Friday, June 01, 2007

Wall of Worry: Are Traders Getting More Bearish As The Market Acts More Bullish?


Rennie Yang, who authors the excellent Market Tells service, notes an interesting pattern. Despite the recent market rise, investors in the AAII survey are 33% bulls and 45% bears. He notes bullish implications when we have a surplus of bears in this survey during a rising market period. That's the classic wall of worry that bull markets are said to climb.

Above we see a pattern that seems to confirm Rennie's observation. QID enables traders to place double-sized bets on a market decline in the NASDAQ 100 Index. As such, it is an excellent speculative vehicle for traders. Note in the chart above that, as QID has declined (i.e., the NASDAQ 100 has risen), the volume in QID has exploded.

In other words, traders are making more and larger bets against the NASDAQ as that market grinds higher. That's more than a wall of worry; more like a wall of pessimism!

Interestingly, since July, 2006 (N = 217; when my price data for QID begins) when QID is down on the day (rising NASDAQ) and QID volume is up for the day (N = 67), the next day in QID averages a loss of -.51% (41 down, 26 up). That is much weaker than the average loss of -.07% (87 down, 63 up) for the remainder of the sample.

It appears that, at least in the near term, a rising NASDAQ (falling QID) with rising bets on QID has been bullish, as the market confounds the pessimists.

RELATED POST:

NASDAQ Volume as a Sentiment Measure

11 comments:

J Arp said...

You don't happen to have such a chart for the year 2000, do you? If memory serves correctly, AAII posted bearish individuals (55%+) starting in March of that year. I think its worth looking at longer time series.

I do not have the data for that year, so i don't know if there was capitulation in subsequent months of that year.

Thanks

xyz said...

I really don't know how valuable this info is as a lot of hedge funds maybe using QID to hedge long positions.

Brett Steenbarger, Ph.D. said...

Hi,

I'll look into a longer-term perspective on the AAII survey. My past research has shown interesting results when considering both survey results and put/call ratios.

Brett

Brett Steenbarger, Ph.D. said...

Hi XYZ,

I agree; some of that QID volume is likely coming from money managers. The fact that they're feeling a need to hedge here, however, may be significant in sentiment terms.

Brett

weightoftheevidence said...

Brett:
The volumes in the QID have increased probably more to popularity but here is a talbe of the 3 month avg vol for the opposing PROSHARES TRUSTS. The top 6 are the ULTRA_LONG and the bottom 6 are the ULTRA_SHORT. The comparative volumes are illogical.
The long QQQQ does only 7% of the short QQQQ.
COMMENTS PLEASE.

SYMBOL Vol (000) Name
DIG 15 ULTRA OIL & GAS
DDM 136 ULTRA DOW30
MVV 85 ULTRA MC400
QLD 1085 ULTRA QQQ
SSO 265 ULTRA S&P500
UWM 80 ULTRA RUS 2000


DUG 33 ULTRASHT OIL & GAS
DXD 925 ULTRSHRT DOW30
MZZ 285 ULTRSHR MC400
QID 14602 UTLRSHRT QQQ
SDS 2475 ULTRSHRT SP500
TWM 352 ULTRASHT RUS2000

weightoftheevidence said...

SYMBOL Vol (000) Name
DIG 15 ULTRA OIL & GAS
DDM 136 ULTRA DOW30
MVV 85 ULTRA MC400
QLD 1085 ULTRA QQQ
SSO 265 ULTRA S&P500
UWM 80 ULTRA RUS 2000


DUG 33 ULTRASHT OIL & GAS
DXD 925 ULTRSHRT DOW30
MZZ 285 ULTRSHR MC400
QID 14602 UTLRSHRT QQQ
SDS 2475 ULTRSHRT SP500
TWM 352 ULTRASHT RUS2000

Here are the comparative volumes of the respective ultra LONG and ultra SHORT PROSHARES. Very bizarre.

Brett Steenbarger, Ph.D. said...

Hi Weight of the Evidence,

It does appear that large traders are using the short instruments as hedges--and are increasingly hedging.

Brett

Brett Steenbarger, Ph.D. said...

COMMENT TO TRADERFEED READERS: A BLOGGER PROBLEM HAS BLOCKED ME FROM POSTING TO THE SITE AS OF 6/2/07. I AM IN THE PROCESS OF TRYING TO GET THIS FIXED. I'VE UPDATED MY PERSONAL SITE, INCLUDING THE TRADING PSYCHOLOGY WEBLOG AND THE TRADER PERFORMANCE PAGE (WWW.BRETTSTEENBARGER.COM). IF THE BLOGGER PROBLEM IS NOT SOLVED TODAY (6/3/07), I WILL PUBLISH MY DAILY POSTS TO THE TRADING PSYCHOLOGY WEBLOG. SORRY FOR ANY INCONVENIENCE.

BRETT

ryang said...

A long-term chart of the AAII Bearish Consensus is available here

While there was one spike over 40% in March '00, notice that this was quickly followed by unusually low readings (below 20%) June through August. By comparison, it's been nearly a year and a half since we last saw a reading under 20%.

In addition to looking at one-week readings, I'd suggest following the 10-week moving average of AAII bears (seen in blue on the chart) as it does a good job of clearly identifying periods of excess optimism/pessimism. Readings over 30% by the average reflect an unusually large number of bearish investors while readings below 20% reflect an unusually small number of bearish small investors. Back in June 2000, we did see the moving average briefly top 30% (a generally positive sign for the market from a contrarian perspective), but it quickly retreated below 20% (signaling overly complacent small investor sentiment) by August of 2000. By comparison, we haven't seen this same type of small investor complacency a single time over the past three years.

Rennie

Brett Steenbarger, Ph.D. said...

Great observations, Rennie; thanks so much for the comment and link, and thanks for your great work with Market Tells.

Brett

Marc said...

Hi Brett,

I think using volume changes in new index like the QID isn't going to give much information. Perhaps a comparison of volume of traditional shorting instruments would be more useful. I would be interested in seeing this.

Thanks,
Marc