A Sentiment Primer (Long)

A reader asked about how to get up and running with sentiment indicators, where to get data, what time frames to use, etc. Since this is not a subject I have broached here, let me use this opportunity to provide an initial overview of some of my thoughts on sentiment indicators.

In terms of context, I consider fundamental analysis, technical analysis and market sentiment analysis to be the three primary legs of the investment stool. I believe it is a common pattern for relative newcomers to the investing world to begin with a fundamental analysis perspective, start wondering why individual stock don’t move ‘like they should,’ then add some technical analysis tools. Often it takes awhile to get the hang of TA; once they do, most investors get blindsided when thee entire forest moves abruptly while they are focusing on an individual tree or two. This is often the catalyst that leads to a more in-depth examination of market sentiment indicators.

Sentiment as a Contrarian Indicator

With that out of the way, where should you start with investor sentiment? First, keep in mind that much of market sentiment is a contrarian tool. One of my favorite quotes comes from John Bender in Jack Schwager’s Stock Market Wizards, “It’s not the current opinion on the stock that matters, but rather the potential change in the opinion that matters.” When your great aunt, housekeeper and taxi driver all start telling you how much money they are making in the stock market, who is left to buy and drive prices up further? On the other hand, if most of the people you know confess to having recently lost over half of their money in the market and swear about “never investing in stocks again,” then the markets have probably just about run out of sellers.

By all means, keep tabs of anecdotal evidence from novice investors you know and consider how much easy money you could make by taking the other side of every trade that a novice investor makes.

Types of Sentiment Indicators

In a nutshell, sentiment indicators attempt to discern what unsophisticated investors are thinking, feeling and doing, then encourage you to take the opposite position when their fear or greed reaches extreme levels.

Some sentiment indicators are compiled based on a direct survey of investors to determine if they are bullish or bearish. Three of the more famous of these are Investors Intelligence, the American Association of Individual Investors (AAII) and Market Vane. Newer additions to the fold are Birinyi’s Blogger Sentiment Poll and LowRisk.com’s Investor Sentiment Indicator. Since we know that what some people say and actually do are often two very different things, much of sentiment analysis looks at the activity of supposedly unsophisticated investors in order to get a better sense of which direction they are leaning. The ISEE (call to put ratio) is one such measure of investor activity; the CBOE Equity put to call ratio is another; and the Public Short Sales data is another good data point. To the extent you are able to, use ratios and other tools to compare and contrast the actions of the ‘little guy’ with that of institutions.

Sentiment Data Sources

In terms of sources, StockCharts.com has a great (and free) Market Summary page that is an excellent snapshot of what is going on in the 100+ most important markets, sectors, countries, etc. Scroll down to the bottom two groups and you'll find many sentiment indicators I follow in the "Market Breadth" section. You may or may not already be familiar with the "Bullish Percent Indices" – if you aren't, these are something you should take some time to educate yourself on in the future. Each indicator has links to three kinds of charts – and if you click directly on the name of the indicator, you will pull up the default gallery view.

Another excellent free site with a different set of indicators can be found at Market Gauge - Today's Indicators. I keep an eye on the “Contrary Opinion” data, in particular. Note that each line has a chart link in the far right hand side of the page. You may want to also bookmark Market Gauge's Market Summary page. Finally, I only recently discovered InvestmentTools.com, which has some superb “Weekly Sentiment Indicators” as well as some valuable “Short Sales” data.

Two other recommended sources for market sentiment are Market Harmonics (consider the "Volume" and "Momentum" data at some point in the future too) and the "Power Tools" at SchaeffersResearch.com. Once again, start with the “Sentiment Data” section, but eventually look at the others as well. In terms of blog sources, HeadlineCharts probably does more with market sentiment than any of the others I read and TheSentimentals.com has as comprehensive a list of links to market sentiment data as I have seen. For a weekly recap of sentiment data and a commentary on what is happening in the world of sentiment and market internals, check out Fred Ruffy's "Sentiment Journal" column at Optionetics.com.

Whenever possible, I suggest you get data directly from the exchanges, such as ISEE data from the ISE and CBOE VIX data and CBOE Put/Call ratio data from the CBOE. You can download historical data from Yahoo in spreadsheet format for the likes of the VIX and many other indices. As a rule, you should try to get this same data from an exchange or web site dedicated to this index first and use Yahoo only as a last resort.

Sentiment and Time Frames

Regarding time frames, your typical holding period should dictate the time frames you look at. Are most of your trades day trades? swing trades of 2-10 days? have holding periods of 1-3 months? Maybe you trade in multiple time frames (dangerous and confusing at times, but ultimately not necessarily a bad thing to do.)

I have never seen anyone articulate this, but my personal experience is that the charting time horizon for support, resistance, moving averages, recognition of common patterns, etc. ought to be something on the order of 20-50x your typical holding period. Most of my trades are of the 2-10 day variety, so most of the charts I look at are in the 20 day to 1 year range. I like to look at charts of 2 years or more for historical perspective on candidates I have already screened and I sometimes to go down as low as 1 minute bars for intra-day charts to fine tune entries and exits, but for the most part, I live in the 1-6 month charting world. Typically the longer term charts identify the opportunity and the shorter term charts trigger the timing of the entry.

I rarely care much about intra-day sentiment and look at a lot of moving averages in the 5, 10 and 20 day range, sometimes up to 200 days/40 weeks. Your sentiment time horizon should probably match your overall charting time horizon, but I haven't found much bang for my energy/attention buck focusing on intra-day or even day to day sentiment movement. Look for some smoothing factor, such as simple or exponential moving averages, to help minimize the noise.

On-Line Resources and Books

Regarding books and other educational sources, I can't say I have any great ones up my sleeve.

As noted earlier, for an on-line source, StockCharts.com does as good a job as anyone in terms of education. I would definitely start with their Introduction to Market Indicators and go from there, perhaps backing up to their entire Chart School series. You should probably bookmark their Glossary for future reference too.

