Predicting an ISEE September Buy Signal

Among the many indicators that have been giving contrarian bullish readings as of late is the ISEE, which narrowly missed making a new record low for its 20 day SMA earlier this week. Long time readers (can I already have any of these after only 8 months?) may recall that when it comes to the ISEE, I have a preference for using the 50 day SMA and for using absolute readings as well as movement away from well-defined tops and bottoms for the best trading signals.

In March I anticipated an upcoming buy signal from the ISEE 50 day SMA in April and suggested that it was a good time to “get long, perhaps in a big way.” Thanks in part to the predictability generated by older numbers rolling out of the SMA calculations, I was also able to see a double bottom coming in April.

With a month of august volatility now in the books, it is becoming clear that September will also likely trigger an ISEE buy signal. My best guess right now is that the buy signal will become official during the second or third week of the month. Given that this a very high probability signal, I see no reason why not to flag it now and look to grab 2-3 extra weeks of upside.

As an aside, consider how surprising it is that given all the recent turmoil, the broad indices are up over 1% going into this three day weekend, with investors large and small apparently more concerned about missing a bull leg than seeing subprime headlines and red numbers on their screen Tuesday morning. We may be turning a corner…

Echo Volatility, Day 10

Following the 2/27 VIX spike, I had a lot to say about echo volatility in this space.

My dog explains the concept best, but a good working definition of echo volatility is along the lines of “the tendency for markets to be more susceptible to volatility spikes in the wake of an initial volatility spike.” Some historical context is available at “VIX Spikes and Echo Volatility” and I looked back at post-2/27 echo volatility in “Lessons from the Post-2/27 VIX Price Action.”

Finally, in response to a reader question, I presented an important element of my thinking regarding echo volatility in “When Is Echo Volatility Safely Behind Us?” Here I offered two key takeaways:

1. The first ten days tell you very little about future VIX prices that you don’t already know by just applying ‘normal’ VIX mean reversion tools to the close on the day of the spike. By day 15, it is possible to predict future VIX price levels [30-60 trading days out] with considerably more accuracy and by day 20 I would say that I am ‘as comfortable as possible’ about making predictions about future VIX levels and the possibility of more echo volatility.
   
   
2. One of the golden rules of VIX mean reversion is that if mean reversion does not play out over the course of 10-20 days, then we are likely headed for a period of extended volatility.

So, here we are on the tenth trading day after the August 16^th 37.50 VIX spike top, with a very small echo volatility bounce that took the VIX from 20.44 to as high 26.67 in two days, still almost 30% below the VIX spike top. It is too early to discount the possibility of an another echo volatility VIX spike in the next week or two, but by the end of next week and particularly by the end of the 20 day window on September 14^th, the window of opportunity should be closed and we should be able to put the 37.50 spike and any subsequent volatility spikes to bed.

Watch XBD’s Implied Volatility

With the DJIA up almost 100 points right out of the gate, I was curious to see GS and BSC quickly fade from green to red – and that weakness reflected in the XBD (Broker/Dealer Index.)

Of course, you probably don’t have to hedge the entire market too catastrophe-proof your portfolio these days. You can probably accomplish the same task by erecting a safety net under just one or two sectors, such as the home builders or financial institutions. So I looked at my favorite bellwether, Goldman Sachs, to see how their implied volatility has fared in the past month or so. While it makes for an interesting visual, I have not included the Goldman chart because the company has a history of slipping punches. A better chart is the XBD, whose components include 12 companies in the broker-dealer space.

The iVolatility chart below shows implied and historical volatility for the XBD going back three months. Prior to July, the XBD IV spent 95% of the past year in a narrow 20-25 range. After topping 50 in mid-August, the XBD IV looked to be headed back down to 30 or so, until the recent spike left it over 40 yesterday.

While it is important to watch the price level of this index to see how it holds up at support levels such as 215 and 208, I also recommend keeping a weather eye on the XBD’s implied volatility to see what the trend and absolute levels of IV tell us about the thinking of options players. It is quite likely that the tip of the next iceberg may be found floating in the IV chart before it shows up on a price chart.

The Relationship Between Volatility and Market Returns

Hans Wagner (not pictured) has an article with the title “Volatility as a Stock Market Indicator” up on Financial Sense. Posted yesterday, the Wagner article examines the relationship between volatility and market performance from a monthly and yearly perspective. In his analysis, Wagner leans heavily on the research of Ed Easterling of Crestmont Research, whose relevant book, Unexpected Returns: Understanding Secular Stock Market Cycles, I confess not to have read.

