Author Archive for McClellan Financial

Capitulation Not Over in Crude Oil

Tom McClellan – McClellan Market Report

Crude oil prices had a seemingly exhaustive washout selloff following the Nov. 27 OPEC meeting.  Oil bulls had been hoping for a production cutback at that meeting, but Saudi Arabia successfully led an effort to oppose such cuts.

But the message from the Commitment of Traders (COT) Report data is that the washout is not yet complete.  An exhaustive move like what we have seen should produce capitulation among the small speculators, but instead the readings from recent weeks showed them them rushing in to buy.

Crude Oil COT Report data

Traders’ positions are reported each Friday in the COT Report, and they are broken down into 3 categories:

Commercial traders are the big money, and presumed to be the smart money.  They are “engaged in business activities hedged by the use of the futures or option markets.”

Non-Commercial traders are ones who are not engaged in such practices, but who have large enough positions to merit individual reporting of those positions.  Think hedge funds.

Non-Reportable traders are those whose position sizes are small enough that the CFTC deems them not worth reporting individually.  They are the small speculators, and reliably considered to be the “hot money”.

This week’s chart looks at the Non-Reportable traders’ net position in crude oil futures.  These traders tend to get more net long as prices move higher, and they get scared out or go short as prices go lower.  Generally speaking they have a bias toward the net long side, and so any time they actually go net short, it is usually a sign of a bottom for crude oil prices.

Given the amount of the drop in crude oil prices, it would be reasonable to expect these traders to get shaken out of their long positions.  But that is not what they have been doing.  The “buy the dips” mentality was still active.  With the COT Report released on Friday, Dec. 5, they are finally starting to make more of a move to unload their long positions, but they are still not yet back to neutral or even net short, which is what it should take to mark the bottom of this decline.

The COT Report is issued every Friday, and we feature a discussion of the relevant insights from that data every Friday in our Daily Edition.

This oil price decline actually ties into the recent Hindenburg Omen (HO) signals which were triggered this week.  I was just on CNBC on Dec. 4 talking about the Hindenburg Omen, and you can see that interview here.  One point to understand about that video is that headline writers like to spice things up, and in ways not necessarily consistent with the actual story, or with the views of those who are interviewed.

The reason why the oil price decline is related to the Hindenburg Omen is that HO requires seeing both New Highs (NH) and New Lows (NL) exceeding a certain number of issues on the same day.  There are also some other requirements.

A cursory review of the list of stocks making new 52-week lows shows a lot of stocks with the words “drill”, “energy”, or “resources” in their company names.  Were it not for the concentrated damage to energy stocks resulting from the oil price slide, we would likely not be seeing an HO signal now.  Not all HOs end up seeing a big market collapse, but they do tend to show up ahead of every big market decline so they are worthy of some attention.  To get a big slide now, the stock market is somehow going to have to fight off the bullish forces of positive seasonality, plentiful liquidity according to the breadth numbers, and more QE coming from other countries’ central banks.  Plus we are now in the year following the mid-term elections, which is nearly always an up year.

That is a tough package of forces for the stock market bears to fight against.

Related Charts

Jun 06, 2013small chart
Hindenburg and Titanic, OH MY!
Oct 03, 2014small chart
Commercials Betting On Big Dollar Downturn
May 15, 2014small chart
Crude Oil Says DJIA Should Trend Higher

Chart In Focus Archive

More Good News for Employment

Tom McClellan – McClellan Market Report

The data on the U.S. unemployment rate have been getting progressively better over recent months, either because of or in spite of the government’s efforts, depending on one’s viewpoint.  And if this week’s chart is to be believed, then the data should continue to get better over the next several months.

What the chart shows is that the data from the University of Michigan’s Consumer Sentiment Survey acts as a leading indicator for the unemployment rate.  I am using a 10-month offset in this case, shifting the sentiment data forward by 10 months to reveal how the unemployment data tend to follow in the same footsteps.  One could make the case that 10 months is perhaps not the perfect time offset; maybe it is 6 months, or maybe a year.  But the point remains that the University of Michigan survey data do seem to lead by some amount of time.

UMich Sentiment versus unemployment rate

Back in 2009, the Fed had the fire-hose wide open, and yet the jobs market did not seem to respond.  The unemployment rate finally peaked at 10.0% in October 2009, several months after the stock market’s bottom, and also several months after the Fed started “helping”.  What had not happened yet in early 2009 was the passage of enough time following the recovery in consumer sentiment.  The University of Michigan survey had bottomed out at a reading of 56.3 back in February 2009.  So some number of months needed to have gone by before the unemployment rate could respond.

