ES -- DX/CL -- isee -- cboe put/call -- specialist/public short ratio -- trinq -- trin -- aaii bull ratio -- abx -- cmbx -- cdx -- vxo p&f -- SPX volatility curve -- VIX:VXO skew -- commodity screen -- cot -- conference board

Wednesday, November 21, 2007


the case for holding on

i was clearly (painfully!) too early in going long. there's nothing wrong with covering a short too soon; taking the other side, though? i might have waited for sustainable bottoming action in more than 20-day new lows!

but there's Real Fear out there right now. to wit:

And what's interesting is that looking forward 90 days, traders are as fearful now as they have been at any time all year. All of the last five years in fact. Converting to SPX options, this particular measure has not been this high since 2002.

in fact:

I hit up volatility charts for 30 Day, 60 Day and 180 day readings going back 10 years, and no matter which one you look at, the story is similar. We are very near five year highs. The 30 and 60 day numbers were a bit higher in August, but other than that, we have not seen these levels since 2002.

something else to note about volatility -- VXO is making lower highs as the market makes lower prices, and that is a good sign. other examples include march 2007, october 2005, october 2004, august 2004 (indeed, from a larger perspective, the entire 2004 decline from february to october), february/march 2003 (and indeed the whole triple-bottom set from october 2002 to march 2003), february 2003, march 2001, december 2000, may 2000... anyway.

and the isee:

The ISEE’s close of 68 on Friday was the seventh lowest end of day reading in the index during the five years for which data is available.

or, in more detail:

Data from the International Securities Exchange, or ISE, also began to show signs that complacency may be breaking down. As I noted last week, the all- securities call/put ratio at the ISE hit 67.85 on Friday (November 16). Readings as low as this are pretty uncommon – counting this latest one, there have been 22 going back to the beginning of 2006. And many of these readings have occurred in clusters.

My colleagues, Bob Becks and Joseph W. Sunderman, took a look at the implications of a low ISEE call/put ratio (below 90). In their quantitative study, they eliminated signals that occurred within 20 days of the first signal, due to the tendency of these extremely low call/put ratios to occur in clusters. That leaves a total of seven unique signals. Ten trading days after these signals, the SPY has been positive 86% of the time with an average gain of 1.1%. Thirty days out, the SPY was positive 100% of the time, with the average return being 3.2%. The 90-days period following such a signal is really impressive. The SPY was in the black 100% of the time with an average return of 9.1%. Unless we're transitioning to a bear market, which I highly doubt, this study suggests that it's a good time to be long the market, no matter how gut-wrenching it may be on a day-to-day basis.

and of course there is the 20-day new lows, which have continued to make lower highs in the s&p 500 even as the index price has ticked lower and as 65-day new lows have grown. dr. brett steenbarger has been following along:

Going back to September, 2002, which is when I first began collecting the data on 65-day new highs and lows, we've had only 22 occasions in which those new lows [across all exchanges] have exceeded 1500. Forty days later, the S&P 500 Index (SPY) was up every single time, by an average of 5.33%. That is much stronger than the average 40-day gain of 1.83% (845 up, 398 down) for the remainder of the sample. The bear case is that this time is different, owing to dynamics of housing, credit, etc. If that's the case, we should see new lows expand from here and beyond the levels registered in August of this year. The bull case rests on a drying up of new lows and then an expansion of new highs going forward. I'll be tracking that measure closely.

there's a battle on today -- the day before thanksgiving, the most bullish day of the year historically -- to again defend s&p 1425, a price level critical to the august bottom. below that is the 1410 level, and the last-ditch defense of 1378.

some new ideas for me to research -- herding sentiment and buy/sell climaxes.

UPDATE: a new low in this leg down to s&p 1416 today with vicious closing action -- but we finally get a nonconfirm from 65-day new lows to pair with our now-weeks-old tapering of 20-day new lows and nh-nl.

it may not turn up right now, and there could be more downside, but chances are that the vast bulk of the losses are behind. i was early to the long side, but i think it would be very dangerous to short here.

another notable bit of action -- just 5 of the nasdaq 100 closed over their 30-day moving average. the last time that was true: august 6, 2004. before that: september 17-26, 2001 -- the september 11 reaction low. before that: august 28-september 2, 1998 -- the long-term capital management crisis low. before that: july 11-17, 1996.

these were spectacular buying points. from 8/12/4, the ndx climbed trough-to-peak over 20% in the following four months. following the 9/26/1 low, the ndx climbed 45% in less than three months. from the 8/31/98 low, the ndx jumped 20% in just 16 trading sessions. from the hammer low of 7/16/96, the ndx jumped 17% is less than a month (and jumped 45% over seven months). so there might be a few more days of pain ahead -- but on the other side is likely a sharp rally.

Labels: ,

This page is powered by Blogger. Isn't yours?