Some Key Ex-US Indices Diverging from $SPX and $DJIA

Global equity markets tend in the same direction. The first chart below on weekly timeframe, comparing $SPX to $EEM, Euro 350 and All World Ex-US, makes this apparent. 

So, when divergences become apparent, it is at least a flag. In the second chart on daily timeframe, EEM has diverged since the start of January and Euro 350 since the start of February. Both are down 3-4% from their high while $SPX has moved up. 

What Happens After the Largest In-Flows Since 1996?

The last time in flows into equity mutual funds and ETFs was as large as last week's was January 1996.  The situation is remarkably similar to today: a strong run up in the market several years after the end of the economic recession. The start of the new year is strong as well. The good news about the market is featured on the front page of the Times and Barron's. Retail investors want to get involved.

This is what happened next:

US Sector Trends: Very Strong But Some Cyclical Weakness

Every week in the Weekly Market Summary, we comment on the trend of US indices and 9 $SPX sectors. Our benchmark is maintaining above a rising 13-ema. Below that raises a flag. Turning the slope negative or losing the 50-dma raises an alarm if it is widespread among sectors.

In the past two weeks we've seen some cyclicals not make new highs with the indices and also relative outperformance by defensives. These are worth watching closely.

Weekly Market Summary

Our overall market view remains positive. The trend is strong: for the 11th week out of the last 12, a majority of indices/sectors closed >13ema. This week: all 4 indices and 9 of 9 sectors.

Two cracks are starting to form in trend and breadth, however. Several ex-US markets are now <20-dma and diverging with the US indices. Within breadth, $NYMO is struggling at the zero line and $NYSI is diverging. This is just like February 2012. $NYHL is also diverging, moving lower while indices move higher. These divergences can persist for a month or so. But, unless corrected, these are likely to be late stage signals.

Sentiment and macro expectations remain the main headwinds.

Breadth Divergence Watch

One of the two main drivers of the Weekly Market Summary is breadth. It has been unmistakably strong since the November low. The chart below looks at $SPX in the top panel versus NYMO (middle) and NYSI (bottom). 

Generally, weak breadth (NYMO below zero) leads to weak prices. See September-November 2012 as an example. This week, NYMO returned below zero. If the stay is brief, then there is no problem. If this persists, NYSI (bottom) will start to fall and price usually follows. 

Note the long negative divergence between NYSI and price between February and March 2012 (yellow arrows). Eventually, it resolved to the downside with a 10% correction in $SPX. 

Insider Selling 9:1 - A 2% Correction Usually Follows

Ironic, in my opinion, that insider selling is at a high the same week as newletter writers recommend their longest exposure in 13 years. This is people with inside information versus those without.

Mark Hulbert: "The sell-to-buy ratio for NYSE-listed shares listed now stands at 9.2-to-1. The last time a weekly sell-to-buy ratio was worse than this was in late July 2011. Over the next couple of weeks, $DJIA lost some 2,000 points. On average over the month following each prior occasion when the sell-to-buy ratio got this high, the broad market fell by 2.1%. " 

Read the full text from MarketWatch here

All-Time High Newsletter Sentiment: Poor Returns Next 1-6 Months

From the excellent site SentimenTrader:

“The survey of investment newsletter writers from Hulbert Financial Digest is theoretically better than most, since it focuses on actual trading recommendations instead of sometimes-ambiguous opinions.”

“There have been a total of 9 weeks when the combined level neared 70% (a couple of them were clustered together).  A month later, the S&P 500 showed a negative return every time, a median of -3.1%.  Its maximum gain during the next month averaged only +0.1% (using weekly closes) while the maximum downside averaged -4.4%.

Even over the next six months, returns were poor.  Only 1 of the 9 weeks had a positive return (+1.9%), and the median was -4.25%.  The most that the S&P rallied during the next six months averaged only +0.7%, while its maximum decline averaged -13.8%.”

Junk Bond Extremes

Junk bond exuberance seems to be extreme:

1. The effective yield on junk bonds is now around 6%. This is an all-time low. During good times, the normal range is about 7-9%. See first chart below the break.
2. Part of this is explained by the low default rate of about 3% versus an average of 4.8%. Note, however, that the default rate a year ago was 2.3%. It's creeping up.
3. On a relative basis, these yields seem extreme. Oscar Schafer: "High yield bond yields are lower than the S&P's earning yield for the first time ever." 
4. Loomis Sayles’s Matt Eagan: "High yield bonds don’t typically trade at higher than par plus half their coupon." David Schawel: "Well, guess what? We are at that place today." Read his excellent take here.

Low Vix Means 5% Corrections Are Few And Far Between

In the past 20 years, there have been two long periods where $VIX was mainly below 20. During both of these periods, 5% corrections were few and far between. The chart below shows only changes in direction of >5%. Some observations:

1. Uncorrected uptrends during these times can last 4-12 months. 1995 went a full year. There are several lasting more than a full quarter. Low volatility has meant higher prices, long bull runs and shallow corrections.
2. The two periods ran from 1993-97 and 2003-07. A third period may be starting now, in 2013. Is this a 10-year cycle, lasting 4 years?
3. 1993 and 2003 came after economic recessions that corrected the market. We are now already in the 4th year of an uptrend. The starting point for this third era is very different.

Duration of Uptrends Since March 2009

Since the March 2009 low, there have been 8 discrete and completed uptrends. Some features:

1. Median length of 12 weeks, mean duration of 12.4 weeks, minimum duration of 10 weeks, maximum duration of 16 weeks.
2. 4 up weeks in a row is common. 6 has been a typical maximum. One was 7 weeks (2011). The current uptrend has completed 6 weeks.
3. Within an uptrend, it is common to have 2-4 red weeks. Overall, about 70-80% are up weeks.

Importantly, after a long uptrend up more than 5 weeks, the down week is usually minor (within the low of the prior week) and there is always a higher high ahead. Strong trends do not end after a long streak of up weeks.

Graham-Shiller PE10 At 85th Percentile

Doug Short (link) has a valuable post on the cyclically adjusted price-earnings (CAPE) of $SPX. Using smoothed earnings (average of the past 10 years) and smoothed price (average over the past month), $SPX is valued at 22 times versus a long term average of 16.4. This is in the top 85th percentile of all periods since 1871. In other words, its overvalued on this methodology. See the charts below the break.

The purpose of using 10 years of earnings is to reduce the volatility of a trailing PE due to the cyclicality of earnings. Because of this, however, over- and under-valued can persist for years. So, this is back pocket material and not useful for timing except, perhaps, at enormous extremes.

You can get a daily update here

Typical Topping Pattern: Not There Yet

Take a look at recent topping patterns as they are fairly typical. There's two steps to each:

1. Green arrows: After a long run, momentum starts to weaken and price tests a key trend line or moving average (here, its the Bollinger mid-band). 
2. Yellow area: Price pushes higher, falters, tries a third time and makes a double top (or lower high).

Look at where $SPY is now. It has not even tested a key moving average yet (part1). When it does, buy the dip and expect price to move higher before making a discernible topping pattern. If nothing else, look for an actual lower low before expecting a top to develop.