Now we have a different situation.  The latest reading for this survey data was 89.4, the highest since the summer of 2007.  It continues to trend higher, which means that the unemployment rate should continue to trend lower (remember it is inverted in the chart).

At some point, consumer sentiment will reach its peak for this cycle.  When that happens is an unknown.  I don’t have a good leading indicator for that question.  When consumer sentiment reaches its peak, then some number of months later we can expect a reversal for the unemployment rate.  The point for now, though, is that consumer sentiment says the unemployment rate should continue dropping for at least another 10 months or so.

Related Charts

Aug 28, 2014

small chart
A-D Line Foretells Job Openings

Jan 23, 2014

small chart
Important Points About The Minimum Wage

Jan 04, 2013

small chart
Employment Levels Stubbornly Unresponsive

Chart In Focus Archive

Deflation Is Getting Too Popular

Tom McClellan – McClellan Market Report

I cannot believe the volume of the news stories I am seeing in the financial media, with people worrying about impending deflation.  And as any card-carrying contrarian knows, when a topic gets too popular, you are near a turning point.

To check that observation, I went to Google Trends and did a quick search on the term “deflation”.  What Google does is to then come back with a chart showing interest in that term over time among the news media.  And sure enough, October 2014 is showing the highest reading since late 2010, and the month is not even over yet.  As I see it, that confirms that the media are getting a little bit too interested in this topic of deflation.

So what?  Well, when “everyone” is expecting deflation, they usually place their financial bets based on that expectation.  And that usually takes the form of T-Bond prices going too far.  So to verify that notion, I took that Google Trends chart and overlaid a chart of T-Bond prices for the same period.

Google Trends plot of deflation versus T-Bond prices

There are ways to create a more elegant chart comparison, but Google does not make it very easy to get the raw monthly readings unless you are willing to hover your mouse cursor over every point on the chart, note the value that the Flash graphic pops up, put that into a spreadsheet, and then adjust the chart settings accordingly.  So I went more old school, and did it like we used to in the days of acetate transparencies on an overhead projector.  I made the background of the T-Bond price chart transparent, and laid it over the Google Trends chart, stretching and adjusting the bond chart as needed to get the timelines to line up.

The result is a crude but workable overlay which allows us to see that when the subject of “deflation” gets really popular in the press, that tends to mark a top for T-Bond prices.  How much of a top may be different from one instance to another, but the principle is relatively consistent.

This is similar to another comparison I featured here back in 2012 comparing actual oil prices to the frequency at which the term “oil prices” had been featured in news stories.  A high frequency of mentions usually coincided with high oil prices, although sometimes it could also be a sign of abnormally low oil prices.

You can play around with Google Trends to find the different rates of popularity of different terms, and if you do you will probably see a larger number of gross headline counts involving the words “Bieber” or “Kardashian” versus “deflation”.  I’m not sure of what financial data to attempt to correlate to the varying interest in the Kardashians.  The term “New York” shows a generalized downward trend.  But “Dubai” is trending upward.  Playing around with it can get addictive.

The basic point for beginning contrarians is to disbelieve most of whatever you are hearing repeated in the financial media.  And the more it is being repeated, the more you should disbelieve it.  Learn to trust your own analysis, rather than listening to the talking head du jour on TV, even if it is me.

Related Charts

Feb 24, 2012small chart
Using Google Trends To Mark A Price Climax
Jan 27, 2012small chart
Traders Like QQQ A Little Too Much
Oct 03, 2014small chart
Commercials Betting On Big Dollar Downturn

Chart In Focus Archive

2nd Presidential Years Are Different

Tom McClellan – McClellan Market Report

By now most who follow the stock market know that the first two years of each presidential term are basically flat, but the good times come in the 3rd year which is nearly always an up year.  What a lot of people do not know is that it matters a lot in the 2nd year whether you have a first term president or a second termer.

This week’s chart helps to make that point.  To create each plot, I took the historical data for the SP500 (or its predecessor the Cowles Index) going back to 1936 and chopped the data into 4-year chunks, starting at the beginning of November in each presidential election year.  November is when the election is, and my assumption is that the market starts reacting to the results of the election as soon as those results are known, rather than waiting until the victor is actually inaugurated.

Presidential Cycle Pattern versions for 1st or 2nd term presidents

With the data chopped up into those 4-year chunks, I then adjust each period to fit a uniform calendar.  This is necessary because up until 1952, the stock market used to trade on Saturdays.  And then in the summertime in the 1960s, the NYSE used to shut down on Wednesdays so that the clerks could get caught up on all of the paperwork for the increasing volume of trades.  Then I would reset the daily values in each 4-year chunk to reflect the percentage change since the beginning, the better to average them all together.

