Consensus forecasting? Can one trust the 'wisdom of crowds'...

The aggregation of information in groups can result in better decisions than could have been made by any single member of the group. So argues James Surowiecki in 2004 bestseller The Wisdom of Crowds, drawing in part on the early work of Sir Francis Galton. Surowiecki's work has inspired three Oxford University graduates to apply its premise to financial markets. A Sunday Telegraph story, How to profit from the noise of the rabble, informs us:

Stephan Bisse, 40, a former City trader at Goldman Sachs who completed an MBA at the university's Saïd Business School last year, has set up ConsensusView.com, helped by two young computer science graduates from Worcester College, Ed Smith and Sam Brightman.

Each day members of the public are invited to vote on whether they think any one of a series of share prices, indexes, currencies and financial futures contracts will rise or fall that day. Voters do not specify an actual forecast; they simply predict a rise or fall, and have to do so between 4.30am in the UK and the opening of the market they are voting on. There is the lure of a $1,000 (£550) prize for the best forecaster each month. The Oxford three hope to test the theory that crowds have wisdom, and if it holds water, create a business out of it.

The website is up and running, but has yet to reach critical mass. As the Telegraph story observes:

The key to all these efforts, however, will be the site's ability to attract voters and, ultimately, to forecast accurately the direction of markets. At the moment, two weeks after its launch, just 52 people are voting on it, but with a larger number the theory that a consensus view will tend to be right will be tested more comprehensively.

As of Sunday evening there were 212 registered users so far, so clearly the newsaper story has boosted the numbers. I encourage readers of this blog to sign up to ConsensusView.com; it's an experiment worth encouraging.

Comments

[Rob Hayward]

Why vote there when you can vote here http://www.igindex.co.uk/

[gary lammert]

As an alternative to consensus forecasting...

The Final Few Lower High Days of Distribution

This Friday will be day 104 completing an ideal daily growth fractal
sequence of 52/130/104
(x/2.5x/2x) dating from August 2004. This sequence represents the
completion of a third fractal of a larger weekly growth fractal
sequence beginning in the early spring of 2003. It is still possible
that some or all of these next trading days from Monday through Friday
of this week will represent the final distribution days where the
final lower high saturation buying occurs, where the final euphoric
traders buy from the willingly-selling economic realists. The
realists, who have made up the smarter half of the recent trades, are
surveying the economic landscape and appreciate the awful ledge that
the wary, as judged by recent sentiment, American consumer finds
himself and herself on.

The six major economic parameters that result in money growth and
contraction and hence valuation fractal growth and contraction are:
assets and their relative valuations; ongoing wages (including new
wages funded by governmental planned and unplanned spending, the
latter of which Katrina is an example); total debt, including pension
obligations and entitlements; cash and savings; prevalent interest
rates; and amplifying(or contracting) lending practices. These major
macroeconomic factors have, within limits, a degree of fluidity and
can and do change the lengths of asset valuation fractal growth which
are directly dependent upon them. But, unless nearly everyone is on
the governmental payroll - negating the realities of a private
enterprise and a real market economy where private business ownership
requires a profit and owners cannot arbitrarily double, triple,and
quadruple wages - boundaries to unlimited growth do exist. With a
significant private sector economy , the capacity for the government
to hyper inflate and cause asset appreciation via governmental
printing presses to increase base wages of all of its citizenry is at
least transiently cumbersome and becomes a less likely immediate
scenario.

Annual money growth represents such a very very small fraction of the
totality of value of - cumulative value of all noncash or cash
equivalent assets, massive collective debt and entitlements as defined
above, ongoing wages, and cash and cash equivalent reserves. The near
perfection in the evolution of the valuation growth fractals dating
from early 2003 suggests both - that relative to the Mount Everest of
cumulative worth of the preceding four entities- the Federal Reserve's
and government's ability to control the economy via money growth is
significantly limited and ultimately unsustainable and that the US
second fractal credit cycle(with a base of about 70 years ending in
1858), now in its 147 year, is near its nonlinear and early end of a
potentially 35 year range: 140-175 years, ideally ranging from
1998-2033.

