Comparing Two Mountains (part II) , by Allen Gillespie

October 19, 2011 |

How to quantify similarities between such "mountains" [i.e. price charts] ?

1) Decide trailing periods and criteria to be used - YTD performance > X, last 5 year performance > Y, etc
2) Build universe/database of similar companies for each year
3) Build correlation table to confirm
4) Build composite model
5) Look at forward if-then test

In my experience, the bearish case on high momentum names, frankly any name, is best fundamentally analyzed as a move from Blue Oceans to Red Oceans and along with general market trends. Blue oceans situations tend to be P/E unconstrained, consistent growers, etc http://www.blueoceanstrategy.com/ but once we move into the Porter world of Competitive Strategy then P/E becomes constrained which leads to compression. Generally, there are subtle clues - RIMM announced a move into consumer markets where AAPL played- so the business market was saturated - NFLX CFO left when the stock was below $200 on its way to $300. They started focusing on cost strategies, changing the story from new subscriber adds. I haven't followed GMCR that closely - but is there a competitive threat that is changing the marketplace - are they experiencing a strategy change - that's the key question.

Solar existed on subsidies granted by bankrupt governments, so it has to compete with more economic alternatives. Hence, the president's loan issue.

Stocks have to compete with bonds, so stocks crashed in 1929, 1987, 2000, 2008, etc

EK lost to digital photography.

My worst mistake ever came from Able Labs - a generic drug maker - had 26 NDAs pending, huge margins and a new lab in NJ - problem: small reference to litigation in the SEC filings that later turned out to be because they were getting their margins by diluting the drugs - stock went from new high list to opening down something like 86%, where I sold before watching it go to $0 in 30 days. Subtle clues. They are really important if one is making the bearish case.

in reply to Victor Niederhoffer's comment:

Strange similarity between those two [NFLX and GMCR] to a person who looks at it as
two mountains of different heights with similarly looking crests
relative to the peak.

Query. How would one quantify similarities between such mountains?
And once quantified, what is best way to see the predictive value of
such similarities. I am reminded of the cotton traders most famous
trade. He noted that 1987 looked similar to 1929. then he knew it was
going to have a crash. The drunk man saw the same similarity and started
out long that Monday, and then sold. Between the two of them, they were
enough to trip the portfolio insurance to sell.

Query. How ridiculous can you get without quantifying the two
questions I asked? I say it wasn't that similar to 1929 as compared to
other years. and also that the ones most similar to a given few years of
bearishness, in the past, the less is the relation between past and
present. i.e. no predictive value to start.

Gibbons Burke comments:

There is another model which incorporates a similar gradual buildup with no appreciable change, then catastrophic breakdown, like the straw breaking the camel's back. A simple model is dropping grains of sand onto a surface. A pile builds up. With each grain the pile gets higher and higher, in an orderly fashion and is stable, until the angle of repose gets to a critical point, at which the next grain of sand sets off an avalache. Similar but subtly different. The concept is known as "self-organized criticality", and I suppose it may have some relevance to how bubbles build up and then collapse:

http://en.wikipedia.org/wiki/Self-organized_criticality

Christopher Tucker writes:

See also Slope Stability Analysis Methods:

http://en.wikipedia.org/wiki/Slope_stability#Analysis_methods

A similar criticality phenomenon is Flashover:

(quoting the wiki - http://en.wikipedia.org/wiki/Flashover )

A flashover is the near simultaneous ignition of all combustible material in an enclosed area. When certain materials are heated they undergo thermal decomposition and release flammable gases. Flashover occurs when the majority of surfaces in a space are heated to the autoignition temperature of the flammable gases (see also flash point). Flashover normally occurs at 500 °C (930 °F) or 1,100 °F for ordinary combustibles, and an incident heat flux at floor level of 1.8 Btu/ft²*s (20 kW/m²).[1]

another is Phase Transition: (from http://en.wikipedia.org/wiki/Phase_transition )

A phase transition is the transformation of a thermodynamic system from one phase or state of matter to another.

see also Crystallization: http://en.wikipedia.org/wiki/Crystallization

see also Nucleation http://en.wikipedia.org/wiki/Nucleation

see also Vitrification http://en.wikipedia.org/wiki/Vitrification

Gibbons Burke responds:

I was lucky to be in the right place at the right time to capture a flashover in a fire near my home (in 2006) in New Orleans:

http://web.mac.com/gibbonsb/Site/Blog/Entries/2006/3/13_Portrait_of_a_flashover.html

Stefan Jovanovich comments:

The sad fact is that the firefighter community still has no agreement on how to deal with flashover risk. They have not even settled on the question of whether to use a wide fog or straight stream!!!!!

http://www.fireengineering.com/articles/print/volume-157/issue-6/features/flashover-risk-management.html

The best teacher I ever had (an instructor at the Navy's Damage Control School in Philadelphia), said that the Navy were the only firefighters who had figured out how to do something besides spray and pray - i.e. use foam to suffocate fires and inert gases to secure the fuel lines - and even so there was a fatal tendency to believe that all you needed to do was get a big enough bucket. He pointed out to the class that the greatest risk of the Forrestal fire turned out to be the water from the firefighting itself, which almost capsized the ship and washed away the retardant foam.

http://en.wikipedia.org/wiki/1967_USS_Forrestal_fire

Comments

WordPress database error: [Table './dailyspeculations_com_@002d_dailywordpress/wp_comments' is marked as crashed and last (automatic?) repair failed]
SELECT * FROM wp_comments WHERE comment_post_ID = '6875' AND comment_approved = '1' ORDER BY comment_date

Recent Posts

List of Authors

Oct

19