For this week’s chart, there is one additional step, which is to segregate the 4-year chunks of time into groups based on whether the president was in a first or second term.

Why is this necessary?  The stock market tends to behave differently depending on whether there is a new president from a different party than the last one, or a reelected incumbent.  Generally speaking, a 1st term president tends to spend a lot of time during his first two years “discovering” that conditions are even worse than he told us during the campaign, and proclaiming that the “only solution” is whatever package of tax hikes/cuts or spending boosts/reductions he is proposing.  Investors generally get discouraged upon hearing that things are worse than previously thought, and that shows up in the movements of stock prices.

An incumbent who wins reelection generally does not spend as much time blaming the guy who preceded him, and gets on with continuing what he was doing.  And the stock market responds accordingly.

By the 3rd year, those differences are gone, at least according to the historical record, and the stock market generally behaves in a similar fashion regardless of what type of new or old president is in office.  We’ll get to the 3rd year of this current presidential term in a few more months, but first we have to get through the unusual period in the summer of the 2nd year.

With a first term president in office, the market typically sees a price peak in November of the first year, and then a decline into the summer of the second year.  That summer bottom typically leads to a rally into autumn, and continuing right into the 3rd year.

But the 2nd year is almost the exact opposite with a 2nd term president in office like we have now.  The market typically reaches a peak in July (not in May, as so many believe), and then drops for the rest of the summer toward a September low.

To help see the difference, here is a chart showing both of these versions of the Presidential Cycle Pattern and comparing them to the action in the SP500 since the 2012 election.

SP500 and Presidential Cycle Patterns 1st and 2nd term presidents

It quickly becomes obvious that the current market action is behaving a whole lot more like the average pattern for 2nd term presidents, which is as one would expect.  The magnitude of the movements has been greater than average, which is also normal, and not all that remarkable.  The Presidential Cycle Pattern and its variants are much more useful for telling us about direction than about magnitude.

Here is another version, showing only the 2nd termer version, and zooming in closer on just the most recent data:

Presidential Cycle Pattern 2nd term presidents

It shows that the steep drop which began from the July 24 top is happening pretty much on schedule.  A brief pause in the decline is due over the next few days, and then a resumption of the decline should appear.  The ideal bottom equates to around Sep. 10, although the bottom to go up out of does not appear until a month later.

A few months from now, we’ll get to the bullish conditions of the 3rd year of a presidential term.  But the market has a less than bullish period to get through first.

Related Charts

Nov 05, 2010small chart
Entering the 3rd Presidential Year
Sep 27, 2013small chart
Seasonal Pattern
Jul 24, 2014small chart
A-D Line Divergence Again

Chart In Focus Archive

A Scary Valuation Indicator

Tom McClellan – McClellan Market Report

If you want to get worried about long-term stock market valuations, this week’s chart should do the job.  I saw a version of this chart recently in a research report by Daniel J. Want, who is Analytics Director for Prerequisite Capital Management in Queensland, Australia.

What is unique about this indicator is that it combines two separate data series into one indicator.  When I first saw the version of this chart in Daniel Want’s report, I thought it was just wrong to put the CAPE data together with the Moody’s Baa yield, since the two are actually strongly correlated.  So combining them together just magnifies the same message.

CAPE ratio adjusted by Baa yield

This is because the P/E ratio is the inverse of the “earnings yield”, which should reasonably match up with bond yields.  If an investor can get a better return on his money in the bond market, then he will flee the stock market, or vice versa.  That is what keeps the earnings yield and bond yields in correlation.  But when investors are bidding up stock prices to a ridiculous point such that the earnings yield is way out of whack from the bond interest yield, then there can be a big problem.

But the more I thought about it, the more I saw the beauty of this approach of combining the two.  Low interest rates are usually associated with periods of high P/E ratios, but not always.  Sometimes the market’s average P/E ratio can get to a high level because earnings fall to a low level in a recession, as opposed to prices going way up above where they should be.  In 2008-09, for example, the SP500’s raw P/E ratio sky-rocketed, not because prices were too high, but because earnings were unnaturally too low.

SP500 Raw P/E Ratio

Analysts have struggled for years to make meaningful use of overall market P/E ratios because of problems like these.  Professor Robert Shiller of Yale University uses one method to deal with this, which is to calculate a “Cyclically Adjusted P/E” (CAPE) ratio.  This is the same Prof. Shiller who wrote the 2000 book “Irrational Exuberance“, and whose name is attached to the S&P Case-Shiller Home Price Indices.

