Saturday, November 1, 2014

Weekly Market Summary

After dropping to a 6 month low in mid-October, SPX has since risen more than 10% in two weeks. The bounce has been at a torrid pace.

We had been expecting a more complex bottom to form at the low. Why? The market had risen 60% since it's last 10% correction in mid-2012.  A more typical basing at the low, to reset sentiment and breadth, seemed warranted before resuming the uptrend. This now appears to have been wrong.

Instead, SPX, DJIA and NDX all rose to new bull market highs this week. SPX has now completed its 9th "v-shaped bounce" since 2013. According to Dana Lyons, "v-shaped bounces" occurred once every 1.6 years from 1950 to 2012; since then, they have occurred every 2 months (post). To us, that reflects an unflinchingly bullish and aggressive investor psychology. The data supports that view.

This market is well-known for doing the unprecedented. According to SentimentTrader, SPX traded more than 0.5% above its 5-dma for 10 days in a row in the past two weeks. In the prior 75 years, this has only happened twice before, both at bear market lows (1982 and 2002). In other words, a rare rip higher, that has only happened after multi-year bear markets, just occurred after a mild, four week drop. It's incredible and completely unexpected.

Perhaps the most distinguishing characteristic of the current rally is this: SPX has made a series of 12 daily "higher lows" in a row. According to Paststat, there have been only 9 other instances in the past 20 years where SPX has made more than 10 "higher lows" in a row (post). This raises the question of what typically happens next.

We normally think of strength like this as initiating a broader move higher. But that was the case in just 2 of these 9 instances: December 2003 and July 2009. Both of these came within a few months of a new bull market, and in both cases, SPX continued higher. December 2003 is shown below (left side of chart).

In the majority of instances (7 of 9), SPX struggled in the weeks ahead. If SPX continued to move higher, it quickly gave those gains back. In the chart above, November 2004 is an example (right side of chart). Over the next 4 weeks, SPX made no progress. In fact, it made no net progress over the next half year.

The same pattern is true recently. SPX made more than 10 "higher lows" in January, March and May 2013, a year renowned for its remarkable strength. But each of these preceded weakness. SPX was lower 4, 5 and 7 weeks later, respectively.

1995 is the year most like 2013. In 1996, SPX finally closed below its 200-dma after a long streak above (circle), similar to what has just happened in 2014. Later in 1996, SPX made a streak of more than 10 "higher lows" in a row. The market was lower 6 weeks later. After a move higher, it was also lower 5 months later.

Trader's expectations now are that December and January seasonality will override any negative factors; this might eventually be true, but that expectation should probably be tempered a bit. In two of the examples shown above (1996 and 2005), SPX struggled in December or January.

So, while the current streak in SPX, and the other indices, is impressive, the precedent has been for consolidation and/or weakness to follow. It may last only a week or so but it has also lasted much longer.

It's not just the stair step higher over 12 days that suggests limited immediate upside, it is also the pace of gains.

Below is an update of a QQQ chart from last week. The past 12 days surpassed any initial thrust out of a low the markets have seen in the past 4 years. The green shaded bars are all the same proportional size and equal the past 12 days' gain. In the past, the market has given back a third or more of those gains in the weeks ahead. Many times, the market did not move substantially higher over the next month or longer (red line).

Similarly, using SPX, the rate of change in the past 12 days is 8.3%. The only time in the past 4 years that has been exceeded was after the 20% fall in 2011 that ended a 5 month correction. Even then, those gains were subsequently given back (right side of chart). Two other instances were nearly as strong; the first was within the first year of the bull market (left side of the chart) and the other after the 4 month correction in 2010 (middle of the chart). In the best case, the market chopped for several weeks.

Breadth was washed out at the mid-October low. It was one of several reasons to expect a rebound over the next month. But now breadth has moved to the opposite extreme. Even after a low more significant than the one recently, consolidation and/or weakness followed.

Below is a chart using McClellan oscillator, NYMO (bottom panel). On Tuesday and again on Friday, it closed above 80. This is rare. Below are the 8 prior times in the past 5 years. In 2010 and 2011, all of the gains from the low were given up in the weeks ahead. The last instance was in September 2013 and initiated a 5% fall over the next 3 weeks.

According Andrew Kassen, the last 12 NYMO readings over 80 resulted in a median fall of 5% in SPX. And when, as happened this week, there were two NYMO readings over 80 in the same week, SPX fell by a minimum of 4% (median of 7%, average of 11%).

Similarly, the rate of change in SPX Advance/Decline reached the highest level in its 5 year history this week. In 13 of the 14 prior times it has reached an extreme, it has given up subsequent gains in the weeks ahead. In the one exception, all gains were given up within 2 months. In 4 instances, SPX dropped hard. This confirms the findings from the other studies above.

Two other studies are of note.  The first has to do with sentiment.

We stated last week that the sentiment picture was mixed: some of indicators suggested a washout but others did not. Perhaps the most bullish was the equity-put call ratio, which had reached a relative extreme and then rolled over. In the past, this has been a buy signal for SPX (chart from McMillan).

But other sentiment measures do not concur. Equity fund flows (including ETFs) show a scant $3b outflow for October. For a 10% correction, this is minor. Flows were only negative two weeks in a row; for comparison, in mid-2012, flows were negative 6 weeks in a row and then continued to be negative off and on over the next several weeks. In the week just ended, $8.8b flowed into equities, the highest since August 20. This is hardly the profile of a bearish investment community.

Similarly, SentimentTrader, via a post by Brian Johnson (here), shows that retail investors were already back at a bullish extreme within a week of a 6 month low. The last time this happened was in June 2010 when SPX was on its way lower in the middle of a 4 month correction.  Overall when this has happened in the past, near term results for SPX were modest but returns past 1-month were negative.

The speed of the recent rebound suggests that investors were unfazed by the sell off and therefore unafraid of buying equities. This is not surprising given the number of sharp, v-shaped bounces the market has experienced since 2013. But if past is prologue, the study above suggests future returns may be limited until sentiment is more fully reset.

(As an aside, this week, NAAIM showed the largest two week gain in bullish investors in the survey's 8 year history and Investors Intelligence showed the largest one week gain in bulls in the past 12 years).

The second study has to do with the sharp bounce in small caps this week. According to Paststat, large gains like those seen this week are given back over the next week, with a median drawdown of more than 3% in the next two weeks (post).

In summary, a number of studies show a poor risk/reward profile in the near term for US equities. That does not mean that equities will fall. This market has shown a strong propensity to defy precedents. This might well be another such situation. The question for us is always risk vs reward. We believe the odds of a 5% rally appear much smaller than those for a 5% retracement. Therefore, initiating new longs here is uninteresting.

We are very cognizant of two strong positives. The first is trend. As we stated last week, a reliable buy signal triggered on October 22 with SPX at 1930, when both its MACD and 13-ema crossed positive. This set up has marked every low since December 2011.  This set up is still active and suggests higher highs are ahead. We don't like to fight the primary trend.

The other is seasonality. November through January are regarded as the strongest three month period of the year. We'll have more to say about this in a future post, but for now, investors should be aware that equities should remain buoyant through next Tuesday's election and often for a few days after. The "perfect" time for some equity weakness would be into mid-November (yellow shading; chart from Stock Almanac).

Non-farm payrolls (NFP) are reported on Friday. This has been a bullish day for 2 years, with SPX closing higher 19 of the last 23 times.

Our weekly summary table follows.

If you find this post to be valuable, consider visiting a few of our sponsors who have offers that might be relevant to you.