Saturday, August 2, 2014

Weekly Market Summary

Despite the large fall this week, the longer term technical picture for SPX is positive. We will discuss the shorter term perspective in a moment, but for traders with a long time horizon, a return to the recent highs is odds-on.

July ended a streak of 5 positive months in a row. Since 1982, that type of strength has never marked a long term top.  The worst outcome was a sideways market for several months (2004 and 2005). When the market fell, it was short and the losses were reversed. For investors fearing a crash, that would be out of character with prior tops. If a top is being put in, price will oscillate positive and negative over the next several months. A retest of the recent high is more likely to happen first.

That said, equities are not currently in an ideal set-up:
  • Trend: Aside from NDX, equity markets in the US and Europe are all under their key 50-day moving averages (dma). Small caps are net flat over the past 10-months and the Dow is negative for 2014. 
  • Breadth: Individual stocks are just as bad: only 30% of SPX companies are above their 50-dma. 
  • Valuation: Valuation premiums are excessive: we place fair value for SPX under 1900 by year end (post). 
  • Sentiment: Fund managers and advisors are excessively bullish, a contrarian warning (post). 
  • Seasonality: The next 3 months are renowned for being unfriendly to equities. 
  • Volatility: Low volatility continues to be a consistent positive for future returns.
Stating the obvious, US equities have been continuously bid higher in the past year despite similar headwinds.

So what happens next? Until the character of the market changes, the short answer is markets are likely to first bounce higher. There is some evidence that markets will then probably move lower over the next few weeks before eventually moving higher in the autumn. The risk/reward for longs is positive, but it's not risk-free.

The context for the current drop is that SPX had approached 2000. In the past, price has reacted by a minimum of 3% at each "round number." In this respect, the drop this week was normal.

In the past 10 instances, 7 have have found a low near their weekly Bollinger mid-band. Friday's low was just 5 points above this level (1911) and the total drop so far is 3.7%. In other words, the drop satisfies the minimum criteria.

A second part of the context for the drop is that volatility had become unusually compressed in July. We have shown this chart several times in the past: when the SPX Bollinger bands become very tight (lower panel), an expansion in volatility is typically close behind (middle panel). The recent range compression was the most extreme since January 2007. A small warning is that SPX fell 7% in February 2007.

Several positives suggest a bounce is near at hand.

First, Thursday's volatility spike was the second in the past 2 weeks. These spikes have a tendency to come in pairs. The second spike has been close to most bottoms in the past 18 months. A conservative buy signal has been when Vix closes back below its upper Bollinger (around 16 on Friday).

Second, we've discussed badly lagging breadth in the past few weeks. The number of stocks making highs has been diverging lower with each successive ATH in SPX. With the drop, just 30% are now above their respective 50-ema and only 18% above their 20-ema. In the last 4 years this has been close to a washout and has presented a positive risk/reward set up for longs. The one big exception was in 2011.

Third, focusing on SPX, Thursday's sell off pushed price right through its 50-dma for the first time in more than 3 months (since April). When this has happened in the past, the one-week returns have recently been a toss up but one month out, SPX has been 2.5% higher and 3 months out it has been more than 4% higher (data from Bespoke).

Fourth, at least short-term, sentiment has turned bearish. The Fear and Greed index (here) closed at 5 on Friday. This is close to a 3 year low. In the past, the bias has been strongly to the upside (yellow marks similar lows; this chart has not been updated since March).

Finally, the equity only put/call ratio spiked significantly higher on Friday. This is considered a "dumb money" indicator: when these investors buy puts, equities have a high probability of moving higher the next several days. In the table below, in the past 3 years, SPX has closed higher the next 5 days at least once more than 80% of the time. The few losses are small and risk/reward is heavily tilted long.

So, there are a number of strong positives for longs. This is a pitch worth swinging at. But it's not without risk; below are reasons to expect a second leg lower after a bounce..