Since early 2003 the weekly growth fractal sequence has been:

1st fractal (x) 22 weeks
2nd fractal (2.5x) 54 weeks (a little short of the 55 ideal
but ending with a hallmark nonlinear gap lower.)
3rd fractal (2X-2.5X). This fractal consists of a 11/27-28/21-22
of (currently) 22 to 23 week period.

This is the 59th week of the third weekly growth fractal. Notice the
last third and final fractal goes beyond the ideal 2.5x or 54-55
weekly limit, but is within its own idealized y/2.5y/2y weekly limit.
The lowering of the fed funds rates and the new lending practices
with no principal payments, 40 year loans, and large amount of
adjustable mortgages have had an identifiable effect on money growth
and hence fractal evolution; albeit a modest one. The new weekly
growth pattern with a base of 11 weeks starting in August 2004
captured and exactly represents the ongoing integration of the six
determining economic elements resulting in a consistent but very
slightly changed (extended length) fractal pattern of the terminal
third growth fractal.

This combination of initially lowered fed fund rates with most
probably the even greater factor of imprudently unregulated lending
practices was the primary cause of the US and world economic growth,
money expansion, and debt expansion - ballooning the
valuation(overvaluation relative to traditional wages) of the high
tech replacement bubble asset represented by the housing industry.
Credit from overvaluation in the housing bubble has been greatly
amplified by the secondary effects of the second mortgage and
refinance industry luring the ready American consumer to further use
this mirage credit and further overextend personal debt in the
purchase of low cost Chinese goods( whose recirculated dollars have
been invested symbiotically in US government debt), inefficient SUV's
and trucks, furniture, durable goods, vacations, college educations,
et.al. The baited American consumer has been the debtor of last
resort, who has been primarily responsible for increasing global
economic activity after the NASDAQ implosion and for continuing the
eking out of a possible small real US GDP - if governmental inflation
figures are accepted as valid.

While this changing and integrative money growth activity has
slightly deformed (lengthened) the ideal long term weekly growth
fractals dating from early 2003, as of Monday 19 September 2005 the
Wilshire 5000, the largest of the world's equity indices has not
exceeded its 3 August 2005 summit and is still contained within the
ideal 22/54/54-55 maximal weekly growth fractal pattern. Possible
terminal saturation buying this week will determine if there has been
enough money expansion in the last 28 weeks for the Wilshire to best
its 3 August peak.

The first daily decay fractal which includes the Wilshire 3 August
2005 top now appears to be 13-15 days (vice 11 days)in length with a
maximum 3/6-7/7 daily pattern. The second decay fractal should
ideally contain all the trading valuations above the slope line drawn
from the first day to the last day of the second daily decay fractal.
While the originally proposed 11 day fractal decay base (first
fractal)did end at a lower point at the expected 27-28 day range, the
slope line containing the first and 28th day of this possible second
decay fractal was not under the inclusive set of the 28 days. The
idealized three sequential dacay fractal pattern is x/2.5/1.6x-2x-2.5.

The coming major nonlinear valuation readjustment will both occur in
a defined fractal manner and decay pattern and will ultimately reorder
the six economic parameters into a new beginning growth fractal.
Because of the severe imbalances contained within the internal
compositions of each of the six primary determinant economic elements,
the realignment will be necessarily severe.
Gary Lammert

[Mark T]

errr.........follow that. I was going to say that there seems to be a difference between asking people what they would do with their money and actually making them do it, which is why IG index may be a better bet (literally) on the wisdom of crowds.After all the whole science of behavioural finance is based on the fact that when real money comes into it people's risk tolerance changes. What that means in terms of fractals I have literally no idea.

[peter vermeulen]

Thank you very much for catching my attention for this site, it was unexpectedly easy to win the monthly prize in October, November and maybe also in December.
It was a pity I didn't know your name, so I could not say you were my referrer at the consensusview website.
(so you missed already $250 two times).
Regards
Peter Vermeulen