Carpenter Analytix

Carpenter Market Remarks
              Notes on Recent Statistics and Inference

                   
www.CarpenterAnalytix.com
                                

                        Carpenter's Market Remarks. 

For referenced attachments, please request at Contact link below.

For current (2015) Market Remarks, click here.

For other Remarks archives see links at foot of page.

November 29, 2010. The attached image of Net Normalized NY Tails pretty much summarizes where things stand. Following the indicator's sharp retreat from April highs earlier this year, the market's Sep-Oct-Nov rally has surpassed April's highs but the Net Tails show considerably less vigor than the Feb-Mar-April rally. That sets up a substantial divergence, but it's incomplete until the Net Tails path falls below the summer's lows. If and when that Net Tails path posts lows or negatives, as it did in mid-2000 and mid-2007, it will be calling for some real selling. Meanwhile, it's a clear set-up for weakness, but not quite ripe.

Value-oriented mutual funds have begun lightening up. They're surprise timers, and pretty good on macro outlook. Our Gurus' exposures data are posted weekly at the Prime Timers page. Graphics updates are posted monthly...with next Friday (12/03) scheduled for December's graphics. .

Robin C.
Hanover NH

November 22, 2010. Our cross-sectional skew metrics continue to run (mostly) positive. I've attached a graphical summary of the past week's five daily returns distributions. (There's almost no description; ask if it's unclear.) Aside from last Monday, which had a returns distribution of remarkably negative skew (setting up Tuesday's sharp decline), the shape of the distributions are otherwise generally positive. Positive daily skew does not itself have enduring implications...except that history shows the skew usually remains positive (or negative) over somewhat lengthy periods.

Separately, I also note that our hedge fund exposures work shows CTA funds, which went net short a week ago, are now suddenly down into their bottom quintile of five year exposures. Bottom quintile is our general threshold of "too bearish," and therefore entering ripe zone for reversal. The data and graphics are presented in Hedge Fund Analytix. If you're considering subscribing to HFA , I'll be happy to send the current report so we can talk about how this contrarian analysis works.

Finally (and mildly contrary to above data), it's worth mentioning that among our Prime Timers, value funds this week finally showed some pull-back from their long held bullish position. Value funds are proven positive timers. A one-week pull-back from high exposures is not alarming, but will become worrisome if it continues. We post our Prime Timer exposures at the web site weekly (next update on November 26), with graphic updates monthly (next on December 3).

Robin C.
Hanover NH

November 15, 2010. The past week's pull-back has not generated any terrible weakness. Weaknesses noted in recent weeks continue relevant, but indicators last week, having been given a chance to show real near-term weakness, have not done so. The likely outcome is that the vigorous up-thrust of the previous week (w/e Nov 5) has not finished playing out momentum.

I've attached a picture of the NYSE Tail Skew. This indicator tracks the difference between the "stretch " in the upside tail and downside tail of daily returns distributions. What's salient here is that it remains positive (good), but plainly weaker than the price line of the past month or so (not good). Various other metrics show the same behavior.

Bottom line: I expect renewed advance, but the greater weakness (failure of our internal metrics to match new index highs...neither November's nor April's...is saying there's a greater down-turn brewing. That's dangerous.

Robin C.
Hanover NH

November 8, 2010. Upward momentum rules, for the moment. Growth indexes have bested their April highs; Value indexes are just behind, and likely to make their own highs shortly. A week ago I wrote here that the now rising Tails paths suggested increasing volatility and soon downturn. We're getting the volatility, but the downturn's been delayed.

With indexes at or above their April highs, I took a look at some Market Remarks passages back when the correction was just getting started. Here is what was said over the weekend just after the end of April.

May 3, 2010. There's a pretty good probability that recent minor weakness has more to come. Maybe a good bit. But the new highs posted by various cross-sectional analysis strongly suggest the larger-scale uptrend is not finished. It would be unusual (to say the least) for the master rally (from March 2009) to have ended with cross-sectional skew making new highs, as it did just seven trading days ago. I've attached here a graph of the NYSE Tail Skew. Note how the Skew peaked in 2007 a couple of months before the S&P peaked. We have no rule about any specific lead times, but there is every likelihood of seeing new market highs (without Skew participation) before the master rally is done.

We did indeed get "more to come" correction through the summer. Now we are just at the point of "...seeing new market highs (without Skew participation)" as expected. That all took longer to play out than imagined back in the spring, but the basic scenario is intact. I've attached the updated NYSE Tail Skew chart, and also the comparable Nasdaq Tail Skew. The red vertical on each chart simply marks the April-to-May transition point when the passage above was written. The key point is that market Skew was posting highs along with the market back in April, but is nowhere near matching those highs now.

So momentum is expected to play out for now, but weakness is growing. Unless the internal stats improve a great deal to support the momentum, next downturn is likely to be more damaging than the last.

Robin C.
Hanover NH

November 5 , 2010. Thursday's run-up was remarkable, and can be expected to have follow-through momentum. But the concluding remark from last weekend's Remarks is not changed:

"Against all this statistical negativity, there are still quite a few individual and sector charts the "look like" they want to make one more upward push. And we have lots of "news" due this week, and seems to be widely presumed as positive. So maybe there's room for temporary reprieve. But a downturn of some consequence seems to be developing pretty darn soon."

The reference to charts that look like they want to make another push is borne out in any number of charts on Thursday. I've attached, for example, the Retailing spdr (XRT). It sure looks like it's starting a legitimate new uptrend.

Also, the expected higher volatility has plainly arrived, and will likely continue. But until those Tail Skew (and other) statistics turn positive, I'd be very careful with this momentum.

Robin C.
Hanover NH

November 1, 2010. Friday's trading left our market outlook a little more ambiguous than expected when the Friday morning "extra" was composed. For example, Friday's closing Cumulative Skew (CSQ) differed sharply as between the OEX and NDX versions. Both plots are attached here, showing remarkable upside skew on Nasdaq and downside skew on S&P.

Friday's Tail Skew analysis had both tails stretched out quite notably (just about canceling each other out for the day). This stretch-out in the tails has been growing recently, and is apparent in the attached "Tail-EMAs-NQ" chart. It's likely a precursor to increasing volatility very soon. Note on this chart, too, how the recent low in the downside tail (blue path) just about matched the 2007 lows. That low doesn't predict that a top of 2007's magnitude is at hand (that's too much predictability to hope for). But when so few of the weakest stocks (the downside tail of cross-sectional distribution) stretch out deeply, a downturn of some notable magnitude is expected....after the downside tail begins expanding again, which it is now doing.

The final attachment here is a graph of the Sigma/VIX ratio. This ratio has fallen quite low of late (now .466), meaning actual S&P volatility is less than half the option-implied volatility. In a way, we can think of this low ratio as saying implied vol is actually high relative to the actual, meaning the options market is expecting the current low actual volatility will not last. Anyway, low values for this ratio are associated with intermediate tops, as evident back in April.