For a broad brush perspective on sentiment indicators, I highly recommend that you check out an excellent post by Barry Ritholtz of The Big Picture, "Contrary Indicators 2000 - 2003 Bear." You should also download the full PDF or Word document that he links to and study the analysis of various internal and external indicators, then look at some of these and how they performed in and around May-July 2006 as well as around February 2007 – or any other period you are interested in.

If you want a treasure trove of ideas on sentiment, I heartily recommend a visit to TraderFeed.com, where a keyword search on "sentiment" will provide you with many hours of reading from the archives of Brett Steenbarger. If this is news to you, then you just aren't paying attention...

I am embarrassed to admit that I have not yet bothered with the free trial of Jason Goepfert's SentimenTrader.com, but from what I have seen of his work, I would recommend you test drive his site with a free trial. At a minimum, follow the link above to see what sort of indicators he thinks are worth following.

The last time I did a summary of recommended investment books was about 9 months ago. I need to update that list and put it on the blog, but I can tell you that there is not a great book on investor sentiment in there. Gary Smith’s How I Trade for a Living is probably my favorite treatment of market sentiment. Though it is not on the above list, I thought Toni Turner did an excellent job with sentiment and many other topics in Short-Term Trading in the New Stock Market. For my money, Turner’s book is one of the better ones for a relative newbie, albeit with a definite short-term bias. Two other books worth checking out for their discussion of sentiment issues are Martin Pring's encyclopedic Technical Analysis Explained: The Successful Investor's Guide to Spotting Investment Trends and Turning Points and New Thinking in Technical Analysis: Trading Models from the Masters, a patchwork of ideas from many respected thinkers, edited by Rick Bensignor, which includes a chapter on sentiment by Bernie Schaeffer, "Enhancing Technical Analysis by Incorporating Sentiment."

ISEE 50 Day SMA Is Bullish

A week ago, I posted about the ISEE and the predictive power of the 50 day SMA. I have since updated that chart to reflect a buy signal at 120 (-1.5 standard deviations from the mean) and sell signal at 192 (+1.5 standard deviations from the mean.)

After hitting a record low on March 8^th and printing three record lows for the 10 day SMA two weeks ago, the ISEE’s 50 day SMA is now comfortably below the 120 buy level. Perhaps more importantly, with extremely high numbers due to scroll off the average in the next 10 days and relatively high numbers set to scroll off over the next month, it is a good bet that the 50 day SMA will be flashing a buy signal for most of April, if not the entire month.

If you put any faith in put to call ratios, then this is the time you want to get long, perhaps in a big way.

CNBC Million Dollar Portfolio Challenge Update

I am happy to report that as of last night, I am now in the top 1% in the CNBC Million Dollar Portfolio Challenge at #4577 out of what appears to be about a half million participants.

I was going to keep this news under wraps until the weekend, but since I like to call market tops a little early, I thought I'd share the news today in the event my position starts to 'correct' a little.

Briefly, energy and copper stocks have helped propel me to my current standing and I am heavily weighted in oil, gas and uranium at the moment. In most stock contests, it pays to take ridiculous risks in hopes of lucking into a triple digit return. I have not yet succumbed to that temptation. Instead, making largely incremental gains here and there, I am up 27%. As long as I continue to outperform the market and move up the leader board, I will not be putting all my chips on my favorite number -- until perhaps the last 2-3 weeks of the competition.

For those who may be interested, the photo to the left is of Clifford Brown, one of my favorite trumpet players. I figured that if I were going to toot my own horn, I should also acknowledge the work of a true master, who, sadly, had so little time to dazzle us with his talent.

Update: According the Portfolio Challenge blog, the top ten contestants are already sitting on 100% gains, so I may have to rethink my strategy sooner rather than later...

SPX:VIX Back to Predicted Level

A week before the February 27^th VIX spike, I talked about the ratio of the SPX to the VIX and posited that the best way to think about that relationship would be to look at the oscillating VIX number in the context of a SPX that is trending approximately 10% over the long term.

The day after the February 27^th spike, I revisited the SPX:VIX ratio, which had dropped precipitously from 138 to 76 (on the monthly chart) and commented that I considered a ‘neutral range’ for this ratio to be in the area of 105-115.

This seems like as good a time as any to update the SPX:VIX ratio, which rebounded all the way to 111 on the weekly chart (below) and now sits at 106. In other words, the SPX:VIX ratio is now just about exactly at the midpoint between the extremes of pre-2/27 complacency and post-2/27 panic – and also very close to the long-term SPX trend line that reflects a 10% annual increase.

If it feels like the forces of bullishness and bearishness are at a standstill at the moment, then it is perhaps because at the current levels, the battle is a draw.

Breaking News on the Ursine Front!

Trading Goddess outs Tim Knight

...and Tim comes forth with a full confession and offers up his side of the story

Fun with Technorati and Keyword Trends

Last week I added a small Technorati chart to this blog in the right hand column, just above the “Recent VIX Readers” widget. As the label indicates, the chart shows the number of posts on all blogs that have included the word “VIX” in them over the past 30 days, as measured by Technorati.

I added the chart largely for informational purposes, but since 30 days have elapsed since the February 27^th volatility eruption, I thought it might be interesting to track the use of the word “VIX” in the blogosphere and see how well it correlates with the index itself. Much to my surprise, the correlation is extremely high, with very little lag, as the graph below indicates.

Theoretically, you can do something similar with Google Trends (including slice and dice by region and date,) but if you try to do this with “VIX” the results look much less compelling.

Getting back to Technorati, you can use their own drop down menu parameters at http://technorati.com/chart/VIX?sub=chartlet. You can insert any keyword or keywords you wish, even search for an exact phrase using quotation marks. I suggest starting with something topical like “subprime” and going from there.

If you wish too fine tune this tool by customizing some of the parameters beyond the options available via the drop-down menu, you can edit the URL directly. Techies will probably know how to do this better than I can, but the URL for the basic VIX graph on my blog is: http://technorati.com/chartimg/%28VIX%29?totalHits=12325&size=s&days=30

You can customize some of the parameters as follows:

• VIX = “any key word(s)”
  
  
• size = y (where y = “s” “m” “l” or “xl”) – I highly recommend “xl” for any analytical work
  
  
• days = zzz (where zzz is any number between 1 and 366) -- unfortunately, I believe this feature is only available for the past year
  

Contrarians and trend faders, have at it!

