- The market rally is being driven by speculation, which has gone parabolic in recent years and is clearly unsustainable.
- Not once since the 1950s has earnings growth persisted for over six years, yet analysts are forecasting 8 consecutive years of growth.
- The trend in the market remains up but momentum is beginning to decline. This divergence is a worrying sign.
- Cyclical sectors such as Discretionary and IT, that led the market higher over the past 5 years are beginning to lose ground.
- Defensive sectors such as the Utility sector is gaining traction, signaling the caution that is starting to prevail in the market.
In the final few issues of 2013, I made my position clear. The US market was overvalued, potentially even in a bubble, but the trend remained upwards. Indeed, the trend still remains up but the warning signs are beginning to flash red. The winners of yesterday are fast turning into the dogs of the stock market, a signal that is consistent with market peaks. What follows is my take on current valuation levels, the strength of the current trend and the signals that have me concerned at this juncture.
Valuations: The Investors Indicator
The most popular metric for determining whether the market is under or overvalued is Price/Earnings. It is very rare that one of the institutional folk paraded in the media will suggest the market is undervalued without making mention of the P/E ratio. The S&P 500 currently trades on a P/E multiple of 17.1x. Based over the past 50 years, the multiple is just above the average of 16.3x. However, there are two issues I take with those who proclaim the market to be cheap currently.
- The numerator (i.e. price) is being fuelled by speculation
- The denominator (i.e. earnings) is misrepresented by just a single year of data
Price - The P/E Numerator
To deal with the numerator and speculation issues first. The US economy has been on life support for some time now and while I disagree with the hard wired capitalists that everything should have been left to crash in 2008 and the people left to rebuild the country from scratch, the level of manipulation has been quite staggering. Slowly, this manipulation is restoring confidence among lending institutions and consumers but most know that even the slightest of setbacks could derail any recovery and plunge the US back into recession. However, in their efforts to prop up the recovery, the Fed has yet again, fuelled a stock market bubble, or at the very least, the makings of one.
Didier Sornette, Professor of Entrepreneurial Risks at the Swiss Federal Institute of Technology, Zurich and also a professor of the Swiss Finance Institute, technically defines bubbles as:
The 'super-exponentially' accelerating rise of a price due to the progressively increasing build-up of cooperation and interactions between investors. The ascent due to positive feedback (also called 'pro-cyclicality') is the translation of the maturation towards an instability or critical point reached in finite time. It is a mathematical certainty that, when the growth rate itself grows, the process will be fundamentally unstable. According to this 'critical' viewpoint, the specific manner by which the bubble bursts and the prices collapse is secondary: a crash occurs because the market has entered an unstable phase and any small disturbance or process may reveal the existence of the instability.
This idea of unsustainable exponential growth is precisely what we are seeing now with regards to speculation in the markets. The attitudes of market participants are perverse in that bad news is greeted with cheers because it means the Fed is expected to come to the rescue. It has fuelled another credit driven boom in the markets where margin debt continues to make new records (Fig.1).
Fig.1 - S&P500 and NYSE Margin Debt
This chart should be enough to at least instill caution as it demonstrates how debt has been the primary driver of this market rally. It is also interesting to note that the market peaks tend to occur shortly after a parabolic like surge in speculation.
- As the markets marched upwards in the late 1990s, margin debt finally went parabolic, surging 53% between November 1999 and March 2000. Between August of that same year and the end of September 2002, it had lost 46% of its value.
- As the markets recovered and moved higher, it was not until 2006 before the next parabolic move in margin debt occurred. It jumped 68% between August 2006 and July 2007. In October 2007, the market peaked and went on to lose 53% of its value by the end of February 2009.
- The latest parabolic-like move in margin debt began at the end of July 2012 after the market had moved 88% higher off the 2009 lows. Margin debt is now 68% higher in the space of just over 1.5 years.
Therefore, it is evident that greed has once again engulfed the markets. Investors have become complacent to the risks, instead choosing to believe that the Fed backstop will be enough to send this market considerably higher. I believe their optimism is based on shaky ground and for this reason, I don't envisage an expansion in the P/E multiple unless it is driven by earnings growth.