Against all this statistical negativity, there are still quite a few individual and sector charts the "look like" they want to make one more upward push. And we have lots of "news" due this week, and seems to be widely presumed as positive. So maybe there's room for temporary reprieve. But a downturn of some consequence seems to be developing pretty darn soon.

Robin C.
Hanover NH

August 9, 2010. The uptrend is still intact, but if it is to survive long it will have to generate more positive statistics than it has done recently. The cross-sectional stats have really been kind of punk of late, and a downturn of any consequence could readily push various indicators into serious negativity.

An example of this problem is attached. The Net Tails EMA is an exponential smoothing of the difference between the length of the uppermost positive tail area and the nethermost negative tail area of the daily cross-sectional returns distribution. As evident in the image, this net difference tends to run from +.25 and -.10 in bull market and from +.10 and -.25 in bear market. That is, the upper tail stretches out much more than the lower tail when the market is strong, and vice versa when the market is weak. So far in the July-August rally, this indicator has reached only up to +.08 and is currently just above +.04. The uptrend will probably extend for now, but danger is brewing. Other cross-sectional metrics confirm.

On a more positive note, our Prime Timer funds remain (mostly) positioned for further advance. Five of seven are at above-median exposure, although most have retreated from earlier outright bullishness. Updates of these gurus' ten-year exposure charts were posted at the web site on Friday. Click on the Analytix site link below and then "Prime Timers" on left-side nav bar.

Robin C.
Hanover NH

October 25, 2010. Last week's Market Remarks noted the previous "so far, so good" condition wasn't quite holding true any longer--likely weakness now approaching. We offered Net Tails graphs as evidence. The market has not turned down since then (aside from Tuesday's sharp retreat), but the Net Tails situation continues to weaken. I've attached here a repeat of last week's NYSE chart ("LW"), and a separate chart ("Current") updating that image.

There's a reasonable chance the major averages may yet reach up to the April highs (S&P and Nasdaq are only 3% and 2% below), but the risk of failing those levels is growing daily.

Robin C.

Hanover NH

October 18, 2010. Last week's so-far-so-good message doesn't quite hold this week. Not quite. The attached images show Net Tails (EMA basis) for the NYSE and for Nasdaq. On Thursday the Nasdaq Tails went suddenly and deeply negative. The series recovered on Friday, but the NYSE Net Tails went somewhat negative. (Because these attached images are EMA smoothings, the extent of the daily hits is not visible, but the effect on the smoothed Tails paths is apparent.)

A couple of days of negative Tails isn't the end of the world. But it's a change for the negative, and probably means the end of the present trend is near. Also note that neither of the blue "ema2" paths has approached its April highs. That could still happen...but for now is a looming divergence.

Robin C.
Hanover NH

October 11, 2010. The uptrend continues to weaken--as evidenced (e.g.) by low cross-sectional dispersion--but also, the trend continues. I've attached here two images; the first is a histogram of S&P 100 stock returns this past Friday (Oct 8). The positive skew is obvious. It's not terribly surprising to find positive skew on a positive day, but (a) the skew really is pronounced, and (b) recent down-days are often showing positive skew as well.

The second image here is an update of the Cumulative Skew (CSQ) path, which is derived from the daily histograms. So long as the positive skew days are larger than negative skew days, or more numerous, the CSQ path rises. This metric is not a leading indicator, but it has the great virtue of mostly "ignoring" market corrections that are unserious (as seen in April-June), but responding promptly when there's a real change at hand (as seen in March 2009). For now, the CSQ is still rising, albeit more slowly than back in the spring.

So the bottom line message might be characterized as "so far, so good," but it wouldn't take much to turn down.

Robin C.
Hanover NH

September 27, 2010. An image of our NYSE Core Momentum (attached) pretty well sums up the current market condition. Momentum remains positive (along with various other metrics, such as our daily skew), but is not at or near prior positive levels. So that presents some vulnerability, but prices can often grind along higher in face of vulnerability. Nothing much to do until there's some sign of actual negativity (rather than merely diminished positivity).

Robin C.
Hanover NH

September 20, 2010. Not much is new. Vulnerability continues, but without signs of imminent downturn. So further advance is the likelihood. In that case, if/when the S&P surpasses recent highs at 1130, we may well see a spirited follow-through.

I've attached a graphic showing five years of our NYSE Core Momentum and NYSE Core Skew paths. (Pay no attention to the relative interaction of the two plots; they're quite different metrics, plotted together only for the convenience of seeing both in one take.) Core Momentum (red) is lacking in oomph, but remains positive. The Core Skew (blue) is also positive and so far holding pretty strong, actually.

Robin C.
Hanover NH

September 13, 2010. Near-term indications are mixed. For example, cross-sectional skew remains positive (good!), but our Core Indexes are now beginning to lag a bit (bad). So direction is unclear.

One reason the cross-sectional skew remains positive is that the downside tail of the returns distribution is extremely limited. That's generally "good," because it helps keep the skew positive. But it can get "bad" (at least for the near-term) when carried too far.

I've attached a chart showing ten years of positive and negative tail paths on the NYSE. These paths track the tail-stretch in the cross-sectional distribution of daily returns; the reach of the outermost outliers. Both the positive and negative tail values have restored (fallen) to levels that prevailed through 2005-06-07. In fact, the negative tail at .0201 is near the lower end of that range. That level is back-traced through the years, with "drop lines" showing where the S&P 500 was when Negative Tail was at current levels. Generally, those occasions were closer to minor-or-intermediate highs than to lows...but often not quite there yet.

The path of the Negative Tail is a kind of "bad news sensitivity index." When the market is edgy, stocks with any adverse news get hammered, the negative tail is extended, and the blue Tail path rises. When complacent, the market is insensitive and even the outermost negative tail is limited. That's where we are today. Note the extreme complacency in the summer of 2007 (when the negative tail path reached down to its nadir). When that complacency shifted to anxiety in the fall, the bull market was kaput.

Robin C.
Hanover NH

September 7, 2010. Last week, Market Remarks noted "Friday was a big up-day, so it's not surprising the skew was positive. But Thursday was a down-day (-0.8% on average), yet both overall Skew and Tail Skew were both quite strong. These metrics do not yet say whether new highs are in store; only that upside is likely to prevail for the nonce." And "upside" certainly prevailed for the week. Clearly positive skew on a clearly negative day (as on Thursday August 26) is a real sign of underlying strength.