Using S&P 500 index data going back to 1962, Easterling notes that when the average daily range (in percentage terms) of the S&P 500 is lowest, it corresponds to a higher likelihood of monthly gains in the index. So…calmer waters make for easier sailing.

A quick and dirty way to monitor these types of opportunities is with a 20 day Average True Range or 20 day Bollinger Band width indicator. As you can see from the charts below, both ATR and BB width provide an excellent means by which to evaluate market volatility graphically, with fairly reliable and easily measurable signals of volatility extremes, not unlike the information provided by the VIX.

Another Look at the VIX:SDS Ratio

Back on August 10^th, when the markets were testing the first set of lows, I toyed with several indicators that I thought might help me better separate fear from volatility. I published a 10-day chart of one of those, the VIX:SDS ratio.

I have been keeping an eye on this ratio during the past 2 ½ weeks and noticed that the extreme reading of .633 it did an excellent job of flagging the recent market bottom. I am still not sure how useful the VIX:SDS ratio may be going forward, but I thought a six month chart might be interesting analytical fodder for those who like to contemplate such matters. As always, comments are welcome.

As a quick reminder, SDS is an ETF that is intended to track at 2x the inverse of the SPX. More information is available from ProShares.

Portfolio A1 Appears to Find a Bottom

After four weeks of performance that you wouldn’t want to be downwind of, it seems appropriate that it took a new addition to the portfolio – and a fertilizer company at that – to turn around Portfolio A1. The fertilizer company, Mosaic (MOS), surged 14% last week, while the other portfolio newcomer, agricultural equipment maker CNH Global (CNH) posted a weekly return of 5.7%. For now at least, the agricultural theme is working.

The aggregate portfolio statistics are still on the ugly side, with the portfolio down 12% since the February 16^th inception, well behind the benchmark S&P 500’s 1.6% gain during the same period. With a Sharpe ratio of -0.72, a winning percentage of 36%, and a maximum drawdown of 29.9%, one has to look long and hard to find a silver lining in the portfolio’s performance. Still, I will ride this portfolio out through the end of the year, at which time I will introduce a new portfolio with a strong discretionary component, as I outlined last week.

This week Portfolio A1 swaps BRIC telecoms by saying goodbye to Moscow-based Mobile TeleSystems OJSC (MBT) and replacing it with Brasil Telecom Participacoes (BRP), an ADR that the discretionary trader in me likes a great deal. There are no other changes to the portfolio this week.

A snapshot of the portfolio is as follows:

Historic Weekly Drop in VIX; VWSI at +2

Last week I toned down my prose a little by noting “we have almost surely seen a top in volatility” and declaring “there is finally evidence that the laws of gravity have been reinstated, at least temporarily.” At least I didn’t hold back in my forward-looking commentary:

“I suspect most pundits will wait until after Labor Day before declaring that it is safe to go back in the water, but I tend to think that excessive caution will likely result in missing out on the first big leg of a post-panic bounce.”

One week does not mean it is safe to announce the resumption of the bull market, but clearly quite a few big players were caught leaning the wrong way as stocks moved up. Starting with the VIX, we saw a record drop of 30.9% from 29.99 to 20.72. This eclipsed the previous record weekly VIX drop of 28.4%, which came in March 1991, just after the end of the Gulf War. In retrospect, this also turned out to be an excellent time to go long the broad markets and short volatility.

The record drop in the VIX has pushed the volatility indicator into a deeply oversold level on a short-term basis, while it continues to remain overbought on a long-term basis. Check out Ron Sen’s blog, Technically Speaking, for a chart of the VIX relative to its 10 and 200 day SMAs. How does the VWSI sort out this divergence? By siding somewhat with the extremes of the short-term VIX readings, where a reading of 31.6% below the 10 day SMA has only happened three times previously (1/91, 3/91 and 12/98.) The bottom line is a VWSI of +2, which is slightly bullish for the VIX, but technically still neutral enough as to preclude a formal directional recommendation.

In the event I fail to blog about this tomorrow, it is also worth noting that the ISEE is flagging record levels of skepticism as well. Readings for this sentiment indicator have been extremely low for the month of August; and an ISEE of 95 of lower tomorrow will mean a new record in the 20 day SMA for this indicator – and contrarian bullish support for a continued bounce in the markets.

It’s no fun being just another cinder block in the wall of worry…

(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: For the second week in a row, I find myself happily recommending a classic Spanish varietal. This time around it is albarino, also known as alvarinho in Portugal. Albarino is an excellent summer wine, with a zingy citrus delivery that will remind some of the best of Oregon’s pinot gris. Don’t limit albarino to just one season, though, as it has enough weight and body to be an excellent year round wine, with the high acid content making it an excellent food match. Albarino thrives in Rias Baixas region in northwest Spain. If you find yourself in a chardonnay rut – or even if you don’t – sample whatever your local wine store has to offer. My local wine store has a Burgans Albarino for $10. I am a big fan, as are the folks at winefoolery and elsewhere.

