Technical update: overbought, overbullish, declining RSI

We finally have the classic syndrome that indicates an intermediate-term top. Upward momentum has stalled, as sentiment has remained elevated for several weeks. The combination of sideways prices on high bullish readings becomes very bearish when it has been sustained for a month or longer.

Here is the RSI and price picture (note declining trend in the momemtum indicator RSI since late August, and its resemblance to the topping pattern last spring):


Charts from Yahoo

A quick glance at sentiment shows sustained optimism:

Looking at the headlines, it is nice to see good news that results in a bump with no follow-through. We saw that in mid-September with QE Infinity, and last Friday with the jobs report. Rallies end on good news and declines end on bad news.

It would not be unusual to see another test of the highs, and for prices to linger at these elevated levels for another month or so, but the odds of a sharp decline are now elevated, and any further gains should be quickly erased.

Topping pattern developing, stay tuned

We finally have a weakening trend on the daily RSI (see RSI on bottom of chart). This is a prerequisite for anyone considering taking a short position, especially with leverage.

It would be typical for some type of quick plunge to develop soon, perhap on the order of 5-8% in SPX, maybe 100 points (1000 Dow points). It would also be typical for stocks to recover from such a plunge and test the highs again, as in spring 2011 or Summer-Fall 2007. In May 2010, this process was compressed in the “Flash Crash,” which was followed quickly by a decline of around 20%.

However, the weekly trend is still up, though finally in overbought territory. Perhaps this begins to turn over the next few weeks as the market chops sideways.

Further strengthening the bear case is sentiment, which has now been elevated for at least two months (I recommend at $30/month for all you can eat indicators). Sentiment is a powerful indicator, but actual price tops lag tops in sentiment by several weeks to months, as prices tend to levitate and chop sideways for a while on weakening RSI prior to major declines. Bottoms have a much closer time relationship to sentiment (extreme bearishness among traders is usually quickly followed by rallies).

The best free sentiment resource that I am aware of is the NAAIM Survey of Manager Sentiment. As you can see in the chart below, sustained bullish sentiment is required in order to register a valid sell signal, as the crowd is often right for a while. The relatively short period of bullishness here is another sign that any decline that develops soon could be short-lived.

I should also mention that John Hussman has noted that the market has exhibited his syndrome of overbought, overvalued, overbullish on rising yields, for several weeks now. This set of conditions coincides with many of the very worst times to be long stocks, and has almost no false positives. Advances made during such periods are soon given up, often in severe declines.

The following set of conditions is one way to capture the basic “overvalued, overbought, overbullish, rising-yields” syndrome:

1) S&P 500 more than 8% above its 52 week (exponential) average
2) S&P 500 more than 50% above its 4-year low
3) Shiller P/E greater than 18
4) 10-year Treasury yield higher than 6 months earlier
5) Advisory bullishness > 47%, with bearishness < 27% (Investor’s Intelligence)

[These are observationally equivalent to criteria I noted in the July 16, 2007 comment, A Who's Who of Awful Times to Invest. The Shiller P/E is used in place of the price/peak earnings ratio (as the latter can be corrupted when prior peak earnings reflect unusually elevated profit margins). Also, it's sufficient for the market to have advanced substantially from its 4-year low, regardless of whether that advance represents a 4-year high. I've added elevated bullish sentiment with a 20 point spread to capture the "overbullish" part of the syndrome, which doesn't change the set of warnings, but narrows the number of weeks at each peak to the most extreme observations].

The historical instances corresponding to these conditions are as follows:

December 1972 – January 1973 (followed by a 48% collapse over the next 21 months)

August – September 1987 (followed by a 34% plunge over the following 3 months)

July 1998 (followed abruptly by an 18% loss over the following 3 months)

July 1999 (followed by a 12% market loss over the next 3 months)

January 2000 (followed by a spike 10% loss over the next 6 weeks)

March 2000 (followed by a spike loss of 12% over 3 weeks, and a 49% loss into 2002)

July 2007 (followed by a 57% market plunge over the following 21 months)

January 2010 (followed by a 7% “air pocket” loss over the next 4 weeks)

April 2010 (followed by a 17% market loss over the following 3 months)

December 2010


This chart from Hussman has data through March 3, but the latest blue band is now about a month wider:


This is clearly time to exit stocks or hedge all market risk. The upside is limited relative to the downside.

