There are many market participants out there whose nerves appear to be pricking up lately as we approach the two worst months of the year, more so given the fact that the S&P 500 has risen more than 50% from the March lows and is overdue for a correction. Last year should have made the point that in terms of the financial markets, anything and everything is possible as we appeared to be on the brink of a systemic collapse. However, while anything is possible, how likely is another 1987-like sell off in the weeks ahead, or even a sell off similar to the 1998 Asian Currency Crisis?
Whether one looks at seasonal patterns over the last 20, 50, or even 100 years, September has consistently proven to be THE worst month of the year by a sizable margin as the enumerable folks at Bespoke Investment Group show below.
Source: Bespoke Investment Group, B.I.G Tips (08/25/09)
While a general correction is health as it lets out some froth and steam from a market’s advance and would allow investors to put capital to work at lower prices, no one wants to live through a dramatic sell off as was seen in 1987, 1998, or even last year’s carnage that took place over September and October. In order to determine the likelihood of a major (15% +) rather than a general correction (5-10%), a few indicators were reviewed to see if any red flags appear that could be heralding a rough September ahead.
Stock-Bond Ratio
Sometimes it is helpful to not look at any one market in isolation but rather in comparative terms to other asset classes to see how far skewed the market is in any given direction. To do this one can use ratio analysis which divides one variable by another, and then comparing the result to the historical norm or to moving averages to look for trend changes. An example of this was done back in the middle of February to determine if perhaps gold had moved too far relative to the stock market, and to gauge whether a correction or consolidation period lay ahead for gold. In the February article, “Gold, Is the Future Still Bright or Fading?,” the gold to S&P 500 ratio was analyzed relative to the 200 day moving average of the ratio which showed that gold was in fact extremely overbought relative to the stock market. Commentary and figures from that article are provided below:
The recent surge in gold as it approaches its 2008 highs has pushed the ratio of gold to the S&P 500 nearly 60% above its 200d MA, the most overbought condition since the peak in gold in 1980, more than 28 years ago. Clearly at this juncture gold’s outperformance relative to the S&P 500 is reaching extreme territory. To put a more quantifiable context to gold’s recent outperformance I measured the standard deviation of the ratio to its historical average relative to the 200d MA. As seen below, the outperformance of gold to the S&P 500 relative to its 200d MA is nearly four standard deviations above the mean, an event that should occur only 1/100 of a percent of the time, or roughly once every 27 years, which is actually roughly the last time this event was witnessed as seen in the chart below.
Source: Bloomberg
Source: Bloomberg
As it turned out gold put in a top within two days and corrected roughly 15% into late March. However, while gold corrected roughly 15% the gold to S&P 500 ratio was still well above its 200 day moving average and has since spent the last few months consolidating while the S&P 500 has risen dramatically, driving down the gold to S&P 500 ratio to levels that have typically marked major bottoms, as shown in the updated figure below.
Source: Bloomberg
Using the same type analyses one can look at the stock to bond ratio (SB) to determine whether one asset class has moved too far relative to the other. Looking at the SB ratio back in 1987 showed a very dramatic rise in stocks relative to the bond market that reached an extreme condition and warned that stocks had likely ventured too far too fast. The relative strength index (RSI) of the ratio on a monthly basis reached just over 75 at the market's peak, a severely overbought condition and warned that stocks were ripe for a selloff relative to bonds.
Source: Bloomberg
Looking at the 1998 selloff in the market showed that the SB ratio was consistently overbought for more than a year and didn’t provide any immediate warning. However the weekly RSI was diverging in the final months leading up to the correction with the SB ratio as it was putting in a lower high and warned of momentum loss in stocks relative to bonds.
Source: Bloomberg
So what does the SB ratio tell us about the current situation? The weekly RSI of SB ratio earlier in the year was pointing to a similar setup as the 2002-2003 market bottom as the RSI was failing to put in new lows to confirm the lower lows made in the SB ratio from the October lows to the March 2009 bottom, which correctly heralded a market bottom. The weekly RSI is currently near 60 and reaching overbought territory near 70, but there is no divergence yet to suggest a major market top is approaching as momentum is confirming price.
Source: Bloomberg
When looking at the monthly chart of the SB ratio, the monthly RSI is still relatively oversold at a value of 38.92, and the last two major market tops of 2000 and 2007 were witnessed the RSI at or above readings of 70, so on a monthly basis the stock market relative to the bond market is considered neutral and isn’t sending any cautionary red flags.
Source: Bloomberg
NYSE New Highs and New Lows
Another helpful indicator to determine the health of a stock market advance is the percentage of New York Stock Exchange (NYSE) issues that are trading at 52-week highs and the percentage of issues trading at 52-week lows. A healthy market advance is characterized by net new highs making higher highs with a market advance, and a divergence between the net new highs with the NYSE warns of a deteriorating rising market, while divergence of net new lows with a falling market points to an improving market as less and less issues are making new lows with the NYSE. A red flag of a possible market correction is when the percent of NYSE issues making new highs diverges with the NYSE, and a secondary red flag is given when the percent of NYSE issues making new lows and making new highs are BOTH above 2.5% of traded issues. This signals that while some stocks are making new highs others are already falling off the beaten path and making new lows, showing significant divergence internally within the stock market.
You can see this below when looking at the health of the NYSE index prior to the 1998 Asian Currency Crisis selloff. The percentage of NYSE issues making new highs peaked in July of 1997 and made a series of lower highs while the market headed higher (declining green trend line) warning of deteriorating internals. In late May of 1998 BOTH the percentage of NYSE issues making new highs and those making new lows were greater than 2.5% (red horizontal line), flashing a sell signal (red vertical line) more than a month before the market top.
Source: Bloomberg
An example of the opposite situation is the character of market internals from October 2008 to the March lows of this year. The peak in NYSE issues making new lows came during the panic selloff seen in October, and with each new low in the S&P 500 fewer and fewer NYSE issues were making new lows, signaling that the underlining market internals were improving, heralding a coming market bottom.
Source: Bloomberg
Looking at the present circumstance, on a weekly basis the % of NYSE issues making new highs have confirmed each intermediate high since the March bottom and aren’t warning of a major market top. Additionally, the % of NYSE issues making new lows is not even at 1/10 of a %, and is far below the 2.5% warning threshold.
Source: Bloomberg
While the weekly chart above isn’t giving any warning signs as the % of NYSE issues are confirming each new high in the S&P 500, the same cannot be said of the daily chart. As seen in the first red box below, the % of NYSE issues making 52-week highs peaked in early June and began to fall off while the market remained flat, with a multi-week correction soon following. A similar set up can be seen over the last two months (second red box) as the % of NYSE issues making new 52-week highs is not confirming the recent high in the S&P 500, which would point to a possible correction ahead. Please note though, that the % of issues making new 52-week lows is virtually muted, as less than 1/10 of a percent of NYSE issues are making new lows, so no red flags being signaled as neither the % of NYSE issues making new lows or highs is currently north of 2.5%.
Source: Bloomberg
As stated above, all things are possible though not all things are probable. When looking at the SB ratio and the internal health of the stock market by the percent of NYSE issues making new highs and lows, it doesn’t appear that any major market top is evident or that a major selloff like the 1987 crash or 1998 Asian Currency Crisis is brewing. However, September has proven to be the worst month of the year and the percent of NYSE issues making new highs have not confirmed the recent market highs in the S&P 500, so a garden variety correction may be what lies in store in the coming weeks.