Daily skews also remained mostly positive through the week. Nasdaq had a bit of negative skew on Friday, however, in spite of the average 1.5% gain on the day. So continuation is likely in store, but last week's clarity is kind of played out. I've attached two-year graphics of the Cumulative Skew (CSQ) paths both for S&P and for Nasdaq. (The Nasdaq CSQ is new; I'll send a five-year Nasdaq CSQ image if you ask.) What's apparent in these CSQ paths is that the positive condition continues (raw path is above EMA), but crossover could occur in fairly short order if skew turns negative for more than a few days....so we're positive but not in condition "all clear."

(Separately, our Prime Timer equity exposures page now has September graphics posted. More positives than negatives, but also not "all clear." To view the updated exposure paths, click on Prime Timers in the nav bar at the home page.)

Robin C.
Hanover NH

August 30, 2010. Friday's rally looks like the start of a fresh uptrend. The NYSE and Nasdaq Core Momentum paths just down around neutral (saying the prior up-thrust to April highs has been fully corrected), as noted here last week. Meanwhile, cumulative cross-sectional Skew has remained positive (higher than its EMA), and cross-sectional Tail Skew has suddenly turned quite strong.

The attachment here shows cross-sectional distribution of returns for this past Thursday and Friday. Friday was a big up-day, so it's not surprising the skew was positive. But Thursday was a down-day (-0.8% on average), yet both overall Skew and Tail Skew were both quite strong. These metrics do not yet say whether new highs are in store; only that upside is likely to prevail for the nonce.

Robin C.
Hanover NH

August 23, 2010. Both NYSE and Nasdaq Tail Skew were positive on Friday. That is, the cross-sectional distribution of returns had a longer positive stretch in the upside tail than negative stretch in the downside tail. Friday's NYSE distribution is pictured in the "X-Sect-100820" image attached here. Positive skew in a down-day is a plus for trend outlook.

We can't draw important inference from a single day, of course. That's why we accumulate various skew values into time series...in order to see how they're progressing over time. The second image attached here shows an EMA path of our Nasdaq Core Skew. The period since the April high is divided into sections (a-b-c) of initial decline, the July rally, and the August pull-back. During the initial decline (a), I repeatedly said we could expect yet another up-trend approaching the April high or surpassing it. This was because we expect the skew indicators to develop weakness at a top, which they did not do in April (so the real top would be yet to come). Then as the July rally developed (b), I began to write of growing danger. This conservative turn was based (largely) on the failure of the various skew series to rise along with the market. Now we have had another price weakness, with the skew paths not weakening (and even rising very very slightly).

So we've seen failure to show weakness in the initial decline, then failure to show strength in the rally, and failure to be weak again in the very recent down. Taken together, the theme emerges that none of these sub-trends has really "meant it." After all the galoomphing around, the reality seems to be there's a kind of unpurposeful drift at hand. Directional bets should probably await more purposeful indications.

Robin C.
Hanover NH

August 16, 2010.  This week's attachment shows our NYSE Core Skew path. Our Core Indexes (NYSE and Nasdaq) exclude far outliers and also "noise" level fluctuation. The Skew path measures whether the cross-sectional returns distribution of Core Index stocks is pointing upward or downward (i.e., which end of the distribution is getting more play). What's notable in the attached chart is its almost total lack of response to the July rally. Other skew metrics show similar lack. These cross-sectional weaknesses are the basis of our various warnings of recent weeks.

On the other hand, it's also notable that the path has not plummeted down close to zero (as it did back in 2007). So it's still well up in positive territory. So we have a mixed picture for now. But the overall high level of the path is a product of months gone by, while July's non-response is pretty current, so the danger warning gets somewhat more (subjective) weight.

Meanwhile, the Core Momentum paths are down just about almost to neutral (zero) again. Last trip to down neutral prompted the following in our Market Remarks for July 6 (distributed July 5):

"The third attachment here shows five years of Core Index Momentum. The Core Mom generally inhabits positive territory in bull markets and negative territory in bear markets. The correction since April has now brought this metric down just about to zero...so far resisting the actual negativity it showed back in late 2007 (then having crossed zero when the S&P hadn't even lost the 16% it's down now). A negative Core Mom cross-over is not a "sell signal" in the magical predictive sense, but if the Core Mom gets to spending time below zero, that will cancel out a lot of other positives. Unless/until that happens, the presumption here is for uptrend renewal."

Those remarks are relevant again today, and we just have to wait to see whether the Moms return to positivity again...or not.

Robin C
Hanover NH.

August 9, 2010. The uptrend is still intact, but if it is to survive long it will have to generate more positive statistics than it has done recently. The cross-sectional stats have really been kind of punk of late, and a downturn of any consequence could readily push various indicators into serious negativity.

An example of this problem is attached. The Net Tails EMA is an exponential smoothing of the difference between the length of the uppermost positive tail area and the nethermost negative tail area of the daily cross-sectional returns distribution. As evident in the image, this net difference tends to run from +.25 and -.10 in bull market and from +.10 and -.25 in bear market. That is, the upper tail stretches out much more than the lower tail when the market is strong, and vice versa when the market is weak. So far in the July-August rally, this indicator has reached only up to +.08 and is currently just above +.04. The uptrend will probably extend for now, but danger is brewing. Other cross-sectional metrics confirm.

On a more positive note, our Prime Timer funds remain (mostly) positioned for further advance. Five of seven are at above-median exposure, although most have retreated from earlier outright bullishness. Updates of these gurus' ten-year exposure charts were posted at the web site on Friday. Click on the Analytix site link below and then "Prime Timers" on left-side nav bar.

Robin C.
Hanover NH

August 2, 2010. Cross-sectional stats continue to suggest--mostly--that the correction from April is just a correction. At the center of this inference is the mostly positive skew of daily market returns. But contradictory evidence is starting to creep in here and there. One such item is the negative skew in Friday's Nasdaq returns distribution.

I've attached here two images of our Net Tail EMAs; one for NYSE and one for Nasdaq. The Net Tails are simply the difference between the upside and downside stretch at the edges of a cross-sectional returns distribution. The NYSE version has so far failed to show a very dynamic upside in the July advance, but no tangible weakness, either. The Nasdaq version is also less than dynamic...but now, Friday gave a sudden downside pulse (downside tail much longer than the upside tail) in spite of purportedly neutral average price action (as the Composite was up 0.13% Friday). Indeed, the Nasdaq downward pulse pushed further negative than any time since the master rally began in March 2009.

One day's negativity in one market doesn't terminate an established uptrend. But it's a warning, and not the first.

Robin C.
Hanover NH

July 30, 2010. Friday's rally looks like the start of a fresh uptrend. The NYSE and Nasdaq Core Momentum paths just down around neutral (saying the prior up-thrust to April highs has been fully corrected), as noted here last week. Meanwhile, cumulative cross-sectional Skew has remained positive (higher than its EMA), and cross-sectional Tail Skew has suddenly turned quite strong.