Jaws of Opportunity?

With the VIX sporting a 30 day historical volatility of 206 (a 52 week high) and a current implied volatility of only 85 (as of EOD Friday), it is reasonable to wonder whether this wide gap is an opportunity to stock up on some VIX calls.

Well…since I was unable to get this post up before the new home sales data came out, some of the opportunity has already disappeared, but, truth be told, the divergence in the chart is now only slightly narrower than is shown in the iVolatility chart below.

At current levels (low 14s) I am generally neutral with respect to the direction of the VIX. You can make a case for the high historical volatility as a signal to buy the market and/or sell the VIX; you can also make a case for the relatively low IV in the VIX as an opportunity to go long the VIX. I consider most of what shows up in the graph below to be an artifact of the sharp correction and elevated emotions associated with a perception of increased risk.

Where some may see opportunity in this divergence, others may recall what happened to Quint when he got a little too Ahab for his own good.

When in doubt, I usually recommend exercising some patience and waiting for better opportunities to present themselves.

Portfolio A1 Update for 3/25/07

Portfolio A1 rallied vigorously last week and now leads the benchmark SPX by 1.7%. Four of the five holdings (all but NTY) are within 3% of their 52 week highs, with TEX logging a new all-time high on Friday.

There are no changes to the portfolio for the coming week.

Though I am still not sure of the significance, the equity curve continues to strongly resemble a duck, although as portfolio performance improves I am ever hopeful that we can fly out of range of Dick Cheney.

(On a distantly related note, I am pleased to report that I have moved in to the top 5,000 at the CNBC Million Dollar Portfolio Challenge. I am currently in the top 2% of all contestants for this challenge, using a discretionary trading system that has recently made strong gains in natural gas and copper. My goal is to be comfortably in the top 1% and taking a run at the money when the contest ends on May 14. I will pass along any news of interest as it develops.)

Portfolio details for Portfolio A1 are as follows:

Complacency Creeping Back In; VWSI at +1

Quite a few commentators, including Tim Knight, have pointed out that the VIX has reversed all of the post 2/27 spike. While volatility measures are subsiding much more rapidly than put to call ratios, it is safe to say that some investors are thinking that the brief correction has passed and the bull market can now safely resume its course. In short, some complacency is creeping back into the market, albeit slowly.

The VIX Weekly Sentiment Indicator (VWSI) reflects some of this complacency, as it has hugged the 0 mark the past few weeks, currently sitting at +1.

While the VWSI generally seeks to predict VIX movements looking out only 1-2 weeks, this is as good a time as any to remind the reader that longer term seasonal cycles are at work as well. You may never have heard, “Sell in March and go away,” but the March-June period marks the downside cycle of the mid-year “V” pattern in the VIX that I have commented on earlier.

I would not be looking to anticipate volatility spikes at this stage, but I would expect that we will see a few more smaller ones in the coming weeks. Better yet, be prepared to play a little Whac-a-Mole and fade these spikes as they occur.

(Note that in the above temperature gauge, the "bullish" and "bearish" labels apply to the VIX, not to the broader markets, which are usually negatively correlated with the VIX.)

Wine pairing: A while back, a reader had the temerity to suggest that I might want to consider offering a wine pairing with the VWSI. Why not? I've been neglecting my wine blog, so why not jump start the latent oenophile in all of us. For the current +1 reading, coming off of some violent volatility spikes and ushering in the new season, I am recommend a dry gewurztraminer. Gewurztraminer is a wine that can stand up to all seasons and all levels of volaltility. It has more personality than any other white wine, is found in more styles and...it's not chardonnay. While the classic version comes from Alsace, I am a big fan of several American versions of this wine: dry; off-dry; and the dessert variety. One winery that gets all three versions right is Navarro. Last weekend, however, I found a stunning example of dry gewurz off the beaten track in the Russian River Valley at Harvest Moon. I hesitate to mention Harvest Moon because it is such a small winery, but the wines are so good that they deserve a wider audience. Enjoy!

20% Under the 10 Day SMA, Then What?

As usual, Adam Warner of the Daily Options Report has been all over the latest developments in VIX. He was the first to comment on the VIX falling 20% under its 10 day SMA on Wednesday (actually -19.3%) and has added two follow-up stories, most recently this morning, where he draws comparisons to the June-July VIX walkabout from last year.

To recap for those why may be link shy, the VIX has closed 20% below the 10 day SMA on seven days since 1990, which I have grouped into four distinct events (one isolated event and three other events with two separate EOD readings,) as follows:

> 8/15-16/2002
> 12/23/1998
> 3/14-15/1991 (3/13 was 19.9% below)
> 1/21/91 and 1/24/91 (1/18 through 1/25 were all at least 18% below)

The only additions I will make to Adam’s commentary are two graphs that appear in one form or another on these pages on a fairly regular basis: a composite look at all 7 instances, from 5 days prior to 20 days after the -20% reading; and a rather busy graph of each of those 7 instances, color coded by ‘event,’ with the second -20% reading for each event indicated by a dotted line. The graphs, not surprisingly, suggest a possible mean-reverting move over the next 10-20 trading days, but given the small sample size, I would consider their entertainment/voyeuristic value to be higher than any informational value.

Cherry Picking Other Opinions…and an Occasional Fact

The current installment of my survey of recent comments about life on the fringe of the VIXiverse and beyond:

• Mebane Faber of World Beta with an interesting look at mean reversion (a tip of the hat to CASTrader.com)
  
  
• Bernie Schaeffer of SchaeffersResearch.com with a deep dive on put to call ratios
  
  
• Adam Warner of Daily Options Report commenting on the rarity of the VIX falling 20% below the 10 day SMA (I get specific in his comments section and will talk in more detail about this in VIX and More soon)
  
  
• Philip McDonnell of Daily Speculations on a high variability distribution model for security prices
  
  
• Bill Rempel has some thoughts (buried toward the bottom of the comments) on how the VIX might be decomposed and better modeled (I’ll offer more on some of this thinking in due course)
  
  
• …and my favorite end of day commentary, The dk Report, is back up and running.

