Short-term Outlook (1-5 Days):  Neutral  Since June 24, 1067 SPX

Go to:  Top  |  Short-term Outlook  |  Int-term Outlook  |  Equity Updates  |  Indicator Summary  |  Commodity Updates

Go to:  Top  |  Short-term Outlook  |  Int-term Outlook  |  Equity Updates  |  Indicator Summary  |  Commodity Updates

Equity Indicators - Updates and Extremes

Head & Shoulders Pattern

The technical support that "everyone" was watching in the S&P 500 was 1050, which coincided with the June closing low and a break of the Head & Shoulders pattern.

That pattern has gotten as much media attention as the "death cross" that we looked at last week.  It's probably just as (in)effective as a predictor...but let's actually test it instead of relying on textbook "shoulds" and vague opinions.

Some interpretations of the pattern differ, but if we're going to test this thing then we need hard-and-fast rules.  So what we're looking for is any time the market rallies, then falls back (creating the left shoulder).  Then it rallies even higher (creating the head) before falling back again to a level near the first pullback.  Then it must rally once more to a point that's relatively close in price to the left shoulder, and finally it must break the trendline that joins the trough between the two shoulders.

Once that occurs, we would sell short on a violation of that trendline.  The profit target is the difference between the head and the trendline, while the stop loss would trigger if prices rally back above the head.

If we test the Head & Shoulders pattern on the 500 stocks in the S&P 500 since 1995 by selling short when the pattern triggered, we get a total of 383 trades.

Out of that sample, only 27% of them were winners.  That means that almost 75% of them did not decline enough to meet the profit target.  Overall, the average return from these trades was -1.2%.  Not good.

For those trades that did work, it took a median of 26 trading days to hit the profit target.  For those that failed, it took a median of 58 days to get stopped out.

For the stocks in the Nasdaq 100, the win rate was 50% with an average return of -3.3%.  The higher winning percentage is likely due to higher volatility, and thus higher likelihood of hitting the price target, even though the 50/50 winning percentage wasn't enough to make the strategy even remotely profitable.  It took a median of 43 days to get stopped out, and a median of 24 days to hit the profit target.

There is some art in defining the Head & Shoulders, which makes it difficult to test, and looking through some of the trades, I personally wouldn't have considered them valid.  But the test gives us a general feeling for how successful the pattern has been across a wide swatch of stocks, without relying on a handful of cherry-picked examples.

Is the violation of that 1040-1050 support area bearish?  Probably, and I'm more cautious as a result.  Does it mean we're going to hit the profit target from the Head & Shoulders pattern.  Most probably not.

Price Oscillator

On an intraday basis, we track a Price Oscillator for the S&P 500 and Nasdaq 100.  The indicator looks at how each 30-minute bar closes compared to its open, high and low.

We can do the same thing on any length bar; it doesn't have to be 30 minutes.  When we do it on the daily S&P, something very notable emerges.  Due to the remarkable stretch of weak closes versus opens lately, the 10-day average of this Price Oscillator is more stretched than any point since July 19, 2002.

On the intraday charts, when the Price Oscillator drops under 30%, we know that we're massively oversold on a very short-term basis.  Well, we just hit that level on a daily basis and it's exceptionally rare.

As a matter of fact, when we apply this to the S&P 500 futures since their 1982 inception, only three other dates can match the current level of intrabar selling pressure:

May 24, 1984:  The S&P bottomed 2 days later

October 19, 1987:  Black Monday bottom

July 19, 2002:  The S&P bottomed 2 days later

Obviously, it's pretty tough to rely on just three occurrences as a basis for what to expect going forward.  But the fact that all three coincided within days of major lows is very intriguing.

Rydex Ratio and Rydex Total Bull / Bear Ratio

The kind of incessant selling we've seen tends to cause consternation among Rydex traders, and we're certainly seeing some signs of that.

As of yesterday, the Rydex Ratio, which divides the assets in the S&P 500 long fund by the assets in the S&P 500 inverse fund (that profits if the market declines), nearly dropped to a new all-time record low.  Currently, there are almost 10 times more assets invested in the inverse fund than the long fund.

The only other time in the past 15 years when the ratio reached this level of pessimism was early March 2009 at the genesis of the bull market.

If we expand the ratio and bit and look at the Total Bull / Bear Ratio, the picture isn't quite as extreme but is still notable.  Traders are betting as heavily on the downside as any other point since the March bottom.

Going back to 2000, when the ratio of total bull funds to total bear funds drops under 0.88, the S&P 500 put in the following performances going forward:

Even considering the fact that the ratio can get "stuck" in extreme territory during bear markets, the performance in the S&P going forward was impressive, especially when looking out a couple of weeks and again a few months later.

Go to:  Top  |  Short-term Outlook  |  Int-term Outlook  |  Equity Updates  |  Indicator Summary  |  Commodity Updates