The attachment here shows cross-sectional distribution of returns for this past Thursday and Friday. Friday was a big up-day, so it's not surprising the skew was positive. But Thursday was a down-day (-0.8% on average), yet both overall Skew and Tail Skew were both quite strong. These metrics do not yet say whether new highs are in store; only that upside is likely to prevail for the nonce.

Robin C.
Hanover NH

July 26, 2010. Our Cumulative Skew path (CSQ) continues strong, with continuing implication that the correction from April highs is just a correction. (Image attached.) Thursday and Friday of this past week, however, were not as internally robust as the averages would suggest. While the S&P is up 3% from Wednesday's close and the Russell 2000 is up more than 6%, the NYSE downside tail exceeded the upside tail both days. So the Net Tails paths declined, with implied weakness. As shown on the attached image, the Net Tails EMA paths (attached) are still OK for now, but they'd better resume their upside progress or there'll be trouble.

Robin C.
Hanover NH

July 23, 2010. Both NYSE and Nasdaq Tail Skew were positive on Friday. That is, the cross-sectional distribution of returns had a longer positive stretch in the upside tail than negative stretch in the downside tail. Friday's NYSE distribution is pictured in the "X-Sect-100820" image attached here. Positive skew in a down-day is a plus for trend outlook.

We can't draw important inference from a single day, of course. That's why we accumulate various skew values into time series...in order to see how they're progressing over time. The second image attached here shows an EMA path of our Nasdaq Core Skew. The period since the April high is divided into sections (a-b-c) of initial decline, the July rally, and the August pull-back. During the initial decline (a), I repeatedly said we could expect yet another up-trend approaching the April high or surpassing it. This was because we expect the skew indicators to develop weakness at a top, which they did not do in April (so the real top would be yet to come). Then as the July rally developed (b), I began to write of growing danger. This conservative turn was based (largely) on the failure of the various skew series to rise along with the market. Now we have had another price weakness, with the skew paths not weakening (and even rising very very slightly).

So we've seen failure to show weakness in the initial decline, then failure to show strength in the rally, and failure to be weak again in the very recent down. Taken together, the theme emerges that none of these sub-trends has really "meant it." After all the galoomphing around, the reality seems to be there's a kind of unpurposeful drift at hand. Directional bets should probably await more purposeful indications.

Robin C.
Hanover NH

July 19, 2010. Friday's relentless decline notwithstanding, it sill looks like the correction form the April highs is still just a correction. Perhaps we're about to revisit the recent lows, but so far no change in the correction-only inference. One example is the Net Tail Skew pictured in the attachment here. The 2nd EMA has refused to push down into negative bear market area, and the last upward push was clearly into positive bull market area. Other skew metrics concur; the tide has not turned.

Robin C.
Hanover NH

July 12, 2010. Our Core Momentum paths (both NYSE and Nasdaq) bounced off of the zero zone this week, as expected. The NYSE Core Momentum path is shown in an attached image. (The small red rectangle shows the bounce from zero.) We've noted for some time that the correction is probably only a correction so long as these Momentums don't sink into negative territory....and also so long as they don't create significant negative divergence as in 2007. (The large yellow rectangle highlights the 2007 divergence.)

The other feature we've been harping on is cross-sectional skew, which also remains positive. Last Wednesday's big 3% rally was a bit troubling in that it was so symmetrical it contributed almost nothing to the positive skew. But Thursday and Friday profiles made up for it with notable positive skew.. So it seems "all's well" for now. I've attached a page showing cross-sectional skew profiles of the four-day week. Look for where the red curve is higher than the blue: if to the right, it's positive skew; if to the left, negative. You can see how neutral Wednesday was in spite of the big advance.

Robin C.
Hanover NH

July 7, 2010. Here (attached) is a "special edition" mid-week Market Remarks. It's issued today because the present positioning of VXX and S&P is just so neatly aligned. Alas, with only 18 months of VXX history, all inference is tentative. Nevertheless, I thought you'd have some interest in seeing it.

Robin C.

Hanover NH

July 6, 2010.  We continue to find cross-sectional skew is positive in various metrics. I've attached a profile of Friday's NYSE cross-sectional returns distribution (T-Skew.gif). Although the average return was negative (about -0.8%), the positive tail (to the right) is plainly greater than the negative tail (to the left). Positive Tail Skew says that although the average and the overall distribution shifted down, buyers were willing to bid further above the mean than sellers were willing to offer down below the mean. That's positive skew, and it's more characteristic of a corrective down-day than of a down-trend down-day.

The positive skew did not just suddenly emerge on Friday. It has (mostly) persisted throughout the correction so far. The Cumulative Skew (CSQ) path accumulates daily skew into a continuous path, and it has simply refused to turn down (much) in the market retreat since April. (It flirted with downturn a couple of weeks ago, but has since resolved upward to new highs.) See CSQ image attached.

The third attachment here shows five years of Core Index Momentum. The Core Mom generally inhabits positive territory in bull markets and negative territory in bear markets. The correction since April has now brought this metric down just about to zero...so far resisting the actual negativity it showed back in late 2007 (then having crossed zero when the S&P hadn't even lost the 16% it's down now). A negative Core Mom cross-over is not a "sell signal" in the magical predictive sense, but if the Core Mom gets to spending time below zero, that will cancel out a lot of other positives. Unless/until that happens, the presumption here is for uptrend renewal.

Finally, I note that five out of seven of our Prime Timer fund managers are carrying above-median equity exposures. They've all retreated from their high exposures of early spring, but only two are in outright bearish position. Fresh Prime Timer equity exposure charts were posted at the web site Friday.

So far, this correction is generating statistics looking like "just a correction."

Robin C.
Hanover NH

June 28, 2010. The Cumulative Skew (CSQ) path, which has been threatening reversion to a state of decline, perhaps began to redeem its uptrend on Thursday and Friday. The positive skew on Thursday is impressive, as most market averages were down more than 1.5%. (And Friday's positive skew is somewhat impressive too, as large cap stocks--on which the CSQ is based--were nearly unchanged on average, yet the positive skew continued.) I've attached a picture of the updated CSQ path.

On the other hand, the outer tails of cross-sectional returns distribution skewed negative on Friday. The second attachment here illustrates the situation. Although the distribution skews positive through the middle, the worst losses (left side) were far more deviant than were the best gains (right side). One day like that is not dispositive, but if they accumulate in the face of general rally, it's a real sign of market weakness. (You've seen the Tail Skew index and how it tanked starting in mid-2007.) We've already had a degree of non-response in the Tail Skew in the recent mini-rally.

As the market pulled back from April highs, I said the highs would likely be revisited or even topped. That view was based largely on the fact that various skew measures were still at or near highs, even as prices receded. Revisiting those highs seems less likely now, but some renewed rally is still probable. If so, it had better generate some really positive data, "or else."