Finally, if you want to have some fun and learn about albarino at the same time, I recommend you watch the upstart Gary Vaynerchuk at Wine Library TV taste four albarinos. (If you really want some over the top wine-related entertainment, watch Gary teach Conan O’Brien how to train his palate.)

Drilling Down on Sector Performance

Yesterday I talked about sectors in the context of the nine AMEX Select Sector SPDRs. While these are excellent high level buckets for analyzing macro sector performance, when you lift the hood on these SPDRs (click on any one for details), you jump down to the individual stock level, without the benefit of sub-sectors to analyze.

Fortunately, there are many other excellent free resources where one can drill down on sector performance. Four great places to start are:

• ETFScreen.com sector performance and RS – all sectors, sortable from 1 day to 1 year
• ClearStation’s sectors and industries – includes relative performance data as well as new high/low numbers (click on the number to see the companies making new highs/lows in that sector)
• Marketwatch 10 best/worst performing industries – sorted by % change from 1 week to 5 years (BigCharts has the same data)
• ADR.com’s top sectors – for an important global sector perspective

I should probably devote an entire post to Prophet.net, which does many interesting things with sectors. The tools I find of particular value are sortable performance for 214 sectors from 2 days to 5 years; and historical sector ranks from 3 months to 5 years in a helpful graphical format (see below), complete with a drill down capability that pops up the charts for all the individual stocks in a particular sector. Also, from the I-just-couldn’t-help-myself category, while the pull down menu only allows for a minimum historical performance of 3 months, if you manually edit the URL, you can produce some interesting charts for shorter time frames. For example, where the URL for the 3 month graphic ends in “…period=3m” it can be edited to “...period=1m” to generate a particularly interesting one month historical chart. Try it!

MarketGauge.com has an industry group summary that is an excellent graphical tool covering multiple time frames, but also offers four fundamental analysis options for analyzing the top and bottom sectors. I particularly like a feature they have that highlights the stocks driving the strongest/weakest groups higher and a perhaps even more valuable leading stocks in today’s top groups page that includes fundamental data and charts on one handy page.

For a different take on sector performance and momentum, you might want to try ETFInvestmentOutlook.com. Two of their features that I get the most use out of are the McClellan breadth ETF rankings and the high-low breadth ETF rankings.

In addition to the above, there are a number of interesting sector-related heat maps available, including two sites of particular note:

• SmartMoney heatmaps – sector map, map of the market, mutual fund map, etc.
• NASDAQ – NASDAQ-100 Heatmap, NASDAQ-100 Pre-Market Heatmap, ETF Heatmap, and a custom heatmap you can build yourself

Finally, in the event that you have not been there in awhile, Yahoo has beefed up their Industry Center a little. A good place to start surfing there is in the leaders and laggards section.

What’s Working? A Sector Overview

Now that the broad indices have pulled back 10% or so and retraced about half of that drop, it is a good time to evaluate what is working in the current market environment.

One of the first places I look for clues is in sector and industry performance. With ETFs it is now possible to take the temperature of just about any micro-segment of the market you can think of, but for today I will stick to broader market segments.

Looking at the peak to trough from July 19 to August 16, you can see that it is the materials sector (XLB) that suffered the most dramatic losses, dragged down by WY, IP, AA and the like.

In the five days since the market has bounced, the materials have been leading the bullish charge, with energy and financials lagging. Given the change in expectations about interest rates, it is not surprising to also see that utilities have had a much better week relative to the past month than the other sectors.

Going forward, look at the relative performance of materials, industrials (XLI) and technology (XLK) to provide some clues about global economic conditions and consumer discretionary (XLY) vs. consumer staples (XLP) to tell us something about the health of the consumer. If energy (XLE) and financials (XLF) can also start to rally, this may help us to discern the difference between a short-term bounce and the resumption of the bull market.

Which Gravity?

In my seemingly never ending quest to litter this space with obscure and occasionally relevant VIX trivia, I have today come across some numbers that I find particularly interesting.

First, let me point out that the VWSI is currently back to reading an even zero, with neither a bullish nor bearish bias. Part of the reason for this neutral reading is the current deadlock in the gravitational tug of war between short-term and long-term mean reversion. Not only that, but in the 17 year history of the VIX, the current 15% under the 10 day SMA and 43% over the 100 day SMA is the largest ever divergence between these two indicators. The question, of course, is whether the gravitational pull of the 10 day SMA (not pictured) will win out over that of the 100 day SMA – or even whether one mean reversion magnet will get the upper hand going forward.

