Why Do Analogs Work? -- Part 2

“The 1101 touch occurred 69 sessions after the May 2 high.  In 2008, the equivalent bottom was 70 sessions after the Oct 11, 2007 high.  The 2011 decline amounted to 19.6% from the high; in 2007 it was 19.4%.    The 2007/8 total point loss of 306 points was very closely matched by 2011′s eventual loss of 296 points.  And, the eventual low in 2011 came 108 sessions after the top; whereas, 2008′s came 107 sessions after the top.”

********

I was compelled to create the pebblewriter blog on May 2, 2011 because I was convinced we were nearing a top.  I knew little about analogs, but I knew chart patterns and was learning about harmonics.  So, I posted:

“A very long term support line comes into play as resistance.  About to intersect with the rising wedge from March ’09 (daily) and the rising wedge from this past April 15 (hourly.)  Then, there’s the .786 retracement off the March ’09 lows coming up at 1381.50.”

And followed up a few hours later, adding:

“Price target if the longer term wedge plays out is 46 – 100% of the rise, indicating 320-700 points on the SPX.”

Of course, I had no idea when I wrote those words that the top had just occurred an hour earlier.  And, I was dead wrong about the minimum loss coming.  Instead of 320 points, SPX fell only 269 points (oh, well.)

The analog came to me as the result of my obsession with patterns.  The basic chronology of my observations (with links to earlier posts) was as follows:

On May 31 [Why P[3] is My Top Bear Count] I noted

 “significant similarities between the past few months and the tops in 2000 and 2007.”

On June 3 [Here We Go] compared the market action between the patterns as related to breaking and backtesting the long term support trendline.

On June 8 [Deja Vu?] detailed the pattern that the 2000, 2007 and 2011 tops had in common, noting how the market…

“…once it drops below its long-term support, looks like it’s in for a free fall.  But, it eventually finds new support in a parallel channel as I theorized a few days ago.  The bottom of the channel is drawn off a recent major high and the top is drawn off the two most recent peaks.  It pencils in nicely for 2000 and 2007…”

On June 10 with SPX below 1270 and amidst warnings of “flash crash!”, [Channel Surfing] showed how the pattern (accurately) promised one last return to the midline of the regression channel at around 1330.  I showed how the channel I’d drawn corresponded well with a 2-standard deviation regression channel.

On June 12 [Update: Channel Surfing] reported that almost every major top since 1928 exhibited the same pattern.  I suggested that the pattern was not only characteristic of tops, but a requirement.  I formalized the description as:

“…characterized by a multi-month pattern within a rising market that has at least two significant touches (of the index or its Bollinger Band) of at least 1.5 standard deviations on the upper and lower extremes of a regression channel commencing after a post-correction new high.  It’s capped off by a third touch on the lower boundary and subsequent return to at least the midline before a final plunge to new lows."

On June 16, [You've Got a Fan in Me] showed how the regression channel could also be defined in terms of fan lines from significant previous high and low pivot points.  I used examples from 2007, 2000 and 1937 to demonstrate.  The prior day’s post [Playing the Bounce] used the pattern to call the bottom at 1261.90 (it was the next day at 1258.07.)

On June 21 with SPX at 1294, [A Different Perspective] discussed how a return to the pattern midline around 1327 and subsequent fall would create a huge head and shoulders pattern that, by mid-August, would begin a decline to around 1200.

On June 23 [Deja Vu, All Over Again] showed how the 2000 and 2007 patterns corresponded, beat for beat, and charted the various trend lines that would govern the pattern’s completion.  With SPX at 1287, the pattern still indicated a 1320-1330 target.  I reiterated the pattern prediction the following day in the midst of the 25 point intra-day plunge triggered by the strategic petro reserve decision.

On June 26, [Cliff Diving] defined the decline from 1370 to 1258 as 1 of (1) of P[3], and the subsequent rise a corrective wave 2.  I suggested the pattern interim target might also follow Fibonacci guidelines.  The midline, at 1322, was virtually the same as the .786 Fibonacci retracement line.  With the market back to 1268 and bearish sentiment through the roof, I suggested the pattern would take us up 54 points in 5 days.  It took 4.  As I suggested on June 29, we would likely even overshoot (we did, hitting instead the .886 Fibonacci retracement.)

In June 29th’s post [Lunatic] I restated that contrary to popular opinion, the rally would not lead to new highs, but would merely complete a corrective wave 2, followed by wave [i] of 3 to around 1300.