Earnings - The P/E Denominator
Of course, this brings me on to the issue of earnings, the numerator in the calculation of P/E. After collapsing in 2008, earnings have rebounded strongly. Earnings growth is depicted in the chart below by light blue bars (Fig.2). Meanwhile, the red bars show the periods when earnings contracted and the dark blue bars are consensus forecasts for the coming three years (2014-2016).
Fig.2 - S&P500 Historical & Estimated Earnings Growth
Over the past five years, earnings have registered year-on-year growth. Furthermore, with the consensus suggesting three further years of earnings growth, investors could be forgiven for getting excited. That is until you consider the viability of the market registering a total of eight years of uninterrupted earnings growth, which based on data since 1951 is highly unlikely.
- Only on two occasions since 1951 have earnings registered five years of earnings growth.
- If earnings grow in 2014, it will mark only the second time in over fifty years that earnings have grown in six consecutive years.
- Not once since 1951 have earnings grown for more than six years without a correction and yet, the analysts are suggesting we are about to witness eight years.
In essence, what the Wall Street analysts are proclaiming is "this time is different", referred to as the four most dangerous words in investing by the legendary investor, Sir John Templeton. Personally, I'm not banking on this time being different and expect to see earnings correct at some point over the coming two years.
For this reason, I place little emphasis on one year of earnings data, whether it past, present or future. Instead, I subscribe to Benjamin Graham and David Dodds preference for using earnings smoothed out over many past years. The purpose of doing so is to remove the variations in the economic and business cycles.
Robert Shiller, a professor at Yale University, took this a step further by creating his cyclically adjusted price-to-earnings ratio (CAPE), using 10 years of earnings and adjusting for inflation. He also took a monthly average price of the index. Based on the historical data available to him, Shiller found the average CAPE of the S&P 500 to be approximately 16x. However, rather than spending time at the average, the market would often considerably extend beyond (overvalued) and below (undervalued) this average.
The following chart paints a telling picture, where the left-hand column illustrates the distribution of CAPE (Fig.3). Put differently, it represents the percentage of time that CAPE stays within that range.
Fig.3 - CAPE Distributions and Historical CAGR Overview
(click to enlarge)Source: mebanefaber.com
Some of the key observations are:
- The largest distribution is from 15 to 20 (31%). Most compelling though is that the market has spent 90% of its time, from 1881 to 2011, between a CAPE of 5 and 25.
- With today's reading of 25x, the market has historically only been more expensive on 9% of occasions.
- The current valuation level also suggests that the CAGR for the coming decade will most likely be in the region of 2.5%-3.5%
- However, this positive CAGR is likely to be driven by the performance in the latter part of the decade as the 5-year data suggests negative returns.
These findings are consistent with those of John Hussman, who writes an excellent free weekly commentary piece, which you can read here. Hussman notes that:
Historically, nominal GDP growth, corporate revenues, and even cyclically adjusted earnings (filtering out short run variations in profit margins) have grown at about 6% annually over time. Excluding the bubble period since mid-1995, the average historical Shiller P/E has actually been less than 15. Therefore, it is simple to estimate the 10-year market return by combining three components: 6% growth in fundamentals, reversion in the Shiller P/E toward 15 over a 10-year period and the current dividend yield.
Based on these findings, Hussman uses a simple formula to determine expected 10-year returns in the market:
1.06 * (15/ShillerP/E) ^ (1/10) - 1 + dividend yield
Applying today's figures to the formula, the expected return for the coming decade is
= 1.06 * (15/25) ^ (0.1) - 1 + 0.019 = 2.6%
The simple formula has proven to be incredibly accurate over a 10-year timeframe. However, Adam Butler and Mike Philbrick (Gestaltu.com) illustrated that CAPE is equally effective in determining the expected return over a 15-year period (Fig.4).
Fig.4 - CAPE and Expected 15-year CAGR
Not everyone is a fan of CAPE. Jeremy Siegel recently penned a piece in the Financial Times (Don't Put Faith in CAPE Crusaders), criticizing the use of CAPE, proclaiming that CAPE's pessimistic predictions are based on biased earnings data. While Siegel makes some interesting points about the effect of accounting changes in the 1990s, I believe the excellent James Montiers rebuttal (A CAPE Crusader: A Defence Against the Dark Arts) to be more accurate and I continue to use CAPE as one of my primary valuation indicators.
