Nomura’s prediction: France vs. Spain in the final
Its approach: The Japanese bank opted to treat the tournament with the same techniques it would use to offer balanced risk to investors:
The results: Its analysis suggests that France, Spain and Brazil are the most likely to reach the semifinals, and it predicts that France and Spain will meet in the final
ING’s prediction: Spain
Its approach: The Dutch bank used one of the more unusual techniques described here, opting to calculate the chances of success using a measure based on market value of the nation’s team and its previous performance. (The assumption here being that value and success are closely correlated.) A team’s worth was calculated from individual player transfer value estimates and their track record from FIFA world rankings
The results: Spain will be crowned world champion, with a total team value of €1.04 billion ($1.16 billion). A close second: France, which is valued at €1.03 billion.
Goldman’s prediction: Brazil
Its approach: Goldman Sachs’s global macro research team opted to use artificial intelligence algorithms to conduct its analysis. Here’s its approach:
The result: An unnervingly algorithmic score in the final of 1.70 goals for Brazil to just 1.41 for Germany.

Bonus: Goldman has predicted a Brazilian victory for the last three World Cups, and has been wrong every time. (Sadly, Paul the Octopus is no longer alive to predict the results.)
UBS’s prediction: Germany
Its approach: A group from the Swiss bank’s chief investment office scored each team based on an objective skill-level measurement called Elo rating, its route through qualification and any home nation advantage. The analysts then carried out statistical modeling known as Monte Carlo simulation.
The result: Germany, Brazil and Spain are most likely to win — with Germany the UBS favorite, with a 24 percent chance of victory

culled nyt