normalizes for these variables and earnings, and suggests that the market is implying 5.82% long-term earnings growth (figure 57). While this is low versus the past and could indicate the market is cheap, it appears more reasonably valued when one also considers that earnings growth has slowed. Absent a revolutionary new impetus that increases productivity, growth is expected to remain subdued to mid-single digits due to lower population growth, an ageing population, and low inflation.
.
.
Not only are stocks performing well, but other assets and expectations variables are moving in a manner consistent with a rising stock market. For instance, credit spreads (Bank of America Merrill Lynch US High Yield – US 10-Year Treasury) is 3.81%, versus an average of 5.33% since July 1990, a high of 18.60% during the financial crisis, and a low of 2.78% during the peak of the financial bubble. Figure 58 shows the correlation of 13 factors to S&P 500 returns. Each factor was converted to a percentile based on its value versus its historic values and whether it relates positively or negatively to stock market returns (e.g., a low value of a factor with negative correlation was converted to high percentile, and high value for a factor with a positive correlation earned a high percentile). Figure 59 shows how the index moves with the S&P 500. The index level is high, and reversals tend to be associated with market corrections. Besides (1) earnings growth and (2) revisions, the index includes (3) interest rates, (4) yield curve factors, (5) the high yield credit spread, and (6) alternative market variables (dollar, commodities, and gold). If each category has a 100% weight, then the maximum is 600%. The current level of about 400% (figure 61) is similar to highs made when Greenspan uttered his famous remarks that the stock market was irrationally exuberant, at the peak of the internet bubble, at the peak of the financial bubble, and during the early years of the recovery following the crisis. Since the index can be impacted by outlier variables with high levels, breadth was also considered. Not only is the index high, but the breadth of factors versus their prior year levels is above average (figure 60).
.
.
A sentiment index using a combination of fundamental – business and consumer – and financial market – equities, credit, and alternatives – variables yields the same conclusion. Sentiment is high (figure 62). As can be seen, over the last year sentiment has moved from a level associated with low returns to a level associated with higher returns, and returns have followed right on schedule (figure 65). A high sentiment index is more likely to decline (53 of 72 months). Future six-month returns are low – but positive – if sentiment declines from the current > 60 reading regardless of whether earnings growth accelerates or slows. However, if sentiment – however unlikely – rises from these levels, then returns over the next six months should be in the upper single digits (if history repeats). See https://coachinvesting.com/2017/03/26/sentiment-rises-modestly-cautious-readings/.
.
Step 4: Doubt but Escalation – Mark that Box!
.
Turn on the news and read a financial paper, and you will likely hear people talking about how volatility is low and how it has been a long time since we had a correction (i.e., much doubt). Yet, we still have low volatility and rising markets (i.e., much escalation)(figure 68). We have experienced 274 market days since a 5% correction and 378 since a 10% correction (figure 66). Over the last 50 years, we have spent 59% of the time 5% or more off a high. The average duration before a first break of 5%+ below a high (without making a new high) is 303 days. When doubt becomes doubt (instead of escalation), the correction could be substantial given that sentiment is high, valuation is elevated, and we are in the latter stage of the expansion. However, fortunately earnings growth continues to deliver. If this falters, alarm bells will be ringing.
.