Shiller’s CAPE looks at each month’s inflation-adjusted (by CPI) real price for the SP500, and compares it to the 10-year average of real earnings.  The intent is to smooth out the earnings fluctuations across one or more business cycles, which is an idea that dates back all the way to the work of Graham and Dodd in the 1930s.  CAPE does a better job than the raw P/E ratio of identifying extreme high and low valuation levels:

SP500 vs. CAPE

The extreme peaks and valleys are smoothed down somewhat, but we still are confronted with the difficult task of determining how high is “too high”.  The CAPE peak at the 2000 bubble top was well above everything else in the record, even including the 1929 stock market top which preceded the Great Depression.  But just because the CAPE level got up that high did not mean that the stock market had to roll over right away.  It stayed at a lofty CAPE ratio level for many months before the Internet bubble finally burst and caused stock prices to turn downward.

The indicator which Daniel Want featured in his report seems to solve some of those problems.  For the record, Mr. Want states that he saw it elsewhere and cannot remember the source, but he added it to his toolbox a long time ago after seeing the nice properties it displays.  I am a big fan of giving attribution to the creators of original ideas, whenever I can figure out who they are.  So I’ll give a tip of the hat to Mr. Want from whom I learned about this tool, even if I cannot yet discern the true originator.

“Baa” is a bond quality rating assigned by Moody’s, and it means that the bond instruments with that rating “are subject to moderate credit risk. They are considered medium-grade and as such may possess certain speculative characteristics.”  The monthly Baa rate used in this week’s chart comes from the St. Louis Federal Reserve’s Economic Database (FRED), and is the monthly average of daily yield data on all of the Baa-rated corporate bonds.

The magic of the composite indicator in this week’s chart is that by adjusting CAPE for the Moody’s Baa yield, the result seems to set a much more uniform ceiling for how high valuations can go.  Getting above a certain level says that the market is really getting to the edge, and is near a MAJOR price top.  And that is the message of this week’s lead chart, which shows that this CAPE/Baa ratio is now up to the sort of level which has always marked a MAJOR stock market top every time it has been reached.  That’s obviously a problem for the market, but it is not necessarily an immediate problem.  And I want to emphasize that the message is that the market is NEAR a major top, and not necessarily AT one.

Looking at past major tops, the final price top does not usually arrive while the CAPE/Baa ratio is in its steepest slope of an up move, like that which we are seeing right now.  We also see that the price peaks in 1966 and 2000 came after this indicator had already peaked and turned down.  The point is that there is considerable variability about the nature of how tops are built, and it can take a while to complete the process.  The message for now is that the rubber band is getting stretched really far, and that is a problem for the really long run for stock prices.  Today does not have to be the final moment for the long bull run, but it is the time to begin planning for the final moment.

Related Charts

Mar 23, 2012

small chart
SP500 Undervalued Versus M2

Nov 19, 2010

small chart
Is Gold Overvalued?

Jan 21, 2011

small chart
Why Even Fundamental Analysts Should Watch A-D Line

Chart In Focus Archive

Alarming Sign in NDX Stocks’ Drawdown

Tom McClellan – McClellan Market Report

The Nasdaq 100 Index is making new multi-year highs, levels not seen since the weeks just after the 2000 Internet Bubble top.  But it is interesting for us to see that the average component stock in that index is down 7% from its trailing 52-week high.

And that 7% drawdown number is actually smaller than it has been recently, but it is still not back to the low drawdown reading of 6.0% that we saw in early March, before the Nasdaq 100 stocks got into a patch of trouble.  And the fascinating point is that divergences like this tend to be really important for the long-run picture for stock prices.

This current divergence is not guaranteed to stay with us.  It looks like a genuine divergence now, but it is still possible that we could see a continued rally that takes the stocks in the Nasdaq 100 collectively up closer to the level of their 52-week closing highs, thereby closing the gap on this measure of average drawdown.  But for now, the message is that the average stock making up the Nasdaq 100 Index is not confirming the bullish message of the NDX’s higher price highs.

Nasdaq 100 stocks average drawdown amount

A similar message comes from a similar indicator, this time examining the 30 stocks which make up the DJIA.

DJI Oscillator Positive Index

This one looks at the 30 stocks that make up the DJIA, and checks each one to see if its Price Oscillator is above or below zero.  What we are seeing lately is a declining number of Dow components participating in the uptrend that way, and a drop of this indicator below its 15-day moving average.  This condition also comes as a divergence appears between prices and the indicator.

As with the NDX stocks’ indicator above, the divergence does not absolutely have to persist; the market could just power through it, but that is not usually the way things usually work out.