Robin C.
Hanover NH

June 21, 2010. We may be approaching the beginning of the end (of the master rally from March '09). That's based on some cross-sectional skew metrics that have yielded somewhat to the correction, and now have been only weakly responsive to the ostensibly stronger up-days of the past two weeks. You can see this in the Cumulative Skew (CSQ) picture attached here. The CSQ has been resisting decline in the correction (good), but should get going again as prices advance...which it's not doing yet (not so good). This indicator is not for generating specific buy-sell signals, so the fact that it's now below its EMA is not a call for action, but does say we're entering a caution zone. (Note how it hemmed-and-hawed around its EMA back at the top in 2007.)

That the sluggish CSQ is only an early warning now is illustrated by the second attachment, showing three years of our NYSE Net Tail series. This tail-skew metric is based on the difference between the stretch of the upside tail and downside tail of cross-sectional returns. (Rather than using the whole distribution, it looks at the extremes.) Note how in bull market periods it reaches up into the +.20 to +.25 range, but down only to about -.10. In bear markets it reached down to -.20 to -.25 range, but only up to about +.10. For now it seems to have finished a trip down to -.10 and heading up.

Robin C.
Hanover NH

June 14, 2010. There's a pretty good chance last week's Thursday-Friday upturn has more to go. Actual volatility is running higher than implied volatility, which usually indicates near-term strength, and the various cross-sectional skew metrics are still resisting decline. (Graphic available on request.) The S&P Cumulative Skew (CSQ) has continued horizontal in spite of market decline, and our Tail Skew has been pretty much neutral in midst of price decline...and now maybe turning positive. (See attached illustration of Friday's positive Tail Skew.)

Meanwhile, our Prime Timers have all reduced equity exposure from their recent highs (and four of seven reduced exposure this week), but we still have four at above-median exposure. So the aggregate message from the Prime Timer managers seems to be growing caution but not yet a net bearish outlook. June graphics of their exposure paths were posted last week at the web site.

Robin C.
Hanover NH

June 7, 2010. This market moment is more difficult to assess than any since the 2009 bottom. Price charts for stocks, commodities, and (non-US) currencies look like hey're about to utterly collapse. And our Hedge Fund Analytix tracking says Managed Futures hedge funds that were way too equity-exposed have just this past week finally relenting.

On the other hand, almost all of our cross-sectional metrics are saying the market's underlying behavior is NOT acting weak. I've attached a plot of the daily NYSE Tail Skew, for example. You will see that it is down notably from its April peak, and in some ways looks like the shape of 2007. But back in August 2007 that skew path had dropped by 0.33 (and was negative!) when the S&P was off only 9.4%. Today in contrast, this skew path has dropped only 0.18 (and remains positive), while the S&P is off 12.5%.

I've also attached an update of the S&P Cumulative Skew (CSQ). It has weakened in this just-past week, but remains above its EMA and has to be considered relatively strong in the face of recent price weakness. Again, compare today with 2007.

Overall, the data here suggest we should still expect a soon substantial rally (perhaps recovering recent highs). If we can just get past the risk of a one- or two-day black hole (either by its arrival, or by quiescence), there ought to be upside available. That hedge fund selling will take its toll, but maybe not quite yet..

Robin C.
Hanover NH

June 1, 2010. Lots of charts (sectors, countries, etc) look much like they want to go for another drop, and Friday's late sinking spell suggests near term weakness, too. But our cross-sectional analysis continues to deny we've entered a new bear phase. I've attached a plot of the CSQ Cumulative Skew path. Even this stalwart metric sank downward on Friday, but it remains above its EMA and has been resisting the correction quite consistently.

Robin C.
Hanover NH

May 26, 2010. A month ago I pointed out that as of April 28 the sequential scatter of S&P vs VXX showed the relationship to be at the upper bound of a nearly linear pattern. That position implied that S&P would likely fall more than VXX would rise, at least restoring the core relationship.

That over-high S&P (relative to VXX) has now reversed. The S&P/VXXscatter point today is nearly at the lower bound of the nearly linear pattern. Implication is that the S&P will gain more than proportionally to VXX loss (or fall less than proportionally to VXX gain). That is, the scatter will likely return at least to mid-channel, or to top-of-channel.

I've attached the original and current scatter diagrams

Robin C.
Hanover NH

May 24, 2010. Last week's high VIX/DVAX ratio gave unfortunate inference. Citing the historical 27 occasion when the ratio was as high, the average 1-week S&P outcome was a gain better than +3%. But this occasion (the just passed week) turned out -4.2%. Moreover that VIX/DVAX ratio is now at an unprecedented 1.43 ... which we can't even scale to any historical outcomes. So (alas) we don't really have any basis to assess whether the past week's result was simply different (historical inference wrong), or delayed...or what?

Be that as it may, other data here--our cross-sectional distributions analyses--continues to NOT indicate cancellation of the master rally from March '09. For instance, having a 10% correction with no downturn in the CSQ index of cumulative skew is remarkable. Other skew measures are also resistant to downturn. Like any market metric, skew can change direction "any time it wants to," but so far it just doesn't seem to want. I've attached a graphic showing three years of CSQ, with rectangles noting the very different nature of current market pull-back vs 2007 top.

Robin C.
Hanover NH

May 17, 2010. I've noted repeatedly that the master rally is unlikely to have ended with our skew metrics having posted new highs at the last index highs. That's still true, even if the present correction continues lower. Cross-sectional skew has retreated as the market retreated, but not unduly (and maybe not even "duly"). Our CSQ (cumulative S&P skew path) is essentially horizontal in the current correction; not declining, and not crossing its EMA. (Send a note if you want to see it.)

Moreover, Thursday and Friday of last week had dispersion shrinking notably, while VIX increased. That has pushed our VIX/DVAX ratio up into rarified areas. The ratio now stands at .897. (DVAX is a key index of cross-sectional dispersion, which is a dominant factor in market volatility, so the ratio is more stable than either metric itself.) In DejaVu analysis covering the past nine years (2,258 days) the VIX/DVAX ratio has exceeded .85 only 27 days. In the ensuing 1-week periods (5 days following each of those 27 occasions exceeding 0.85), the S&P 500 gained an average of 3.15%. All other 5-day returns averaged slightly negative, and the difference is statistically significant.

No metric will tell us whether we're about to renew the master rally from 2009. But (a) the rally is probably going to resume (and post new rally highs) in due course, and (b) there is good likelihood we will see at least an attempt at renewal in the coming week.

I've attached a graph showing he VIX/DVAX ratio, and a table showing the DejaVu outcomes following ratio > 0.85.