Based on this analysis, my outlook for the market is rather subdued based on my belief it is speculation rather than solid fundamentals that are driving this market higher. However, we have seen this happen before in the 1990s. So, rather than selling out of this market rally, I look to devise entry/exit signals. Ultimately, I am looking for triggers to alert me as to when the indicators I observe start to play out. This is where the attention switches to trends and trend following.
Trends: The Investors Trigger
There are three things I am trying to determine when it comes to the market trend
- Direction of the trend
- Momentum in the trend
- Condition of the trend
In the strategy piece later in this issue, I discuss trend following itself in more detail so let's get straight into the data itself now.
With regards to the trend direction, I use a price moving average crossover strategy. My preferred time frame is 12 months. The reason being I have backtested it since 1929 and found that over that period, it has outperformed the market. Furthermore, in the harrowing decade 2000-09, this trend trigger proved to be particularly robust.
Therefore, my acknowledgement to the trend being up in previous issue was in direct reference to the price being above the 12-month moving average. This remains the case today with the price of the market at 1,872 and the 12-month MA at 1,752 (Fig.5). I am watching this support closely.
Fig.5 - S&P500 Trend Direction
I use moving average convergence divergence (MACD) to determine the extent of momentum in the any market trend (Fig.6). From the lows of 2009 to mid-2011, the trend was relentless. However, it then turned and it did appear like the market rally was in danger of coming to a crashing halt. However, the market recovered once more and it was backed by another surge in MACD to record levels. While the momentum has continued to rise, the rate of advance has slowed in recent days, potentially exhibiting signs of exhaustion.
Fig.6 - S&P500 Trend Momentum
I use the relative strength index (RSI) as a gauge to determine whether the trend is overbought, neutral or oversold (Fig.7). Although markets can maintain extreme readings for extended periods of time, it is a useful gauge when considered along with the trend direction and momentum of the market. The S&P 500 re-entered overbought condition at the end of March 2013 and bar a brief period where it dipped below this reading, it has been overbought ever since.
Fig.7 - S&P500 Trend Condition
So, what I am seeing is the following:
- Up trending market
- Record levels of momentum with signs of exhaustion
- Overbought market conditions
Looking back at the markets since 1929, based on the parameters I use there has only been one other period in history where the market has been in an uptrend, momentum at extreme levels and the condition being overbought.
- That period was the end of February 1998. At the time the market was at 1,049 before a 9% correction ensued over the next six months, removing the overbought condition.
- Market then recovered and these extreme levels were reached once again at the end of 1998.
- Even then, a final blow-off period in the market occurred and it was not until October 2000 that the trend direction finally changed.
- What developed was a collapse of over 40% in the market until the trend turned up once more in April 2003.
This perfect cocktail mix has me concerned that the market could be susceptible to a significant sell off. It is still too early to sell because as I have stressed, for now the trend remains up. However, for some, they want to know what, if any, are the telling signs that the trend direction itself is ready to turn. For this, we turn our attention to the individual sectors and stocks.
Sector and Stock Trends
This bull market that has been in place since 2009 has been led by the cyclical sectors, particularly discretionary. This sector has been the best performer over the five year period from 2009-13, growing 213%. It has also been the best performer over the three year period 2011-13 (+79%) and again over the course of 2013 (+41%).
Digging down further, it has been the Discretionary subsector "Internet & Category Retail" that has led these gains. This subsector surged 811% from 2008-13, placing it as the best subsector among a total of 62. In 2013, this trend showed no signs of letting up. The almost 74% gain was only better by the biotechnology subsector.
However, there are signs that this phenomenal performance is coming to an end. The discretionary sector has shed 3.2% as of the close on the 9th of April and is the worst performing. Meanwhile, the Internet & Category Retail subsector has plunged 10.6%, also placing it as the worst performing subsector (Fig.8).
Fig.8 - Year-to-date performance of Discretionary sector and Internet & Category Retail subsector
There are five stocks within this subsector; Amazon (NASDAQ:AMZN), Priceline (NASDAQ:PCLN), Netflix (NASDAQ:NFLX), TripAdvisor (NASDAQ:TRIP) and Expedia (NASDAQ:EXPE). Amazon is by far the biggest of these. Their market cap of 152.4 billion accounts for 60% of the combined market caps of the five stocks. Therefore it is worth looking deeper into Amazon as this will have a large influence on the trend of the others.