For market historians, there are two instances of possible historical precedent which may be of interest.

In the end of July 2002, we had the largest previous divergence, with the VIX 17% under the 10 day SMA and 32% over the 100 day SMA. This set of circumstances followed the WorldCom bankruptcy filing and came close to signaling the bottom of the 2000-2002 bear market. In fact, in the days leading up to this divergence, the VIX fluctuated wildly to a peak of 48, then dropped to 31 just three days later. Within a week following the maximum divergence, the VIX was back over 45 again; and it remained elevated over the course of the next two months as the markets finally confirmed a bottom.

There is some similar historical precedent in the wake of 9/11, during which period the VIX hit 49, then dropped to 31 five days later, resulting in a VIX 16% under the 10 day SMA and 33% over the 100 day SMA. What followed was a temporary market bottom and VIX readings that went sideways for about five weeks, then began to subside for about nine months, before the July 2002 craziness noted above kicked in.

I am not sure what to conclude, if anything, about the historic divergence at present, other than it is the almost inevitable residue of an unprecedented VIX spike. In a few weeks we will all know whether the liquidity/credit crunch has swallowed up one or more of our trusted financial institutions or, as is usually the case, if investor fears just got a little too far ahead of the reality.

VIX:VXN Ratio Extremes

Earlier this morning, Adam Warner at Daily Options Report posted a chart and commentary about the VIX:VXN ratio, which volatility groupies will recall compares the implied volatility of the S&P 500 to that of the NASDAQ 100.

The interesting factoid is that the VIX:VXN ratio is at an all-time high, which the chart below highlights (while the VXN was launched by the CBOE in early 2001, StockCharts.com only has VXN data back to February 2003.)

The key question is why the VIX:VXN ratio is printing such extreme numbers at the present. Adam concludes the following:

“Best guess is that there's a perception out there that tech is relatively *safe* now. And I suppose it is given that it's not the focus of the periodic poundage.”

This take makes a lot of sense, as various ratios of technology stocks to financials (e.g., XLK:XLF and XCI:XBD) reflect that the current environment is one of those rare instances where technology is considered less risky than financials.

A look at two SPDRs tells the story even better, in my opinion. XLF, the financial sector SPDR (XLF top holdings), shows a 160% spike in IV from mid-June, while XLK, the technology sector SPDR (XLK top holdings), shows an IV spike of about 120%. Even more interesting, at least to my eye, is that following the February 27^th VIX spike, XLK retraced all of its IV spike, while XLF only retraced half of that spike before starting to rise in mid-June. Was the half-hearted XLF implied volatility spike retracement in March through June a warning of what was lurking under the surface? For those who may be interested, of the nine AMEX Select Sector SPDRs, four of the sectors – financials, consumer discretionary (XLY), industrial (XLI), and utilities (XLU) – did not retrace their February IV spike; and these sectors do not correlate with relative sector performance over the past month.

Can the VIX See a Year Out?

Even before Adam Warner and I were humbled in our attempt to forecast the VIX one month out, I have been slow to warm up to the predictive value of the VIX looking out more than a month or two – all of which makes Eric Boughton’s “The Predictive Value of the VIX: Room for Divergent Opinions” particularly interesting reading.

First, the bad news: Boughton concludes that “knowing what the VIX is today is not likely to give you any aggregate useful information about whether returns will be high or low over the next year.” The good news is that the VIX does a reasonably good job of predicting future volatility over the course of the next 12 months.

While Boughton’s target time frame of one year may not be the best way to test the predictive value of the VIX, the tidbit I found most interesting was tucked away at the bottom of Boughton’s article, in which he notes that “buying the S&P on every day the VIX exceeded 30, and holding it for a year, resulted in an average return of 17.15% (as compared to the average return of 10.28% available for all twelve-month periods during the period in question).” His conclusion? Even if there is no correlation between the VIX and one year returns, “the market seems to pay an outsized return in exchange for the risk of buying when fear is high.”

My thinking continues to be that middling values in the VIX are generally not worth talking about. In the end, it is VIX spikes and mean reversion that matter most – and those factors are easiest to predict over the course of about 5-10 trading days.

Ticker Sense on VIX Spikes

This morning Ticker Sense published an analysis of the four previous instances where the VIX has risen more than 100% in a three month period.

Among their findings is that the greater the drop in the SPX during this period, the greater the rebound in this index during the following month, as shown in the following chart, which is courtesy of Birninyi Associates / Ticker Sense:

An even more interesting finding – and one consistent with the capitulation theory – is that in each of the five instances (assuming last week’s VIX spike will turn out to be the top this time around), the largest move in the VIX came during the last 20% of the VIX run-up period.