As a final aside, I appeared on CNBC this week to discuss last week’s Chart In Focus topic.  You can see a portion of that interview at

Related Charts

Oct 11, 2013

small chart
New Highs’ Divergence

Jan 06, 2012

small chart
DJI Oscillator Rising Index Signals Trouble

Jul 28, 2011

small chart
NDX Stocks Show Troubling Drawdowns

Chart In Focus Archive

VIX Below 12!

Tom McClellan – McClellan Market Report

The CBOE Volatility Index (VIX) has dropped below 12.  Below 12!!!  Surely this must be one of the signs of the apocalypse.

Not so fast.  A low VIX is a sign of option trader complacency, and complacency is a problematic sign for the market which can lead to a price decline.  But when the VIX gets really REALLY low, the message changes.  Sure, it is a sign of an absence of worry, and a correction is possible.  But the really big price tops do not appear when the VIX is this low.

In 2000, the SP500 topped after the VIX had climbed up from the really low levels of a few years earlier.  The final price top came when the VIX was at 16.54.

In 2007, the SP500 again topped with the VIX above 16.  It had been much lower earlier that year, bottoming below 12 in April 2007 well in advance of the final price high.  This is a consistent theme.

VIX below 12

When a toy top spins, it is at its most stable point when the RPM are high enough to maintain stability.  Gradually the friction against the tip spinning on the floor slows the top, and the top starts to wobble.  The wobbling is a prelude to the top falling over completely.  In other words, instability increases ahead of the tipping point.  When everything is as stable as it can be, that is not yet the tipping point.  So while a VIX reading below 12 is a quite spectacular condition, history says it is not the sign of an ultimate price top.

On a shorter term time scale, the VIX can give us a hint of a brief change in trend direction.  It is normal for the VIX to make a lower low as the SP500 makes a higher high.  But when there is a disagreement, THAT is when we get useful information from the VIX.  We have a tiny little divergence right now, with the VIX making higher postings at the same time when the SP500 is also making higher highs.  On a short term basis, this is a sign of the spinning top starting to wobble, and sets up the opportunity for a countertrend price drop.

VIX SP500 with 50-1 Bollinger Bands

The divergence is a setup for potential trouble.  The real trouble for short term price movements typically comes when the VIX rises up above its 50-day moving average (50MA).  That is a topic I addressed here.

But we should not confuse a short term setup for a corrective move with the signs of a more important long term top.  Seeing the VIX go this low is NOT the sort of condition consistent with the major tops of the past.

Related Charts

Jul 23, 2010

small chart
Golf Balls, Geese, and the VIX

Sep 10, 2010

small chart
Year-End VIX Plunge Coming

Jan 02, 2014

small chart
VIX ETN Not Right For Investors

Chart In Focus Archive

Equity Options vs. Index Options

Tom McClellan – McClellan Market Report

Most traders who look at options market data for market insights like to focus on the “Put/Call Volume Ratio”, which examines the differences between the number of trades in bullish versus bearish bets.  I use that data in my work, and even get into differences between the Put/Call Ratio for equity options versus index options.

There can also be useful information just in looking at the total volume in each category, which is what this week’s chart does.  It originated from a desire to know what it means when comparatively more of one category gets traded, and with a little bit of smoothing it makes for a nice indicator.

What we are looking at is a 10-day moving average of the daily ratio of total equity options volume versus index option volume as traded on the CBOE.  It sums together the puts and the calls in each category.  The range of values tends to wander a bit over time, and so I have also added 50-1 Bollinger Bands (50-day MA, and bands are set 1 standard deviation above and below).  When this indicator goes up above the upper 50-1 band, it can indicate a topping condition for prices, and the opposite meaning comes from dips below the lower band.

Equity versus Index option volume

The one huge caveat is that this indicator tends to work poorly around the time of quarterly options and futures expirations.  The artificial volume related to exercising options and reestablishing positions can skew the numbers for a brief period.  So that is an important factor to remember when using this indicator. 

That is not a condition we face right now, as this indicator has just dipped below the lower band.  It is an unusual condition to see the indicator this low while prices are making higher highs, but it has nevertheless happened  before.  The meaning is still bullish even at a higher price high.  Some examples are highlighted.

The implication is that the breakout move in the SP500 should be able to continue higher because of the sentiment message revealed by this indicator.

Related Charts

Jan 11, 2013

small chart
The Disappearing Volume

May 22, 2014

small chart
ETF Volume is Different

Jan 02, 2014

small chart
VIX ETN Not Right For Investors

Chart In Focus Archive