Fed Links

In addition to the links below, I intend to have an in-depth analysis of the VIX on Fed Days for the next FOMC meeting on May 9th.

General Fed links:

• Calculate the probability of a rate cut with the guidance of the Cleveland Fed
• Keep track of the latest from Ben Bernanke and other breaking news at the Fed's main site
• Get FOMC minutes, a meeting calendar, etc. (even bios of your favorite hawks) on the FOMC page

If you want links to speeches from any of the FOMC members, unfortunately it gets a little messy, as you have to go to the individual site for each regional Fed branch:

• Atlanta
• Boston
• Chicago
• Cleveland
• Dallas
• Kansas City
• Minneapolis
• New York
• Philadelphia
• Richmond
• San Francisco
• St. Louis

(FWIW, the slow market pre-announcement period is a good time to get caught up on your reading...)

Now for the good stuff:

• A database of 5 minute charts of the E-Mini S&P 500 futures for each Fed Day going back to May 1998 courtesy of MyPivots.com
• TickerSense looks at SPX performance on Fed Days and in the week that followed since the beginning of the tightening cycle in June 2004 in this handy chart
• The crack research staff at VIX and More takes an initial look at VIX price movement around Fed Days

Simple Moving Averages and the ISEE

In "A First Look at the ISEE" a reader asked about September 2006 and the predictive power of the ISEE.

Like most oscillators, the ISEE is great for getting you to buy on a pullback and less effective in timing when longer term bear trends will turn around. Depending upon where you set your trigger (say, a 10 day SMA of 120, 110 or 100), you can fine tune the percentage of downturns you want to catch and how early you might be in calling a bottom.

As an accomplished over-trader (not to be confused with uber-trader), I have a tendency to try to call turns too early. If you suffer from a similar affliction or just want to eliminate some noise from your charts, you might prefer to use a longer SMA with the ISEE, such as the 50 day version.

In the chart of the 50 day SMA below, you can see that while absolute values of the average are important, the rolling over of the trend is even more telling.

Getting back to September 2006, no the 50 day SMA had not rolled over then and it would not roll over until October. While looking at absolute numbers in the 50 day SMA might have had you long back in July, if you waited for the 50 day SMA to roll over, you would have missed out on the July to October rally.

This raises three points:

1. No indicator is perfect and the ISEE is a good illustration of that point
   
   
2. It is possible to maximize the value of the ISEE using both absolute SMA numbers and the rolling over of these SMAs
   
   
3. It can be very expensive to try to time markets exactly as they turn; generally, it is much more profitable to target the middle 50% of a move than to try to capture the entire move

A First Look at the ISEE

I briefly touched upon the ISEE yesterday; and while the ISE web site has a nice little graph of the ISEE going back a year, it is probably not sufficient to answer even the more casual questions about the usefulness of the index. Fortunately, the ISE has data going back to October 2002 available to download so that you can conduct your own analysis…or you can just keep reading.

For now I’m going to skip over the details of the debate about the usefulness of put to call ratios to help call market turns. If you want to do some research on this indicator, a good place to start would be Bill Rempel’s “The Significance of the Equity Put to Call Ratio” commentary at MarketThoughts.com.

Suffice it to say that I believe that it is worth the effort to watch this contrarian sentiment indicator and suggest that a good place to do so is the SchaeffersResearch.com page that graphs the CBOE equity (non-index) put to call ratio. You can create your own CBOE put to call charts at Stockcharts.com using $CPCE as the equity put to call ticker, $CPCI for the index put to call ticker, and $CPC for the total equity + index data.

The ISEE is actually a call to put ratio based entirely on purchases of calls and puts in which a customer is opening a position through the ISE itself. For more details and a testimonial of sorts I recommend a Wall Street Journal article by Mohammed Hadi.

What do the ISEE data look like? I have graphed the 10, 20 and 100 day SMAs going back to 2002. They show a mean of 154, with obvious extreme readings falling conveniently at round numbers: below 100 and above 200. As a contrarian indicator, a reading of 100 or below is a good cutoff for overly bearish sentiment and a high likelihood of a bullish move ahead; 200 signals extreme bullishness and an increased chance of a bearish turn in the markets in the near future. In many respects, a high ISEE number is a lot like a low VIX number and vice-versa.

Turning to the ISEE as a predictor of future market performance, this is a subject that I will come back to regularly in this space, but for now I have posted a graph of the SPX and the 10 day SMA of the ISEE for the period 10/1/02 to 3/19/07. Note that the four largest spikes in the ISEE all preceded significant moves down in the SPX; also, low ISEE readings frequently marked the beginning of bullish moves.

Before wrapping up this subject for today, I will leave the reader with one last thought: the lowest 10 day SMA ever recorded for the ISEE was an 87.00 last Thursday.

Good Time to Take Some Profits

With the VIX at 13.52 and expiration coming tomorrow, this is a good time to take some profits on any remaining fade the spike trades.

Note that the lowest VIX close since 2/27 is 13.99, so the risk/reward profile on any mean reversion trades continues to deteriorate with each click the VIX makes down.

Dr. Brett on Put to Call Ratios

I was away for the weekend and am still catching up in my reading, but it looks like Brett Steenbarger saved me a post on the abnormally low put to call ratios.

According to Dr. Brett, Jason Goepfert’s SentimenTrader.com apparently has a more in-depth discussion of the put to call issue. I should add that I am not a subscriber to SentimenTrader, but have had it on my To Do list for a couple of months.

Finally, some of the more observant readers may have noticed that I added a link to the ISEE at the upper right hand corner of this blog last week. The ISEE is a sentiment index (in this case, a ratio of call to put options, multiplied by 100) compiled by the International Securities Exchange or ISE. In some respects, it is similar to the put to call ratios at the CBOE, but without the heavy institutional action in the indices that tends to dominate the activity on the CBOE.

I have been looking hard at the ISEE in the past week because of the unusually low readings, including consecutive all-time lows (data go back to 2002) in the 10 day SMA last Wednesday and Thursday.

