This Reliable Long-Term Indicator Could Go Bullish at the End of the Month

Sixty-Five Year Record of Success

One of my favorite long-term indicators involves the 12-month ROC of the S&P Composite falling below the -5% level and subsequently rallying back above zero. This approach is shown in Chart 1, which illustrates the period from 1995 through April 30 of this year. Chart 2 covers the period between 1953 and 1995. The -5% line is shown in blue while zero is in black. The green arrows in Chart 1 flag the  signals that have been triggered since the mid-1990s. Based on a one-year holding period, only one of the 19 signals that have been generated since 1953 has been unprofitable. That exception developed in 1957, but the ROC only held above zero for one month.

Chart 1 also shows that the model is not yet bullish, as the ROC remains below zero. The reason why this technique interests me is because the closing price for May of last year was 2752 and the latest May 7 price is around 2880. That means that the May 2020 closing low could fall around 120 points or so from current levels and still turn the model bullish. Since the market has risen in excess of 30% off its March 23 intraday low, the question naturally arises as to whether that's possible. The answer is unknowable, of course, but it is possible to point out some benchmarks that place things in greater perspective.

Extreme Overboughts and Oversolds

First, take a look at Chart 4, which features the 10-day MA of the S&P Composite's 12-day ROC. Two events recently happened. The first was an exceptionally extreme oversold reading in March, while the second was a close-to-record overbought condition in April. Normally, these extreme oversolds develop at the tail end of a bear market, when prices have been dropping for a year or more. If they occur close to an all-time peak, as in the 1987 example, this usually indicates a change of sentiment to the bearish side. Such conditions, known as mega-oversolds, offer the first early bird (in terms of time) signal that a new bear market is underway. In the 1987 case, that did not pan out because the mega-oversold was immediately followed by an above-average overbought reading, which had the effect of cancelling it. Remember, markets are nothing more or less than people in action, and people can and do change their minds. It's not therefore unreasonable for markets to do the same thing, as in the 1987 situation (from an extreme pessimistic oversold to a more optimistic above-average overbought reading).

The current setup is similar in that both developed close to an all-time high, with a close-to-record oversold reading being followed by a record overbought one. The April overbought reading, however, almost tied the previous record and is therefore far more impressive. The three previous bullish extremes that touched or exceeded the green horizontal line, in 1982, 2002 and 2009, were all followed by big bull markets. The other point to bear in mind is that none of these extreme overbought readings were followed by new bear market lows. The experiences in 1932 and 1938, when similar overbought setups developed, were not followed by new lows either.

This is not to say that there will not be a test of the March 23 low, merely that the March 23 bottom is likely to hold for a long time.

A Couple of Discrepancies

One of the reasons we could see a test, regardless of the extreme overbought reading in the MA of the 12-day ROC, lies in the fact that all of the strong rallies from 1929-2009 cited above were associated with one or more overbought readings in the 9-day RSI. That may happen down the road, of course. So far, though, none of the major averages we follow, with the exception of a couple of sectors, have reached an overbought condition on either a 9- or 14-day time span. That's a problem.

Charts 5 and 6  feature both RSI timeframes; neither the 9-day nor the default 14-day series have reached the 70 overbought zone since the March 23 low. Note how the oscillator had no difficulty in achieving overbought status in late 2019 during a bull market environment. The benchmarks, which may offer an indication that a test is underway, could well come from a violation of the two trendlines in Charts 5 and 6.

Good luck and good charting

Martin J. Pring

The views expressed in this article are those of the author and do not necessarily reflect the position or opinion of Pring Turner Capital Groupof Walnut Creek or its affiliates.