Thursday's selling created a major distribution day (MDD). This means that sell volume was 9 times larger than buy volume on the NYSE. In the past, these have also come in pairs, most recently in January of this year. Think of this as downward momentum that takes time before it slows and reverses. There has been only one recent exception and that was in August 2013; importantly, this came after 4 weeks of selling whereas the current sell off is just 6 days old. This suggests further downside lies ahead, perhaps not immediately, but within the next few weeks.

Relatedly, the McClellan oscillator (NYMO) closed Friday at -89, an extreme. There have been 16 similar cases since October 2009. It marked the next 20 day low only half of the time; in the other half, the low was retested and exceeded in the coming weeks (table from Ricky Roma).

August seasonality is typically weak. The average return is the weakest of any month in the year and September is only slightly better (chart from Stock Charts).

This year is also the 6th in the presidential cycle. SPX has tracked this pattern quite well this year. Without being too literal, a July peak typically precedes a fall in August and into September (chart from Stock Almanac).

Without belaboring this point, we should also note that every mid-term election year since 1998 has seen the SPX fall below its 200-dma and turn negative YTD before rallying into year-end.

While short term sentiment is a positive, longer term sentiment is not. In July, fund managers established their most aggressive long position since February 2011 and their second most aggressive since 2001. When investors have been this bullish, SPX has always fallen (chart).  Looking at the similar extremes in bullish sentiment (yellow shading), SPX fell a minimum of 5% over a period of 4-5 weeks from the time of the recent high (in this case, July 24). It then took an additional 5 weeks from the low before the high was retested.

If the current situation is like the others, SPY could fall further during the month of August and reach a low near 189 and perhaps lower to 183. While this seems like a large fall, recall that SPY fell 6% in January to 1 standard deviation below its 20-week average. A similar fall now takes SPY to 185. This is also the current 200-dma; RUT is below its 20--dma and the Dow is just 1% above its 200-dma.

Similarly, keep in mind that SPX made its most recent ATH just 6 days ago. The drawdown cycles have been getting shorter but 6 days would be exceptionally short. Since 2011, they have typically lasted at least 4 weeks from top to bottom.

Given the persistence of the current uptrend and weak seasonality, a more typical correction might be in store. This would likely catch most investors by surprise. 

The Week Ahead

NDX remains the strongest index. It ended the week on its 50-dma and on July-low support. There's no foul here. Should this support fail, break-out support from the March-May trading range is near 3700, 4% lower.

The drop in SPY also held well above its March-May trading range high (yellow). SPY ended the week on its mid-June low at 192. 

What is most notable is the volume by price pocket below (green arrow): SPY moved from 189 to 192 in one week at the end of May. There is very little trading history between these levels, making this tenuous support. The "ideal" low is to retest the prior March-May trading range high at 188-190. First resistance is 194.2 (weekly pivot) and first support below is 189.9 (weekly S1).

On the daily scale, SPY ended on a long term trend line (green), a positive. But daily momentum as measured by RSI finished at just 12. The blue vertical lines are similar instances: only 2 of 7 marked the final low and both followed several weeks of selling (unlike now). So while SPY is 'oversold', the watch out is momentum might carry lower.

The Dow, in comparison, is already back in its prior trading range. Its 200-dma has held prior corrections (green arrows) and it is also support from the May lows (16,300). This is just 1% lower. Like SPY, an RSI below 10 has often led to further weakness.

The weakest index is RUT. It is back in a range dating from last October. The bottom of this range is critical support: its 2% lower (1090). In 10 months, it has gained net zero. However, RSI has a positive divergence and MACD is setting up for a low. This bears watching closely. 

Investors' least favorite asset class, treasuries, remain in an uptrend much stronger than all the equity indices except NDX. TLT tested and held its 50-dma this week, despite strong GDP, NFP and corporate earnings reports. $114 remains critical support. 13-ema has inflected down, representing a loss of shorter-term momentum.

Unlike many months, August does not typically start strong. There's no tailwind from seasonality this week (chart from Sentimentrader). 

Our weekly summary table follows: