Monday, June 1, 2009

Eerie Déjà Vu as VIX and SPX Both Jump More Than 2.5%

If you follow me on Twitter (VIXandMore on Twitter), you probably saw my surprise when I asked rhetorically, “When is the last time the VIX was making new intraday highs with the $DJIA up 230?”

While I can’t answer that question (though my guess is that today was a first), I did check to see if today’s close set a record for greatest percentage gains by both the VIX (+3.87%) and the SPX (+2.58%) on the same day.

It turns out there was one previous instance that topped today’s double spike, back on November 27, 2002, which was the day before Thanksgiving. On that day, the VIX was up 4.93% and the SPX gained 2.80%.

I checked the charts to see how November 27th fit into the 2002-03 bear market lows. As it turns out, November 27th, which I have indicated with a purple arrow in the chart below, was just two days before an intermediate high which preceded a 17.3% drop that led to the final bottom process some 3 ½ months later. From this point, the markets began to rally and were bullish for more than four years.

The chart below recounts the remarkably similar history from 2002. In that year, the bear market associated with the dot com meltdown bottomed first in July and then again in October, where it hit 768. From that October low, the market rallied to 939 on November 27th – numbers that are strikingly similar to the November 2008 low and the subsequent January 2009 high.

In 2003, the SPX put in a third and final bottom of 788.90 on March 12th, before stocks rallied and never looked back.

The question of the day is whether the current market, like November 27, 2002, still has (at least) one more sharp drop in it…or does the current situation bear a closer resemblance to the May 2003 beginning of a new bull market?

For the record, when I relaxed the conditions on the simultaneous VIX and SPX jumps, the day that comes next closest to today and November 27 is May 11, 1990. At that time, stocks were in a bullish uptrend that would continue for another two months, before being interrupted by a 19.8% drop that would last from July to October of that year.

Statistically significant? Of course not. Anecdotally interesting? I think so.

[Finally, if the chart from 2002-03 looks familiar check out Sunday’s Chart of the Week: Emerging Markets. I had to double check to make sure I had not posted the wrong chart.]

[source: StockCharts]

10 comments:

  1. Another interesting aspect in the chart you are showing: look what happened when the SPX touched the 200 MA in March 2003.

    ReplyDelete
  2. Great post. Your blog is definitely now a daily read. Thanks

    ReplyDelete
  3. There's something more to this (i think). I usually consider not only the SPX index, but also the "stability weighted index" computed by dividing SPX:VIX (or the same with NDX:VXN).

    This is a quantity which displays big flat areas (that i interpret as a segmented market consensus: the higher, the more stable) separated by sharp transitions (when operators change their minds). Studying such a ratio shows that the low of march 2009 has no significant meaning as SPX:VIX is higher than at the low of november 2008 (despite SPX being much lower, meaning that VIX is much lower too). Further, we are now in a very slow increase of SPX:VIX which seems to be bounded by a limit around 32.

    Having both SPX and VIX up today allows SPX:VIX to remain steady and (perhaps) repeating what happened in march 2009 (in the opposite direction of course) when prices were decreasing together with the volatility.

    This goes in the direction of anticipating a reversal for the next days to come.

    Excellent blog that you have!
    --
    Laurent

    ReplyDelete
  4. I think this is nothing more than large long accounts rolling up their put hedges from lower strikes to atm, thus pushing up the vix. I saw a number of large short dated put spreads trade today. It feels like we keep seeing this. A rally where the vix doesn't come in as much as you'd expect at first but then over the coming days it drifts down and then has a one day plunge.
    lather, rinse, repeat

    ReplyDelete
  5. I'd like to go back to the definition of VIX. Say on 9th March, VIX was 51.34(H) and S&P was 676.53(C). By definition, there is a 68% of chance that s&p should be over or down by 51.34/sqrt(12) = 15% or S&P = 800 over the next 30days. But on 9th April, SPX was 829.

    It implied it is a black swan again (or outsize one sigma region). In fact, the latest market movement seems quite unpredictable. I wonder if VIX is still a good indicator to the market sentiment anymore? I'd like to know your opinion.

    Man

    ReplyDelete
  6. Something else I noticed on Friday:

    The VIX and VXO moved in opposite directions. The VXO is now at a 180 trading day close low (you have to go back to 11 September 2008 for a lower VXO).

    Bill, do you have any views on VXO and the VIX moving strongly in opposite directions?

    ReplyDelete
  7. Sorry - I meant Monday I think.

    ReplyDelete
  8. Bill

    I have done a bit of analysis on one of my excel models.

    I find 42 instances in the past where both the S&P 500 and the VIX rose by more than 1%.

    The last 6 occurrences of this were in 2003....

    I found that the next day was an up day on 21 occasions (exactly half) and the others all finished down.

    The average increase on the next trading day was 0.06% and the max increase was 2.9% (trading day after 25 Sept 1990) and the maximum fall was 1.59% (trading day after 3 July 2000).

    I then wondered if the high 2 day RSI (that we had yesterday) would make a difference and noticed something immediately. The AVERAGE daily RSI(2) was 90.33%. The lowest RSI(2) was 62.19 - again on 28 Sept 1990.

    So, it seems to me that the VIX and the S&P both going up by over 1 % may be something that is common after a steep rise in the market and fairly unknown at other times.

    Douglas

    ReplyDelete
  9. I also looked at +2% (both).

    Of the 5 times this occurred, 27 Nov 2002 was the only one that was followed up by more down days than up (5 May '97 was almost all up days from +5 days onwards).

    27 Nov 2002 was also the only date where the RSI(2) was less than 90 (at 73) - all of the others (except yesterday) were over 95 (RIS(2)).

    This worries me as I am short the market...

    ReplyDelete
  10. I read VIX from a slightly different angle and am more interested in how VIX could help me in projecting the market in the next trading days.

    If interested, take a look at blog www.blash168trader.blogspot.com/

    ReplyDelete