Relief rally coming?

We’ve got a clear divergence on RSI now, as each impulse lower over this week has been weaker than the last. This is a sign to tighten up stops or close shorts. You could make a decent case for a quick long trade here with a stop just under the lows, but on a wider time frame market risk is still very high.

Here’s a chart of SPX futures:

TD Ameritrade

Time for a bounce? (euro, stocks, copper, oil)

Here’s ES, 1-min scale:


RSI could use one more low, ideally at a higher level than the last to increase the odds of a rally.  The Nasdaq has diverged from the S&P 500 already though, and copper has a clear upslope in RSI:


EDIT: Here’s that higher low in RSI on ES that I was talking about and breakout from the downsloping trend:


Hard to deny there’s a bearish pattern here:

TD Ameritrade

The precious metals have ambivalent correlations with stocks these days, so I’m not sure what the above means in the scheme of things, other than the commodities echo bubble slowly deflating. Sometimes the metals are high beta, sometimes negative, and sometimes they seem to have no correlation at all.

Toppy action in dollar index, another rally in stocks coming?

The dollar index is showing waning strength on advances, and its rally is very long in the tooth. There could be worse times to take a short position on DX. At least there is a clear and close stop price at the highs.

Here’s the 2-hour bar, then the 5-min:

TD Ameritrade

Here’s what ES looks like to me this evening (5-min chart). There’s an upward divergence in RSI, indicating waning oomph on the sell-offs:


I took a long position in CHF and ES (SPX futures) after the close of regular trading today, and I’ve shorted some Treasuries and yen, but I’m still short oil, palladium (added to my position today), copper, silver, gold also added more today) and copper. I’m agnostic about the intermediate-term direction of the risk/inflation trade and just trading what I see: commodities look broken, but stocks still have rebound potential, as do the euro, franc and pound.

This chop and weak rally action in stocks precludes neither a big new rally nor a fall into the abyss. Big rallies like August-October 2007 or Feb-April 2010 have started slowly with chop like this and kept the bears confident. The timeframe for such a rally is limited, however, and if we don’t take off in another week or so momentum will peter out and we can start to roll over. This is what happened in late spring 1930 after the first hard leg down from the top of that post-crash rally.

Weakness developing in commodities

Checking the 15-min bar chart, copper and oil are not looking too spirited.

Here’s oil:


And copper, same scale:


And of course platinum and palladium are looking busted. Daily charts here.





The technical damage in these metals is probably not a good sign for gold and silver either.

Silver’s daily chart leaves a thing or two to be desired:

Dollar set for a fall?

Checking the early futures action, I see some strength has been building in several currencies vs the dollar over the past few days, despite their longer-term weakness. The dollar enjoyed a small rally against the european set late last week, putting the CHF, EUR and GPB in attractive positions for long trades. With sentiment entering its 5th month of extreme negativity, it may pay to be alert for signs of a rally (that is to say a dollar decline, and further yen decline too against the dollar).

Here are a few charts (1-hour bar), noting the RSI uptrends that have developed:





To wrap it up, here is the dollar index futures contract:

Week in review and intermediate-term thoughts

Today I’m going to lay out what I’m watching for clues about the intermediate-term prospects. The action of the last 4 weeks has been more suggestive of a reversal than any we’ve had since the March 2009 lows. However, even if a top is in hand and we are finally at our spring 1930 moment, I’m not willing to throw caution to the wind and discount the possibility of another few weeks of rally.

Let’s start with the 5-day average equity:put call ratio, which has nailed so many intermediate-term tops, not least of which this last, which it suggested would be followed by a serious decline:

The put:call ratio could use a bit more of a reset, which could be achieved by just a few more days of calmer or rising markets. Nothing major required here, just a pause.

This break has shaken up a few traders, but judging from interviews this week, the majority remain fairly sanguine about a continued bull phase and consider this a healthy correction, much as they did the first declines off SPX 1550 in late 2007 and early 2008. Also, the past year has established a very clear pattern of modest declines followed by new highs on extreme bullishness. Traders and their machines may be programmed to buy this dip, but with the recent technical damage (we busted the last low for the first time in the whole rally and experienced a mini-crash) I’d expect any rally to be relatively shallow.