Amazon's share price has risen 678% since the end of 2008 but thus far in 2014, has lost 16.8% of its value (Fig.9). Now, the price is threatening to confirm the break below the 50-week MA, suggesting the support at the 100-week MA could next be tested at $288. Momentum is also in free fall and looks to be heading into negative territory. Despite the 20% selloff, the stock has not yet reached an oversold condition so it should not come as a surprise if over the coming months, Amazon continues to head lower.
Fig.9 - Amazon share price including 50 & 100-week moving averages
The effect is being felt across the board. Priceline, Netflix, TripAdvisor and Expedia have lost 12%, 22%, 21% and 12% of their values respectively since late February/early March. This subsector is under pressure and when these momentum winners of yesterday start to become the losers of today, you can be sure that the speculators decline mutiny and abandon what they believe to be a sinking ship.
As mentioned earlier, the other sector that has performed so well in recent years is that of IT (Fig.10). It is the second best performer over the five year period 2009-14, having recorded impressive gains of 153%. The sub sector here that has driven these returns is that of Internet Software & Services. Over the period 2008-13, this subsector has returned 255%, placing it 7th among the list of 62. The gains kept coming in 2013 (+49%) but 2014 has been sluggish (-0.6%).
Fig.10 - Year-to-date performance of IT sector and Internet Software & Services subsector
There are six stocks within this subsector; Google (NASDAQ:GOOGL), Facebook (NASDAQ:FB), eBay (NASDAQ:EBAY), Yahoo (NASDAQ:YHOO), Akamai Technologies (NASDAQ:AKAM) and VeriSign (NASDAQ:VRSN). Google and Facebook account for 81% of the combined market cap so again, it makes sense to look at these in a little more detail.
Googles share price of $552 is still above the 50-week MA support at $501 (Fig.11). However, with momentum having turned sharply lower and the stock in a neutral condition, there is room for this $500 level to be tested. The share price has shed 10% of its value since later February.
Fig.11 - Google share price with 50, 100 & 200-week moving averages
Facebook, meanwhile, has lost 16% of its value since the 11th of March. Like Google, the 50-week MA support is still intact at $46 but momentum has now started to fall at the same time the company retreats from overbought territory. Provided these levels hold, the losses will be viewed as a healthy correction but the concern, as I have stressed is that if these momentum stocks begin to sell off, it will be exacerbated by fleeing speculators.
I should also note on the issue of sector and subsector trends. The defensive utilities sector has been one of the worst performing in recent years. This was particularly true of the electrical utility subsector, which has actually lost almost 10% over the five year period 2008-13. It was also the 3rd worst performing sub sector in 2013. However, thus far in 2014, it is the 3rd best performer having recorded gains of 11.3%.
One final thing that investors should consider is that 40% of best performing subsectors this year were actually among the worst 10% of performers in 2013. There are increasing signs that the trend is beginning to bend and that the outperformance of the cyclical sectors will be replaced by a bias toward the defensive sectors such as Utilities.
With regards to valuations, the evidence illustrates that the price of the market is being driven by speculation. With this speculation having gone parabolic in recent years, the risks are to the downside and as such, it is difficult to make the case for higher prices unless it is driven by earnings optimism. Unfortunately, the market is already pricing in earnings growth for a further three years based on analyst expectations. Should this transpire, it would mark eight consecutive years of earnings growth, something we have not seen over the past 50 years.
Meanwhile, from a technical perspective, the market trend is up but there are some worrying signs that this trend may be starting to bend. The momentum stocks that have led this market higher, such as those from internet related spaces, are finally coming under pressure after incredible runs since 2009. Furthermore, the defensive sectors that have lagged are gaining some traction, the utility sector in particular. Added to these concerns are the record levels of momentum, which ties in with the parabolic move in speculation and the overbought condition of the markets.
Timing is now everything in this market and despite even my best efforts, I have to acknowledge that I do not know when the market will ultimately correct. However, what I believe is undeniable is that we are indeed in for a rough ride once the tide starts to go out. As Buffett says "it is only then will we get to see who has been swimming naked."