On July 4, [Final Destination] suggested the next move down would stop north of 1298.61 in order to keep the bullish count alive.

On July 7, [Confidence Fairies] and [Then and Now] suggested we had reached the pattern high at 1356.87 right at the .886 Fibonacci level.   The next day began a 5-day, 46-point decline.  In [Friday the Bear Came Early], I suggested the peak was in — earlier than the 87-day cycle would indicate because that’s what happens in market tops.

When a big decline the next morning was quickly reversed and seemingly everyone turned bullish [She's Come Undone]  explained this was simply a throw-over, and that the pattern 1-standard deviation line had stepped in for 2007′s channel midline in defining the top and preventing any further advances.  The McClellan Oscillator gave a warning.

On July 11 [The Deathly Hallows] with SPX off 24 points at 1319, I theorized that the -1 standard deviation channel line at 1299 would catch the fall — supported by the H&S target, a key Fibonacci retracement and fan lines.

On July 15, with SPX at 1316, harmonics indicated an imminent low of 1300 and subsequent rise to 1345 [The Waiting Game.]  The actual low the following day was 1295, and the subsequent rise was to 1347.

On July 20, I posted about an inverse H&S pattern forming that would signal a huge updraft, but felt we would end up 10 points short — as happened in 2007 [Ten Lousy Points.]  In reality, we would end up 8 points shy.

The next several posts were simply warnings of the impending fall.  [Merry Christmas] warned that July 21 was the equivalent of December 24, 2007.  [Pulling the Trigger] on the 21st correctly advised that…

“1347 might be the last best chance at an excellent short.”

[All Aboard] on the 26th gave visual clues for those who don’t like to read.  And, on July 26, with SPX closing at 1331, [Happy New Year] declared that the 27th would be the equivalent of December 31, 2007.

“Those who have been following this blog for any length of time know why I’m wishing you a Happy New Year in July.  Under the 2011 is 2007 theory, tomorrow is the equivalent of December 31, 2007.”

SPX was off 27 points the following day, and ten sessions later, bottomed at 1101.  It was  down 269 points from the day I felt moved to post that it looked like the market was topping in a way that was similar to 2007.

The 1101 touch occurred 69 sessions after the May 2 high.  In 2008, the equivalent bottom was 70 sessions after the Oct 11, 2007 high.  The 2011 decline amounted to 19.6% from the high; in 2007 it was 19.4%.    The 2007/8 total point loss of 306 points was very closely matched by 2011′s eventual loss of 296 points.  And, the eventual low in 2011 came 108 sessions after the top; whereas, 2008′s came 107 sessions after the top.

**********

Did the analog play out precisely as did 2007?  I’m too lazy to tally all the data points and calculate it; but, just spitballing, I’d say it was about a 95% correlation.  We could point to the damage done by Fukushima, the US downgrade, the troubles in the Euro Zone, S&P’s GDP revision, etc., but how could we possibly explain those events falling on the necessary dates in order for the 2007 pattern to repeat so precisely?  Exactly.

When all is said and done, I believe the analog played out as scripted because the same conditions and patterns were in place as in 2007/8.  The rising wedge, the harmonic patterns, the various H&S patterns, etc. worked the way they were supposed to.  And, people being people, their reactions to rising/falling markets was the same each time.  Fear and greed will always be with us.

At any time, a black swan event might have come along and knocked the analog off track.   Equally effective are government and industry players that share a vested interest in a steadily increasing DJIA.  A strategically placed leak or massaged jobs report can work wonders when a pattern completion is on the line.  I have no doubt this is what happened last October.

As I watch some of the current analogs play out, I’m wary of the countless curve balls that might come our way: downgradings, upgradings, wars (trade and/or shooting), earnings surprises, elections and the ever-present threat of more QE.  All we can do is keep a watchful eye, try to avoid bias and use stops.

Analogs are a process.   They take time and hours of study -- as my daughter whose basketball game I missed today will tell you (hopefully, she’ll forgive me when it comes time to write college tuition checks.)  When done correctly, we won’t discover nearly-complete analogs days from paying off.  We’ll find them weeks or months in advance — seeing similar patterns setting up, watching for divergence.

The ones I’m watching now look very good — including the one posted last week.  In the days ahead, we’ll see some that work, and others that don’t.  But, in an investment environment where so many forces are working to manipulate the markets for their own financial and political gain, I think analogs represent an excellent way to make money.

Good luck to all.