Stay tuned for more on put to call data, what it means, and its predictive power.

IT In and PCCC Out at Portfolio A1

As I suspected might happen last week, PCCC’s performance lag has finally caught up with it and it will be dropped from the portfolio tomorrow and repaced by IT, otherwise known as Gartner, the research powerhouse that has been on a tear as of late, up almost 200% since April 2005. IT made a new 52 week high on Thursday, the 31st new 52 week high in the last year. IT has also raised guidance three times in the past nine months and announced a $200 million share buyback just last month.

The portfolio rallied relative to the S&P 500 this week and now sits 0.32% below the benchmark index. RKT, which became the first addition to the portfolio two weeks ago, has also become the top performer in the portfolio, up 6.9%.

A snapshot of the portfolio is as follows:

VWSI at -2

After a classic mean-reverting snap back last week, the VIX Weekly Sentiment Indicator (VWSI) took more of a meandering approach in the most recent week, registering only a -3 reading following Tuesday’s 30% spike and a -4 on Wednesday, when the VIX topped out at 21.25, a reading not seen since June 2006. While technically any reading below zero has to be considered to be somewhat bearish for the VIX, the current -2 reading is still within the -3 to +3 neutral zone, where I do not consider the statistical edge and risk/reward profile of the VIX to warrant taking either a long or short position.

At a little less than 4% over the VIX’s 10 day SMA of 16.18, it is difficult to make a case that the VIX is over or undervalued at current levels. It is worth recalling that the VIX tends to fizzle on Fed Days and almost never spikes 10%, so it is particularly difficult to make the case for being long the VIX going into the upcoming Fed announcement. For those looking to trade the VIX, this is a good time to be patient and wait for better setups.

(Note that in the above temperature gauge, the "bullish" and "bearish" labels apply to the VIX, not to the broader markets, which are usually negatively correlated with the VIX.)

Meet the Spikers

Quadruple witching day seems like as good a day as any to meet the family. Having already introduced VDAX, the VIX’s German cousin, today we look at the US relatives. First, the handy chart:

Due to marriages, divorces, name changes and the like, a more detailed look at the US extended family can get complicated, but here are the five key players, roughly in order of their current significance in the markets:

VIX – measures implied volatility for S&P 500 (SPX) options for all near term at-the-money SPX puts and calls and out-of-the-money puts and calls. Deep-in-the-money options are excluded. The current methodology has been in use since 9/22/03. The VIX was officially introduced on 4/1/93, but the CBOE has calculated synthetic historical VIX data going back to the beginning of 1990.

VXN – measures implied volatility for the Nasdaq 100 (NDX) options for all near term at-the-money NDX puts and calls and out-of-the-money puts and calls. Deep-in-the-money options are excluded. The current methodology has been in use since 9/22/03; the index was originally introduced on 1/22/01.

VXO – is calculated by taking the weighted average of the implied volatility of 8 S&P 100 OEX calls and puts with an average time to expiration of 30 days. Note that this is the methodology that was used to calculate the VIX prior to 9/22/03. On that date, the method used to calculate the VIX was changed and a new ticker symbol and name was introduced to provide continuity with the historical method of calculating the "old VIX" prior to 9/22/03.

From the CBOE site:

"VIX measures market expectation of near term volatility conveyed by stock index option prices. The original VIX was constructed using the implied volatilities of eight different OEX option series so that, at any given time, it represented the implied volatility of a hypothetical at-the-money OEX option with exactly 30 days to expiration.

The New VIX still measures the market's expectation of 30-day volatility, but in a way that conforms to the latest thinking and research among industry practitioners. The New VIX is based on S&P 500 index option prices and incorporates information from the volatility ‘skew’ by using a wider range of strike prices rather than just at-the-money series."

VXD – measures implied volatility for the Dow Jones Industrial Average options for all near term at-the-money DJIA puts and calls and out-of-the-money puts and calls. Deep-in-the-money options are excluded. Introduced on 4/25/05.

RVX – measures implied volatility for the Russell 2000 (RUT) options for all near term at-the-money NDX puts and calls and out-of-the-money puts and calls. Deep-in-the-money options are excluded. Introduced on 5/5/06.

While there is a very high degree of correlation among the volatility indices, like any family, this family does not always move in lockstep fashion, as the graph below demonstrates. In the coming weeks, I will talk more about divergence among the volatility indices and attempt to provide a framework for interpreting them.

More Unprecedented Elasticity in the VIX

Clearly, the markets and the VIX took some interesting twists and turns yesterday, but how unusual were those movements?

One interesting statistic is that yesterday’s VIX trading range was 24.8% of the prior day’s close. This has only happened nine times in the last decade and fully 5 of these 9 instances have occurred in the last three weeks!

Another interesting statistic is that the VIX, which is known more for spiking up than spiking down, finished yesterday 18.7% below the intra-day high. This has happened just six times since 1990 and twice (last June and two weeks ago today) in the past eight years. For those who may have some interest in these matters, each of these six occasions happened to fall during a bull market.

I put together my faithful composite historical graph of those six instances. It shows evidence of some additional echo volatility in the subsequent 20 trading days and the lack of a dramatic reversion to the mean which most spikes tend to demonstrate over this period. (An important heuristic is at work here: intra-day spikes tend to have much less subsidence than end of day and multi-day spikes, partly because the nature of intra-day spikes is that much of the subsidence happens before the close.)

If you look at the individual data points, however, (which include the first nine trading days following the 3/1/07 turmoil) some echo volatility is present, but, for the most part, volatility lessens substantially after the first ten trading days.

CBOE Introduces New Benchmark Index - CBOE S&P 500 VARB-X Strategy Benchmark Provides Performance Measure for Volatility Arbitrage Trading

I normally don't like to quote press releases, but I found the CBOE’s announcement of a new benchmark index for selling volatility to be particularly interesting:

“The Chicago Board Options Exchange (CBOE) today announced that it will begin publishing a new benchmark index, the CBOE S&P 500 VARB-X Strategy Benchmark (VTY) on Friday, March 16, 2007. The new index tracks the performance of a hypothetical volatility arbitrage trading strategy designed to capitalize on the historical difference between S&P 500 Index (SPX) option implied volatility and the realized, or historical, volatility of the S&P 500 Index.

