Friday, August 9, 2013

Risk-Reward Is Again Biased to the Downside

In general, stock market odds favor the long side. Since 1980, the probability of making a positive return each year has been 76%.

When the year starts strong, like 2013, the probability of making a positive return in the second half of the year is even better. Since 1950, when SPX has been up more than 10% in the first half (1H) of the year, it has been up again in the second half (2H) 79% of the time.

SPX gained 13% in the 1H; so far in the 2H, it's up another 5%. Should investors expect more gains through year end?

The answer is yes, but expectations should be modest.

In the years with strong 1H gains referenced above, gains for the 2H have averaged 7%, 2 percentage points more than the gains so far. Looked at another way, total 2H gains have averaged about 40% of 1H gains; in 2013, that implies net gains from here through year end of less than 1%.

A strong 3Q helps. When 3Q finished up more than 5%, 2H averaged gains of 9.8%, implying another 4.8% upside for 2013. This, coincidentally, gets SPX to 1775 where JPM has its year end target, the most bullish forecast on Wall Street.

Against this upside (reward), what is the downside (risk)?

While the trend in the US indices is strong, historically, the pattern from the June low has been bearish. In the past 20+ years, the SPX has never successfully V-bounced and continued higher without a retest of the recent lows.

The first chart is SPX weekly since 2003, the second chart is from 1991-2001. Notice that after every long uptrend, SPX corrected (red shading; the width of the shading is 3 months). Every V-bounce, where SPX returned to the recent highs within 3 months eventually failed and either retested or exceeded the lows (shown with red arrows). When the correction was longer and slower (blue arrows), the ensuing uptrend was more successful.

In the current situation, SPX has already chosen the V-bounce pattern; it would be the first time since at least 1991 if it does not make a retest of the June low in the weeks ahead. This implies downside (risk) of 8%. If the retrace is a Fibonacci 62% (not common, in the examples below), then downside risk is 5%, or equal to JPM's upside target.

Net, risk-return is probably at best 1:1 and probably closer to 2:1 to the downside. Importantly, should a retest of the lows occur in the weeks ahead, recall that the odds favor a strong 2H.

A more traditional way of looking at the current set up is shown below. Notice two things. First, the RSI negative divergence (RSI lower while SPX higher). SPX is losing momentum, which typically implies downside risk ahead.

Second, after becoming very 'overbought' (RSI over 70), a washout low occurs when RSI becomes very 'oversold' (RSI below 30; green arrows). When SPX bounces before the washout is complete (red arrows), the ensuing rebound has failed. This is what occurred at the June low, another pattern that suggests downside risk before greater upside returns.

As you consider these patterns, keep in mind the present positioning of investors (read today's post here) and the seasonal patterns (here). Both of these suggest the V-bounce will fail in the weeks ahead.