Robin C.
Hanover NH

May 10, 2010. Well, that was an interesting week. But (strangely) we don't really see much that's new. The correction is no surprise, except for its ferocity, and our various metrics (mostly) say not much has really changed. We still have high and declining cross-sectional skew, and we still have high and declining momentum paths. No surprise there. The Core Indexes (NYSE and Nasdaq) are strong relative to fixed-comp indexes.

I've attached here two graphs showing the NYSE and Nasdaq Core Skew paths. These two are typical of the metrics that are high-and-declining, but which are not particularly bearish. Note that in 2000 and in 2007 these indicators were way off their highs when the market topped. There's no law saying every top has to develop like that, but tops do tend to take longer time than rational minds expect. It's possible that a day like Thursday's panic changed everything (resetting the mindsets, making all the positive metrics of recent months is moot). If so, we need to wait for some evidence.

I've also attached an update of the VXX and S&P scatter diagram I sent on April 28. At that time, I wrote "...either S&P should decline disproportionately or VXX should rise disproportionately....perhaps to the lower bound of the trade-off band." It may seem we got both. As you will see, Friday's VXX and S&P values put the scatter point right about mid-band. Thursday's low would be just about at the end of that little "peninsula" of red (market's February low), at the lower bound of the range.

Bottom line: Very next moves imponderable, but so far data still suggest the master rally is still alive. The one thing we know for sure is that volatility has persistence. Expect substantial fluctuation.

Robin C.
Hanover NH

May 3, 2010. There's a pretty good probability that recent minor weakness has more to come. Maybe a good bit. But the new highs posted by various cross-sectional analysis strongly suggest the larger-scale uptrend is not finished. It would be unusual (to say the least) for the master rally (from March 2009) to have ended with cross-sectional skew making new highs, as it did just seven trading days ago. I've attached here a graph of the NYSE Tail Skew. Note how the Skew peaked in 2007 a couple of months before the S&P peaked. We have no rule about any specific lead times, but there is every likelihood of seeing new market highs (without Skew participation) before the master rally is done.

Robin C.
Hanover NH

 

April 28, 2010.  Here is an interesting development. Last fall I wrote a report examining how the S&P and VIX can be viewed in context of conventional risk-return trade-off, but with the specific trade-off relationship shifting over time. Now we find that near-term VIX futures, represented by the VXX ETF, illustrate the relationship even more plainly, because the trade-off with VXX is more stable (than it is with VIX itself).

The attached scatter diagram shows the trade-off, with VXX levels plotted horizontally and S&P vertically. The dots are connected and color coded so you can see the sequential development. (The color changes at each new calendar quarter.) Although the VXX history is short (1 1/4 years), it is plain that the relationship to date is nearly linear, within a band illustrated by the yellow diagonals.

Right now, the position is at the high end of the linear band. That is, S&P is high relative to volatility (or equivalently, volatility is low relative to S&P). If the relationship holds within the parallel boundaries, either S&P should decline disproportionately or VXX should rise disproportionately....perhaps to the lower bound of the trade-off band.

Robin C.
Hanover NH

April 19, 2010. Friday's market drop was the largest since early February (S&P -1.61%), accompanied by significant negative breadth and volume. The WSJ cited the Goldman fraud charges and a "rush to safety," which are probably valid observations. So maybe we're at a nascent correction of the market's relentless and shapeless upward plodding since February.

But in spite of the overbought condition, and in spite of the long term risks, our metrics do not suggest an end of the major rally (from March 2009) is at hand. For example, not a single one of our cross-sectional skew models has begun to head lower (as they would normally do approaching a major downturn). Cross-sectional skew measures the shape of the daily returns distribution; in a weakening market, the "upper tail" contracts (even while the distribution is still positive). As the upper tail contracts, the overall distribution skew becomes less positive, or negative. We have seen none of that. I've attached a graphic of our Cumulative Skew (CSQ), and of the NYSE Core Skew, noting the difference between current condition and 2007 weakness.

The implication is that even if Friday was the start of a more-than-plausible correction, the major rally is extremely unlikely to have topped yet, and therefore highly likely to post new highs in due course.

Robin C.
Hanover NH

April 12, 2010. Statistically speaking, the market is overbought and overdue for correction (as everyone knows). So I guess that's likely soon. (But of course, it was "likely" a week ago, too!) And a number of metrics are saying the market is vulnerable. (For example our over-bullish equity exposure of CTA hedge funds, and the extremely low ratio of actual/implied market volatility.)

Nevertheless, it seems quite unlikely the next pull-back will be the end of the 2009-2010 rally. Various momentum and skew metrics are at new rally highs (or very nearly so), and big up-moves very rarely end on new high momentum (or skew). Bottoms can often spring right off of the worst down-thrusts, but tops don't usually do that; market tops tend to just wear themselves out.

By way of example, I've attached a graphic of our NYSE Core Index Momentum path. The current position is up in a range that adds its own suggestion of vulnerability, but it's certainly not looking tired or worn out. Unless Momentum were to reverse really stunningly (in which case, that would be apparent in itself), expectation should favor further rally highs yet to come, whether soon correction emerges or not. We've invoked this logic earlier in the rally, and so far it's proven out.

Robin C.
Hanover NH

April 5, 2010. The market hasn't done much of late, so data analysis shows not much new. We expected a down-ish week, which didn't happen, but not much has changed. There are positives (various measures of skew, and generally expansive portfolio position among our "guru" funds) and negatives (low volatilities, over-bullish position among CTA hedge funds, and a truly scary weakness in the bond market), and our expectation is that the weaknesses will soon prevail in a correction, but after that we will find new rally highs yet again before this monster rally is done. This latter expectation (new highs before it's over) is based on the various skew metrics still posting new highs. We would be astonished to find the rally end until we have weak skew metrics.

Speaking of the weakness in volatility metrics, take a look at the attached graphic showing implied and actual volatilities. Implied volatility is at lowest point since May 2008. Actual volatility is far lower, where it hasn't been since June 2007. The chart shows "drop lines" back where those prior occasions occurred...and they were occasions for defensive position.

I've also attached a bar chart showing one-month forward returns from each decile of Value Funds equity exposure. We added Value Funds exposure to our Prime Timers analysis last year, because their timing scores outperform even our stable of the six fund gurus. The Value Funds are currently in their 8th decile, which you'll want to consider when you examine the bar chart. The Prime Timers exposure charts at the web site were all updated to April 1 this weekend.

Robin C.
Hanover NH

March 29, 2010. Last week we showed the Friday (March 19) plunge in Tail Skew, noting it increases the likelihood of soon downturn, but that the downward spike often produces about a week of higher prices first. Now we've had a week of mildly higher prices (+0.58% for the S&P), with no particular reason to change expectation from impending downturn.