…The CBOE S&P 500 VARB-X ("Volatility ARBitrage") Strategy Benchmark tracks the performance of a simulated trading strategy that systematically sells Three-Month Volatility Futures* and holds the short position through expiration. The value of the benchmark is calculated from the profit or loss on the short futures position, plus the interest income derived from the available capital used to finance the portfolio.”

The CBOE has a strategy paper that spells out the details of this benchmark index and the underlying strategy. You can find more information at the CBOE VARB-X site. A graphic of the back-tested performance of this strategy over the past 32 months is below:

Reminder: VIX Options Expire on 3/21 This Month

This means that any positions put on today still have 5 days left until expiration -- long enough to give most fade the spike strategies a chance to benefit from a short-term mean-reverting subsidence in volatility.

In Case You Missed It…

A periodic look at ideas of note from other blogs:

• Ticker Sense with thoughts on winning strategies for when the S&P500 falls 2%
  
  
• Brett Steenbarger on using NYSE TICK volume to track the sentiment of large traders
  
  
• Profiling Leadership takes a look at how quickly sentiment turned around
  
  
• Mark Hulbert on the usefulness of put to call ratios (I think they have more predictive value than he does)
  
  
• Headline Charts with a reminder to watch corporate bonds for tells
  
  
• Adam Warner at Daily Options Report on what the divergence between the VIX and the RVX might foretell
  
  
• Roger Nusbaum at Random Roger states the obvious that “volatility is back,” but when Roger talks about volatility, we all should pay attention

Fade the Echo Volatility Spike

VIX currently at 18.06, +29% for the day.

Looks like some echo volatility and another opportunity to fade the spike.

Streaking

History may eventually realize that disco and streaking rose to extreme popularity at the almost the exact same moment in time. I don’t know if anyone else has attempted to link the two together, but I’m not one to let the short-sightedness of others stand in my way…

In spite of my efforts here, 1974 was a year to forget. In one stretch from January to September, the S&P500 fell 46%. Of course, Watergate and President Nixon’s resignation stole most of the headlines, but the OPEC oil embargo also sent shock waves throughout the US and the world. As if in sympathy, 1974 saw Babe Ruth’s all-time home run record of 714 fall to Hank Aaron and the UCLA basketball team finally lose after a record 88 consecutive wins.

After selling stocks, cursing the government, waiting in gas lines, and shaking their heads at the downfall of their iconic sports heroes, people looked anywhere they could for inspiration – and it came in strange places. Evel Knievel tried to jump the Snake River Canyon – and failed miserably. Muhammad Ali went to Zaire to fight George Foreman in the “Rumble in the Jungle” – and succeeded. Heiress Patty Hearst got kidnapped by the Symbionese Liberation Army, robbed a bank, and…well, things get a little fuzzy thereafter, but it was a nice distraction at the time.

So what about streaking and disco? Well, how the hell else were we supposed to distract ourselves from all that strange crap going on in the world?

Speaking of streaking (finally, he gets to the transition), the VIX has been down five days in a row. Does that tell us anything? Not much, really.

Here are some (potentially) interesting factoids about VIX streaks:

• Five consecutive VIX down days happen about 1.43% of the time versus about 1.15% of the time for five consecutive VIX up days (24% less frequently)
• 53% of all five days down streaks result in a sixth down day; VIX streaks up get extended to a sixth day only 38% of the time
• the VIX has never (since 1990) been up on eight consecutive days
• the VIX has been down ten days in a row once (0.02%), nine days in a row five times (0.12%), and eight days in a row six times (0.14%)
• VIX streaks of 3, 4, 5, 6, 7, 8, 9 and 10 days each occur less frequently than would be expected from random probability – again suggesting mean reversion at work

The VDAX and the VIX in the Wake of 2/27

When I introduced the VDAX in this space back on 2/22, volatility was hibernating with the bears on each and every continent and my commentary focused on the high degree of correlation between the VIX and the VDAX. I noted that the two indices often trade in tandem, but pointed out that the VDAX sometimes lags the VIX by one trading day or even two.

With the surge in volatility on the heels of 2/27, this seems like a good time to revisit some of those ideas.

Looking at the data for the four months leading up to 2/27, the difference between the VDAX (VDAX-NEW) and the VIX as a percentage of the VDAX ranged between 17% and 36%, with a mean of about 26% (plotted below as a dashed gray line.)

On February 27, the German markets closed well before the US markets; by the time the VIX had closed for the day, it was trading 5% higher than the VDAX. Over the next week or so, you can see where the VIX and VDAX played lead-lag cat and mouse, as global players tried to place bets ahead of any signs of increasing or lessening international contagion. Only in the last several days, has the VDAX-VIX spread ratio settled into a relatively narrow trading range, which I would interpret as a sign that the major players in the volatility markets believe that the probability of increased volatility or contagion is starting to subside. Whether these traders can accurately help predict the future remains to be seen, but when they stop playing the intermarket volatility game, you should at least incorporate that information into your market outlook.

Portfolio A1 Update for 3/11/07

As the graphic below indicates, Portfolio A1 is currently trailing the benchmark SPX return by 1.5%, in large part due to the performance of PCCC, which is down 14.6% since it was purchased as part of the initial group of five holdings on 2/20. In spite of the poor performance, PCCC continues to be the top rated stock in our portfolio, though it is likely that without some near-term buying support, the RSI component of our stock ranker will force a sale in the next week or two.

It should be noted that the one stock from the original group of five that has been sold, RIO, did bounce back 6.9% in the past week. The stock that replaced it, RKT, was up 3.9% in its first week in the portfolio. There are no changes to the portfolio this week.

The equity curve, which is starting to look suspiciously like a duck to me, continues to show a high beta performance.

Current portfolio details are as follows:

VWSI Snaps Back to Neutral

If you wanted to find an example in your investment textbook to illustrate “VIX Mean Reversion,” last week would be a classic case study. After soaring 75% in the week ended March 2^nd, the VIX snapped back 24% this week, so that it now sits comfortably between its 10 and 20 day SMAs.

