$$ AAII Sentiment: Wildly Bullish

Bullish sentiment hit 51% this week in the AAII survey. It is the highest bullish sentiment of the year, topping the January 14 (47%) and April 15 (48%) readings.  

What happened earlier this year after bullish sentiment made a yearly high?  The S&P topped on January 19, falling 100 points in the next 13 trading days.  And the S&P topped on April 26, plunging 155 points over the next 7 sessions.

The spread between bullish and bearish sentiment is 27%, the highest reading since May 8, 2008.  What happened after May 8, 2008?  The S&P topped on May 19 at 1440 before losing 240 points in the next 38 trading days.  

Be careful out there.  

$$ Fed Funds Target Rate & AAII Stock Screens

Every few years, the Federal Reserve changes direction on its Fed Funds target rate.  I classify a change of direction as a 50 basis point move.  It usually happens over two 25 basis point moves, but occasionally the Fed will move 50 basis points all at once.  Here's the history over the last ~15 years:

December 19, 1995:  Target rate coming down. 
August 24, 1999:  Target rate going up. 
January 31, 2001:  Target rate coming down. 
August 10, 2004:  Target rate going up. 
September 18, 2007:  Target rate coming down. 

The American Association of Individual Investors (AAII) has tracked the performance of a variety of stock screens since 1998.  Theorizing that some companies and therefore AAII screens do better in times of cheap money, while other companies/AAII screens outperform in times of money tightening, I wondered whether the change of direction in the Fed Funds target rate could be a useful signal in choosing between the AAII stock screens over the last 13 years.  

AAII member Todd (screen name "tschoepflin") came to the rescue.  What he found was quite remarkable:  using the Piotroski 9 screen during periods when the Fed Funds target rate was coming down, and switching to the Kirkpatrick Value screen once the Fed Funds target rate started to climb produced annualized compound returns in excess of 48% from 1998-2009!  And you'd be up another 158% from January 1, 2010 through April 30 2010! 

Switching between Piotroski 9 and Kirkpatrick Value significantly outperformed sticking to one screen or the other exclusively.  From 1998-2009, Piotroski 9 returned ~28% a year (one of the top 4 performing screens), while Kirkpatrick Value returned ~19% per annum.  Further, during periods of money tightening, Piotroski 9 returned only 15.7% cumulative from September 1, 1999 - January 31, 2001, and lost money between September 1, 2004 and September 30, 2007.  During periods of money easing, Kirkpatrick Value essentially broke even between February 1, 2001 and August 31, 2004, and has lost money since October 1, 2007.

This performance was measured using AAII's methodology, which is a monthly re-balancing model with 100% re-invested in all passing companies each month.  If only one company passed the screen, you'd be 100% allocated in that one stock for the month.  If no companies passed the screen, you'd sit the month out.  No stop loss orders are used.   

To use the Fed Funds target rate change of direction as a signal, you would simply switch screens the month after a Fed Funds target rate change of direction.  Here's the yearly performance breakdown, with the S&P cited for comparison:

1998

Piotroski 9:  17.9%

S&P:   26.7%

1999

Piotroski 9/KV:  26.6%

S&P:  19.5%

2000

Kirkpatrick Value:  63.9%

S&P:  (10.1%)

2001

KV/Piotroski 9:  48.3%

S&P:  (13.0%)

2002

Piotroski 9:  (15.9%)

S&P:  (23.4%)

2003

Piotroski 9:  154.6%

S&P:  26.4%

2004

Piotroski 9/KV:  91.3%

S&P:  9.0%

2005

Kirkpatrick Value:  88.2%

S&P:  3.0%

2006

Kirkpatrick Value:  18.5%

S&P:  13.6%

2007

KV/Piotroski 9:  44.2%

S&P:  3.5%

2008

Piotroski 9:  32.6%

S&P:  (38.5%)

2009

Piotroski 9:  78.3%

S&P:  23.5%

2010 ending 4/30/10

Piotroski 9:  158.3%

S&P:  6.4%

Switching between Piotroski 9 and Kirkpatrick Value using the Fed Fund target rate change of direction as a signal beat the S&P every year except 1998, and most years it crushed the S&P by a net of at least 40%.  The S&P returned a compounded ~1% from 1998-2009.  Also, Piotroksi 9/KV only had one losing year  in thirteen (2002). 

With taxes, slippage and transaction costs, it's impossible to duplicate these results in the real world.  Especially in some of the thinly traded stocks that pass through these screens.  But thanks to Todd for giving us something the chew on if the Fed ever decides to raise the target rate again.

$$ Other Sentiment Indicators

Besides the put-call ratio, I look at a couple of other sentiment indicators.  The American Association of Individual Investors (AAII) polls its members once a week and asks whether they are bullish, bearish or neutral on the stock market for the next six months.  

Like the put-call ratio, the AAII survey is useful for identifying extreme greed at tops and fear at bottoms.  Consider that on August 21, 1987, 66% of AAII members were bullish while only 6% were bearish.   Or January 6, 2000 when 75% were bullish to 13% bearish.  Contrast March 5, 2009 when 70% were bearish and 19% were bullish.

I also subscribe to Market Vane.  They publish the "Bullish Consensus" which is compiled daily by tracking the buy and sell recommendations of leading market advisers and commodity trading advisers.  As such, the Bullish Consensus tracks "market sentiment" as opposed to public sentiment.  Market Vane covers 36 futures markets, including:

Gold

Silver

Platinum

Copper

S&P 500

Nasdaq 100

Light Crude Oil

Natural Gas

Treasury Bonds

10-Year Treasury Notes

US Dollar Index

The markets are ranked from 1% to 100% in terms of bullishness.  Like AAII and the put-call ratio, I look at extreme bullishness and bearishness on Market Vane.  For example, on April 29 the NASDAQ was rated just 3% less bullish than its two-year high Bullish Consensus.  QID (the double inverse NASDAQ ETF) closed on April 29 at $15.35.  By May 6, QID traded over $20 a share.

I also use traditional divergence analysis on the Market Vane Bullish Consensus charts.  For example, if price is trending higher but Bullish Consensus cannot make higher highs, this is an indication that a correction is due.