How big a downturn? Don't know. But large or small, I expect the downturn will not be the end of the year-long rally. As in the January pull-back, cross-sectional skew (in various measures) has been posting new highs (and almost new highs). Typically, we will see some negative skew (or at least, not-very-positive skew) before serious decline. I've attached a picture of the S&P Cumulative Skew (CSQ). You can see it's way above its EMA---so ready for correction---but for the same reason, not hinting at substantial weakness.

Robin C.
Hanover NH

March 22, 2010. There's a pretty good chance that Friday was the beginning of the correction everybody's been watching for. Take a look at the attached graphic showing two EMA smoothings of net Tail Skew on the NYSE. Although the downward price changes for the day Friday were really quite ordinary (barely perceptible on the S&P plot), the short-term Tail Skew (ema1, the red path) plunged dramatically from the longer ema2 (blue path) up in its top range. You can see several examples of this behavior just in the three years shown. This doesn't mean everything turns on a dime; in fact, statistically, prices are often higher a week after a plunge like that..

I've also attached a one-page review of the current VIX situation. This analysis is prompted by the VIX stories in Saturday's WSJ and Monday's Barron's. The WSJ story concludes that VIX is low relative to VIX futures, implying that futures traders expect rising volatility (and falling prices). The ratio analysis here says it's even simpler; implied volatility is way higher than actual volatility, saying that option traders are expecting higher vol. The Barron's story wonders "...why implied volatility is so low when investor fears are so high." The answer is that recent actual volatility is the key ingredient in VIX, and recent actual volatility is even lower.

Robin C.
Hanover NH

March 15, 2010. The S&P has posted a new rally high, as expected here. Now the DJIA is the only major index without new rally highs. Although the February-March advance now has everyone anticipating pull-back from overbought position, it's likely the Dow will get it's chance too...if not "directly," then after a rest.

I've attached here both an update of the NYSE Core Momentum, and the NYSE Core Skew. The Momentum path is still kind-of, somewhat in negative divergence. So a pull-back, it emerges promptly, could look like maybe end-of-big-rally for momentum. But look at that the Skew path! It's posting new highs most every day. So in terms of the Skew, a down-turn here would likely be followed by yet another push to new highs, in which the Core Skew would finally fail to confirm.

Robin C.
Hanover NH

March 8, 2010. Last weekend we said "The year-long uptrend seems to be intact, and remains likely to post new highs." Most major indexes have now posted new highs. The S&P is a little behind (within 1%), but will likely make its own highs soon enough.

We also drew a comparison between now and the weakness of late 2007, illustrating with our NYSE Core Momentum path. On further reflection, some clarification seems appropriate. While the negative divergences themselves are comparable (now and then), it's also worth noting the difference that current MOM levels are a good bit higher than 2007. Also, no matter how comparable any two set-ups may be, the play-outs will differ. There is no implication that the next down-leg will (or will not) be as severe as the prior occasion; only that there is indication of underlying weakness waiting to surface.

Our Hedge Fund Analytix work now shows equity over-confidence among CTA funds. Subscribe now if you haven't yet.

I've attached an update NYSE Core Mom graphic.

Robin C.
Hanover NH

March 1, 2010. The year-long uptrend seems to be intact, and remains likely to post new highs. But the thrill is gone, and risk is high (specifically the risk of not getting to those new highs). I've attached a graphic of our NYSE Core Index Momentum, which more or less summarizes the situation overall. This momentum path remains suitably positive, but has so far failed to match the highs it posted last September. That failure looks somewhat like the failure to confirm new highs in October 2007. (Compare the red rectangles.) Various other metrics are positioned similarly, while volatility and dispersion are quite low.

We still find cross-sectional skew is positive (the upper limb of returns distributions stretch out more than the lower limb), which is the primary basis for expecting the uptrend is still intact. It's not an attractive bet, however. For the risk averse, better to just watch.

Robin C.
Hanover NH

February 22, 2010. Not much is new (again) this week. Cross-sectional metrics remain positive. Momentum remains positive, too, although not terribly vigorous. An image of the NYSE Core Index Momentum is attached. Our Core Indexes track aggregate price action of stocks that move notably but not extraordinarily. That is, they exclude the trivial "noise" of minor daily fluxion and we exclude the outliers. The result is an index that is extremely trendy, and whose Momentum index tends to remain positive in bull market and negative in bear market.

A growing cloud on the horizon is the recent retreat from high exposures among our Prime Timer funds. The Prime Timers are five high-profile fund managers (plus a fund aggregate) with historically positive market timing histories. All but one have been persistently bullish since the bottom in 2009, but are now backing away from that posture. Moreover, the three best timers have been retreating the most. Last Friday's exposure tabulation is posted at the web site ("Prime Timers" button on the navigation bar), along with recent exposure graphics for all six.

Growing caution among top managers is cause for great caution for all, but there is no objective rule or threshold to tell us when the jig is up. So we sit on the edge of our seat, and watch for change in the action-specific metrics like Core Momentum (above) or cross-sectional skew. As in prior Market Remarks, we expect the market will post new rally highs first.

Robin C.
Hanover NH

February 15, 2010. It's harder than usual to guess which way this cat will jump. Many price charts look just dreadful. But the market's internals seem pretty much OK-to-good. I've attached an update of the Cumulative S&P Skew (CSQ). This is a coincident, not leading, indicator. But it has not only not turned weak; it's actually posting new highs! Just after the market highs of 2007 the CSQ (red) crossed under its EMA (blue) before the S&P had dropped as much as it has here in 2010.

I've also attached an image of the NYSE Normalized Net Tail Skew. It's declining from the high end of it's upside range (a condition of potential danger), but it's not nearly so steep a decline as in 2007, and not nearly so protracted as in 2000. Using 2000 and 2007 as a guide, it would be expectable for the market to post a new high with the Normalized Net Tail Skew failing to do so. That has not happened here, so it seems likely we have yet to see the final high of the 2009-10 rally. How to reconcile this with those dreadful price charts, I can't say.

Robin C.
Hanover NH

February 8, 2010. The DejaVu outlook was right again last week, at least directionally. Instead of the indicated average -2.6% at week-end, we had only -0.9% (S&P). But that Thursday-to-Friday-morning drop hit the model's indicated average outcome pretty closely.

Alas, this week doesn't find a crisp data situation to model for another DejaVu go. But the vigor of that Friday afternoon rally was impressive, reminding that in spite of all the dangers (and there are plenty), we have those cross-sectional skew metrics saying the market is nowhere as weak as it seems (or for Nasdaq, actually showing cross-sectional strength).