Perhaps surprising to some, this 24% drop is still less of a move than the VIX drops in each of the three previous instances in which the VIX Weekly Sentiment Indicator (VWSI) was at -10, as I discussed last week. It is enough of a move, however, for the current week’s VWSI to return to a neutral reading of zero. This roughly translates as the odds times the magnitude of another VIX spike are about the same as the odds times the magnitude of continued subsidence over the coming week. Said another way, there is a higher probability that the VIX will be lower on March 16, but if the VIX makes a significant move in the coming week, that move is more likely to be up than down.

In the event that you still have open positions resulting from a post-2/27 fade the spike strategy, this looks like a good time to start taking those profits. Keep in mind, however, that the VIX has a tendency to drop during options expiration. With the 10 day SMA of total put to call ratio currently sitting at 1.28, you can safely assume that it will need a Sherpa to stay at that altitude, suggesting that the market is poised to move upward and the VIX may have another leg down this week.

(Note that in the above temperature gauge, the "bullish" and "bearish" labels apply to the VIX, not to the broader markets, which are usually negatively correlated with the VIX.)

Canary or Canard?

Earlier in the week, I introduced a ratio chart of the percentage of S&P500 stocks above their 50 day SMA divided by the VIX. I described this as an attempt to find a rough approximation of greed ÷ fear. Since this chart received some favorable reviews, let me unveil another VIX ratio chart that attempts identify the same market sentiment extremes and provide a warning about the increased probability of near-term market highs and lows.

The chart below is a weekly chart of the ratio of the new 52 week highs in the NYSE divided by the VIX. I focus primarily on the raw number and the 4 week SMA. You can see from the chart that raw reading above 30 and a 4 week SMA readings above 20 tend to signal that a top is near.

In fact, this chart provided excellent advance notice of the 2/27/07 top, the 5/11/06 top, and previous tops in March 2004 and March 2005. With bottoms, the record is not quite as good, with an excellent advance call of the October 2005 low, but calls that were too early for May-July 2006, March-August 2004, and July 2002-March 2003.

You can decide if this ratio chart deserves a spot in your toolbox or bird cage or whatever it is that you use to try to divine the future. For me, it’s a keeper.

CBOE Press Release: Volatility Index (Vix) Options Trade 230,847 Contracts, Setting New All-Time Record Today, Thursday, March 8, 2007

Chicago, March 8, 2007 - The Chicago Board Options Exchange (CBOE) announced that reported volume in options on the CBOE Volatility Index, "VIX" (ticker symbol VIX), was 230,847 contracts today beating the previous record of 162,856 contracts traded on August 14, 2006. VIX is the widely disseminated benchmark index of market volatility and investor sentiment, sometimes referred to as the market's "fear gauge."

Today's trading activity in VIX options was spread across strike prices and months, with heaviest volume in March 18, March 20 and May 14 calls. Open interest across all VIX options stood at 950,164 contracts (706,949 calls and 243,215 puts) at the start of trading this morning. VIX options trade on the February expiration cycle with March, April, May, August, September and November 2007, and February and May 2008 expirations currently available. VIX closed today at 14.29, with a high of 14.70 and a low of 13.48.

VIX had recently been hovering near the lower end of its range, until a surge in volatility during the week of February 26, 2007. So far in 2007, the lowest VIX reading, of 9.7, was recorded on February 14, and the highest level, of 19.01, was on February 27, 2007. The VIX price is calculated and disseminated by the CBOE throughout the trading day, and historical values can be found on the CBOE website at http://www.cboe.com/. A table with yearly high and low closing prices is included at the end of this release.

Launched just one year ago, on February 24, 2006, VIX options have been one of the most successful new products launched at CBOE.Average daily volume for VIX options during February, 2007 was 45,605 contracts. For all of 2006, average daily volume was 28,050 contracts. VIX futures also are available on the all-electronic CBOE Futures Exchange (CFE).

CBOE has created a new section of its award-winning website, offering education on VIX trading strategies, historical price data on the VIX going back to 1986, charts, answers to frequently-asked questions, and a free options toolbox, at: http://www.cboe.com/VIXOptions. The CBOE Futures Exchange recently launched a monthly newsletter, "Futures in Volatility," that focuses on volatility and volatility trading strategies. The newsletter is free of charge.

CBOE, the largest options marketplace in the U.S. and the creator of listed options, is regulated by the Securities and Exchange Commission (SEC). For additional information about the CBOE and its products, visit the CBOE website at: http://www.cboe.com/.

The Credit Default Swap Canary

Raise your hand if you were following were watching the Bombay Sensex for clues about the future of the Chinese market or were wise to the unwinding of the Yen carry trade before it happened. Maybe you were tipped off to the subprime mortgage debacle. If so, you did a better job than most investors.

I won’t say that keeping a weather eye on the VIX would have guaranteed that you found yourself on the right side of the markets on 2/27, but it would have helped. Careful study of the equity put to call ratio would also probably have helped.

Let me nominate another canary to add to the investment coal mine: credit default swaps. I know, I’m not a bond guy either, but keeping an eye on how the markets are pricing default risk (the spread versus the US Treasury yield curve) is somewhat of a bond analog for how equity risk is priced in with the VIX.

I highly recommend watching the credit default swap index for high yield corporate bonds. A snapshot of this index is available at Yahoo, but much more information is available through Markit (click on the Dow Jones CDX.NA.HY link to pull up the graph below). Note that in the high yield graph, the market discounted 1/3 of the risk premium from September to Feburary and is now going through wild gyrations in an effort to re-price the risk premium to better match current expectations.

I should also note that while StockCharts.com does not provide charts for the high yield credit default index, they do offer charts for a sister index, the investment grade credit default index, whose weekly chart I have included below. The IG index chart suggests that the recent turmoil in the bond markets is less than what transpired during the May-July sell-off of last year and does little to reverse the 20 month trend of increasingly narrowing credit default spreads.