There is a pattern of reduced oomph on each subsequent rally phase, which you can see in the diminishing slopes of each rally. You can also see the weakening of the larger trend in the angles between subsequent lows.  (Click to enlarge the image.)

TD Ameritrade

I’ve drawn those ovals on RSI to show a technique I have for picking bottoms. It gets most intermediate term bottoms, and perhaps more importantly, it has a low false-positive rate. A rally becomes likely when you get a double bottom in RSI. This works on any scale you chose, from 1-minute to daily or higher. The likelihood of a rally increases if the second bottom on RSI is higher than the first. This is common because the middle wave of a decline is usually more intense than the final wave.

Now, this current juncture has such a double bottom signal, though the second RSI bottom is not higher than the first. It is also trickier because the first was formed by the latest black Thursday, May 6. I’m not sure how such an event should factor in, but it throws off our analysis somewhat. Perhaps the May 6 event should be discounted (it was really just 1 hour of trading that produced the reading) so that we can’t actually count this RSI bottom as a 2nd.

In terms of time, we’re just over 4 weeks into the decline, which is approaching the average for an intermediate-term decline over the last 3 years (the last one was very short at 3 weeks, and others have lasted up to 8 weeks).

Also of consideration is the extreme complacency that we are correcting. Look again at CPCE in the first chart above. From what I can tell, it set a record for complacency going back to at least the year 1999. This suggests we may have more decline ahead before an extended relief rally. Sentiment has turned negative, but not overwhealmingly so, and it has only been negative for a couple of weeks, so this is not a contraint to a further decline.

One more consideration is the 1930 parallel. Once stocks broke that April after their rally from the crash of ’29, they failed to rally hard for years. The decline was steady all the way down to the bottom in July ’32. In this analalogue, we would have another week or so of choppy and weak rally, followed by the bottom falling out, an outcome that would elegantly resolve our situation. The dip-buyers pile in, but the oomph is gone, momentum weakens and RSI turns down, then BAM, we’re back to SPX 750 this summer.

Prophet charts

I am approaching this situation by being neutral on stocks at the moment. I am holding a core position in December 2011 and 2012 SPY puts and some calls I’m short on IYR and GDX, though I sold a portion of the puts on Tuesday morning and and the rest are hedged with a short in VIX futures (I do this because spreads on options make them costly to trade in and out of). Essentially, I’m flat on equities.

I closed a ton of shorts from last Thursday to Tuesday morning, and went long SPX, ASX and Nikkei futures (and long CHF, EUR, GBP and short JPY and VIX) early this week when I saw divergences in the VIX, currencies and commodities (ie, stock indexes made a new low that was not confirmed with new extremes elsewhere, a buy signal) as well as a glaring RSI divergence on the hourly scale. Those “long risk” positions I closed for profits on Thursday and Friday, since we’ve already corrected the extreme short-term oversold condition and are in neutral territory. Equity-wise, I ended the week where I was on Tuesday morning, since the drop in volitility hurt my puts as much as my various longs made me money. Vol is a bitch that way — sometimes you time prices right, but it’s not enough.

Speaking of the VIX, I think it could settle down for a few weeks, though to a higher level than in April, before the next decline pushes it up again. I think it will remain elevated (as from Oct 2007 onwards) for many more months or a couple of years:

Prophet charts

In the commodity space I’m even more convinced that a major top is at hand, since some trendline breaks have been decisive (platinum, palladium, oil) and the declines have been so violent all around. Commodities tend not to rally as hard as stocks once the trend changes to down, so I entered shorts on oil, silver, gold, palladium and copper near their highs late in the week. The precious metals are looking particularly suspect to me here, and I still think my July 2008 double top analogue is in play.

The euro, Swiss franc and British pound are still looking very weak. Sentiment has been in the dumps for four months now, which is a set-up for a spectacular rally, but judging from their heaviness this week as stocks and commodities and CAD and AUD rallied, I think they may slide to one more low before that rally.