I've attached three graphics; NYSE Tail Skew, Nasdaq Tail Skew, and the S&P Cumulative Skew (CSQ). The first might have topped, but in fact is holding pretty strong for three weeks of price decline. The second is definitely strong, posting new rally highs most days. (Even though it's way up in reversible territory---like 2007---it's still strong until it stops being strong.) The third attachment, CSQ, could yet cross negative,but so far is doing pretty well. Without guessing whether very recent price weakness is finished or not, there's still a pretty good likelihood the rally from last march is unfinished, with one more up-leg to new highs yet to come.

Robin C.
Hanover NH

February 1, 2010. One of last week's images was the NYSE Tail Skew. It was (and is) dangerously high, and its most recent nearby peak had just barely almost matched the prior (October) peak, thus creating a near-term negative divergence from market. The market's decline over the past week makes that situation more plainly a finished divergence.

Last week we also noted our Nasdaq Tail Skew was quite different from the NYSE, posting new-new highs pretty readily. Indeed, the Nasdaq Tail Skew and Core Skew have posted fresh highs again this week...with the exception of Friday, which ticked down. This positive Nasdaq behavior is essentially what's kept us from outright bearishness, thinking the correction (quite possibly sharp) is likely to be followed by another run at new highs (with even the Nasdaq Skew then failing to confirm).

That remains a plausible scenario. But in the meantime (and nearer-term), we've run our DejaVu model scanning for times when NYSE Skew is falling for ten days and Nasdaq Skew is rising. Over the past 2,250 market days, there have been 11 days of such difference with magnitude similar to today's. The average one-week outcome from those 11 occasions was an S&P loss of -2.6%, significantly different from the average of all other weeks. An image of the DejaVu outcomes table is attached.

Robin C.
Hanover NH

January 25, 2010. Last week's Market Remarks noted the excessively high Nasdaq XUB reading, pointing out that prior occasions have resulted in average one-week declines of 3.29% . Here we are a week later with the Composite down 3.61%, so the precedent holds.

Alas, that doesn't help much going forward. Frankly some sector indexes (and country indexes!) look just terrible. I've attached a picture of XLK, the SPDR Technology ETF. Although I don't put a lot of stock in trendlines per se, this case is saying one thing clearly: the dynamic of the past year is no longer operative.

Also, some metrics are way up in readily reversible range. I've attached a picture of our NYSE Tail Skew, with a red line showing current/recent level to be about where other declines have started. On the other hand, the fact that the NYSE Tail Skew almost posted a January high matching the October high (and the Nasdaq version did post a January high) suggests that we may well get another upside try before we come to full correction (or full reversal) of the run-up since last March. But on the other-other hand, even a run to new highs could occur after a painful correction (say maybe 15%?).

All of which is to say; New rally highs are still quite possible, but now risk is too great for this scaredy-cat.

Robin C.
Hanover NH

January 18, 2010. Most indexes were down at least a percent on Friday, but almost all our skew measures were positive. That is, the positive tail of daily return distributions had longer reach or more stretch than did the negative tails. So this is good, right?

Well, maybe not right away. On the Nasdaq market the upside tail was just a bit too positive. This excess upside gave our Nasdaq XUB a reading of +.44. Our DejaVu model tabulates past occasions of present conditions. And DejaVu shows there have been only five days in the past ten years when Nasdaq XUB was greater than +.40 while the Nasdaq Composite was down more than 1%. In the week ensuing from each of those five days, the Composite lost 3.29% on average. All other week-long spans averaged -0.11%, and the difference is statistically significant. An image of the DejaVu printout is attached.

The XUB metrics are not generally "contrary" indicators, but they do sometimes show little bursts of over-indulgence at the end of minor trends.

Robin C.
Hanover NH

January 11, 2010. Upward drift continues. An interesting and possibly important aspect is that cross-sectional skew has turned surprisingly positive. Cross-sectional skew refers to the shape of the daily returns distribution. When the upward reach of the positive end of the distribution is greater than the downward reach at the negative end, skew is positive; when downward reach is greater, the skew is negative. As with most market metrics, we concatenate successive values to generate a continuous path.

The Core Skew paths (NYSE, Nasdaq) have two salient characteristics. First, they tend to remain positive in bull markets and negative in bear markets. Second, they tend to top-out (and sometimes bottom) ahead of average prices. Back in November I grew quite skeptical of the rally in part because the skew paths started to drop sharply.

But now both the NYSE and Nasdaq Core Skew paths (and other skew metrics) have strengthened a great deal. I've attached two graphics to illustrate. The NYSE version has almost posted a new rally high, and the Nasdaq version has actually done so. Because these metrics almost always top-out ahead of the market, the implication (especially if the NYSE skew also posts a new high) is that a correction, when it comes, is quite unlikely to be end-of-rally. Much more likely, following correction, is to find another market high in which cross-sectional skew fails to confirm.

Robin C.
Hanover NH

January 4, 2010. One might expect that after many weeks of writing about the up-trend's weakening condition, that Friday's end-of-day drop would drive me into downside outlook. One would think. But no.

Friday turned out to have some remarkable internal behavior inconsistent with new downtrend. For one thing, various measures of cross-sectional skew were positive. Our CSQ (cumulative skew) path posted a new rally high. And the Core Indexes gave positive Tail Skew, both on NYSE and Nasdaq. Both Core Skew paths had been remarkably weak, but have recently begun sneaking a bit of strength.

I also point out that contrary to almost all journalistic commentary, the VIX volatility index is not low...at least not when seen in context of actual volatility, and not relative to market dispersion. I've attached here a five-year plot of VIX (implied vol)and the S&P Sigma (actual vol), and you can see the Sigma has been just plummeting. The Sigma/VIX ratio is now below .50, which has only occurred on five days over the past 2,257 days (nine years).

The second attachment shows VIX in comparison with our DVAX dispersion index. Here too, the DVAX has plummeted, while the VIX has held. Friday's ratio to dispersion (VIX/DVAX > .90) has only occurred on 8 days over nine years...with average ensuing one-week S&P +5.7%

As noted last week, holiday-shortened days and weeks can produce strange data. So maybe the positive skew and exceptional volatility ratios are just holiday artifacts. But prior year-ends haven't looked like this. The more straightforward reading is that Friday's late sell-off is/was not as weak as it looked, and NOT yet the end of the rally.

Robin C.
Hanover NH

 

 Market Remarks Archives:

2006        2007      2008       2009    2010   2011   2012   2013   2014

 

 

Carpenter Analytix offers data and analysis, not advice. Comment is limited, so the data can speak for itself.  Market inference requires objectively significant relationships. Absent significance, there is no inference. 

Market Remarks are posted Monday mornings, with emphasis or explanation to the outline of current Daily States.

If you are not yet receiving e-mail delivery of Market Remarks  please open the Services page to send in your request.

Nothing at this site is offered as advice or recommendation.
All site content is subject to full Disclaimer statement and Terms of Use.
Copyright© CarpenterAnalytix.com 20010. All rights reserved.