The bottom line is that it is always good to have a canary that you watch closely to help call market tops and bottoms. It can’t hurt to have several canaries, each with their own unique sensitivities.

Elast-o-VIX

The events of last week gave us an opportunity to examine what happens when the VIX spikes 30% or more in one day. For the most part, the conclusion was that spikes of this magnitude generally represent a near-term VIX top; save some echo volatility spikes, they also indicate that the VIX should trend down from a large spike.

We are in uncharted territory once again this week, this time with the mean-reversion engine working overtime to snap the VIX back down from the low 20s to 15. In an unprecedented move, the VIX fell over 15% twice within five trading days, dropping 16% on 2/28 and 19% yesterday. Interestingly, the last time two 15% drops fell in reasonably close proximity was in June 2006, when they were ten days apart. Prior to that, there were only nine 15% drops in the previous 17 years and only two of those occurred in the decade leading up to the events of last June.

At one point yesterday, the VIX was down 20%. While it edged away from that number by the end of the trading day, I thought it might be instructive to look at the only three times in VIX history that it has fallen 20% in one day: 9/22/93; 4/15/94; and 6/15/06.

For those who are not familiar with the composite historical graphs I have included here in the past, the chart below takes the three data points above and normalizes the aggregate end of day values at 100 for the day in which the VIX fell 20% (also indicated with a “0” on the X axis and a vertical dashed line through it.) Here I have included the aggregate pricing data for the five days leading up to the 20% VIX drop (-5 to -1) and the twenty days immediately following the drop (1 to 20.)

The graph itself should tell the rest of the story. Personally, I found it most interesting that the 20% VIX drop was – at least in this historical context – a mean-reverting reaction that reversed three days of increasing volatility rather than striking out in a new direction. Interesting perhaps, but not surprising, as it is consistent with one of the recurring themes in this blog.

Are We Done Yet?

The answer, of course, is that nobody knows. For the most part, panic and fear are not fleeting emotions.

For the curious, the photo on the left is of Colca Canyon, located in southern Peru. The canyon is twice as deep as its more famous relative, the Grand Canyon. What I find particularly interesting about Colca Canyon is that despite a vertical drop of over 10,000 feet, much of the canyon is terraced and relatively flat.

In thinking about the current investment topography, one of the more obscure charts that I refer to from time to time is a ratio chart of the percentage of S&P500 stocks above their 50 day SMA divided by the VIX. It is a rough approximation of greed ÷ fear and prints some interesting extreme numbers from time to time, particularly for market bottoms. Because the values of this ratio chart are one ratio divided by another, it is possible to compare values over time and get meaningful results.

In looking at the weekly ratio chart below, one of the first things that jumps out is the toppy readings (6.0 – 8.0) that persisted over the past six months or so, as compared to the previous high readings, which tended to last only a month or two.

The downside spikes are even more pronounced and tend to be much more short-lived. We know that the moves in the VIX over the past week have been unprecedented, but what the chart indicates is that given the steepness of the decline and the current low reading of 1.55, the ratio is not likely to deteriorate much further, if at all. Don’t expect the ratio to stay in the Colca Canyon area – or even below 3 – for much longer.

I am certainly not going to declare that this ‘correction’ is over, but I will go out on a limb to say that the worst of it is likely behind us and that further declines, if any, should be a lot more orderly.

Long Volatility with the Exchanges

In the rush to capitalize on the sudden volatility in the markets, many have jumped in to buy VIX calls not fully realizing what they were getting in to. Adam Warner summarizes Barron’s take on some of the shortcomings of VIX options as an investment vehicle and reiterates his own perspective, which is essentially that the more sensible and direct play is SPY options. For the record, I am in agreement here and have stated some other reasons why VIX options may not be appropriate for all by the most experienced options traders.

Here is another way to play volatility without even resorting to options or ETFs: the exchanges. CME just reported record options volume in February and will likely continue to be a beneficiary as concerns about increased risk and volatility create new demand for derivatives to reduce risk and volatility going forward. This morning, CME is down more than 50 points from the January high of 596.

Another options exchange play is ISE. This smaller electronic options exchange had a strong February, but has more question marks than its larger rivals. ISE is down about 12% since February 23^rd and may be an attractive takeover target.

Portfolio A1 Update for 03/04/07

When I first announced that I would be using this space to talk about a live portfolio I was operating, I was more than a little concerned that the timing might not be ideal and that I should wait for a possible market correction before I rolled out a new portfolio. I based this opinion on my experience that the predecessor portfolios to A1 had demonstrated a greater propensity for outperforming the SPX benchmark in up markets than in down markets. Perhaps I should have listened to my gut…

Well, whether we have that correction already in hand or are in the early stages of a more substantial bear market, it is time to drill down on the A1 portfolio. As you can see from the attached equity curve below, A1’s performance has slipped below that of the SPX as a result of a week in which the five holdings were battered more severely than the indices. Due to last week’s performance, the system’s ranking of RIO (a favorite holding over the past two years or so) has dropped, triggering a sale. RIO has been replaced by RKT, a packaging company whose stock has tripled over the past year. RKT sports a relatively modest P/E 16 for 2007 earnings and a 1.9 PEG that puts it well below the industry average. In addition to a revenue growth story that is in sharp contrast to the industry trend, RKT has demonstrated continued pricing power it its markets. This is not a sexy stock, but the company appears to be executing on all cylinders and is attractively valued.

A glance at the equity curve suggests that this portfolio has a higher beta than the SPX; while this is case at the moment, largely due to the volatile AMKR, swapping RIO for RKT should decrease overall portfolio beta.

For the record, PCCC and AMKR are currently rated as the #1 and #2 stocks in this portfolio. In this challenging market environment, the performance of these two stocks should set the tone for the early performance of this portfolio and dictate the size of any drawdowns which may need to be scaled to return the portfolio to the green.

Finally, I should probably have already explained that this portfolio is a long only equity portfolio. It is not allowed to short stocks, buy or sell options, or avail itself of ETFs of any kind. As a result, the system does not make an effort to hedge any positions, regardless of the market conditions. The coming week should have a lot to say about how well this strategy is suited for the current market environment.