Portfolio A1 Looks to Ag to Stop the Bleeding

The big story for Portfolio A1 is the recent drawdown, which is now registering a peak to trough drop of 29.7%, approximately triple that of the S&P 500 index benchmark. Since its inception (2/16/07), Portfolio A1 is now down 18.2%, vs. a 0.7% drop for the S&P 500.

Clearly, this has not been the place to have your money during the liquidity crisis – and the performance of this portfolio has given me a fair amount to think about. First of all, I usually have target drawdowns beyond which a portfolio or trading system automatically gets consigned to the trash heap. As I am an aggressive investor, the drawdown threshold usually falls in the 30-40% range. With Portfolio A1 now hugging the 30% drawdown line, I am tempted to shut it down. The question, however, becomes what to replace it with or whether to even bother with this portfolio feature on my blog. Here is my thinking at the moment: I will keep Portfolio A1 up and running until 1/1/08 or it reaches a 40% drawdown, whichever comes first.

Starting 1/1/08, I will unveil a new portfolio that is a hybrid between an automated system like Portfolio A1 and a discretionary system. As a result, it will incorporate more of my contemporaneous thinking about the markets and about individual stocks. It should also be a portfolio for which I feel greater ownership and accountability.

I toyed with a bunch of other ideas about what to do with this space on the weekend, but I kept coming back to my desire to highlight a specific portfolio and specific stocks. I look forward to the transition and hope that in the remaining 4 1/3 months, Portfolio A1 can regain some respectability.

Note that Portfolio A1 has finally decided to drop former high fliers Terex (TEX) and Southern Copper (PCU), replacing them with two companies with a strong agriculture component: Mosaic (MOS), a fertilizer company; and CNH Global (CNH) a Dutch manufacturer of agricultural and construction equipment.

A snapshot of the portfolio is as follows:

Normalcy Again? VWSI at -4

In spite of Warren Harding’s calls for a return to “normalcy” back in 1920, Americans – investors or otherwise – ended up with a decade that was anything but normal.

With the VIX a hair under the magic 30 barrier and Hurricane Dean taking aim at the Yucatan instead of the Texas refineries, we have almost surely seen a top in volatility. That being said, the VIX managed to close up on a weekly basis for the fifth time in the past six weeks, finishing the week at 29.99, up 6% (1.69) from the previous week. Is normalcy really just around the corner? My guess is that it is not too far away. The new high water mark in the VIX is now 37.50, which is the highest reading since October 2002, when the markets were just beginning to bottom, before turning up for the current five year bull run.

The VWSI continues in uncharted waters at -4 this week, but with the VIX already 25% off of the 37.50 spike top, there is finally evidence that the laws of gravity have been reinstated, at least temporarily. As the graphic below shows, the VWSI is now mildly bearish on the VIX, with mildly bullish implications for the overall markets. If you consider the strong possibility that the markets can no longer dwell on options expiration, a Fed with it’s head in the sand, or Hurricane Dean threatening the Texas-Louisiana oil infrastructure, then I suspect the coming week will more likely bring bullish surprises than bearish ones, especially given that it is a relatively slow week on the government data front.

I also suspect most pundits will wait until after Labor Day before declaring that it is safe to go back in the water, but I tend to think that excessive caution will likely result in missing out on the first big leg of a post-panic bounce.

(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 VWSI of -4 comes along only a couple of times each year, so it is a good excuse to get off of the beaten varietal track. Also, it never hurts to diversify one’s portfolio internationally. With those thoughts in mind, I am recommending some tempranillo for a VWSI of -4. Tempranillo is the classic Spanish red grape that has long dominated the wine scene in Rioja and Ribera del Duero.

Americans looking to dive into tempranillo should recognize that the grape is not widely planted in the US and focus instead on Spain, which remains the center of world class tempranillo. For a good introduction to the grape and some affordable entry level recommendations, try Spain’s Early Ripener by Blake Gray in the San Francisco Chronicle. Considering that Spanish tempranillo is comparable in quality and stature to an Italian Super Tuscan, a French Bordeaux, or a American cabernet sauvignon, you should also be amenable to higher price points. Some excellent places to get ideas about premium tempranillos include the 2007 Shanghai Tempranillos al Mundo awards, the San Francisco Chronicle wine competition awards, and the tempranillo tasting notes on Cellar Tracker.

To read more about what is going on with this grape in Spain, I encourage you to check out Tempranillo, a blog about the Spanish wine industry.