ExtZy 2012 The World In Review 

Professor Michael Mainelli & Xueyi Jiang

March 2013


ExtZy is a set of prediction markets which Z/Yen run and fund in order to learn about markets.  The “Countries” Market is one where players buy and sell shares which pay dividends based on a country’s web mentions, as calculated by search engines.  Each share has a price and dividend earnings.  Thus the price/earnings ratio (P/E) of countries in such a market is an indicator of whether or not players predict a higher than normal rate of mentions (high P/E) looking forward, or a lower than normal rate of mentions (low P/E) looking forward. 

The Price-Dividend Graph from ExtZy - http://www.extzy.net/index.php?option=com_extzy&task=PEgraph&Itemid=141 - shows low P/E countries on the bottom left and high P/E countries on the top right.  On the right, there are positive dividends (In 2012, Extzy has introduced the new algorithms, including the idea of negative dividend.  Players holding shares with a P/E ratio below the median P/E (get negative dividends) lose Zoints from their accounts).  As shown in Figure 1, Iran has the highest P/E, indicating that people feel it will grow in importance.













Figure 1. Price-Dividend Graph

A brief review of the predictiveness of the market for 2012 seemed worthwhile to share with the players.

Countries Market 2012

During 2012 the ExtZy Countries Market was slightly less active than in previous years.  There are 245 registered countries and 90 traders.

  Total Trades  

   Total Volume  





422 38165 90 90.44 1.72 4.69

The biggest movements in prices indicate the biggest changes in opinion over 2012:


Rank Country Name Max Price Min Price Difference
1 Iran; Islamic Republic of 15 1.74 13.26
2 Bouvet Island 12.75 0.16 12.59
3 Czech Republic 14 1.51 12.49
4 Turks and Caicos Islands 12.99 0.84 12.15
5 India 13.9 2.5 11.4
6 Nigeria 10 1.64 8.36
7 South Africa 9.99 2 7.99
8 China 12.76 4.96 7.8
9 Somalia 8 1.53 6.47
10 Brazil 9.6 3.6 6

as well as the historic volatility indicating countries on which opinion changes most often:


Rank Country Name SD                  Historic Volatility
1 Palestinian Territory; Occupied   0.668335 4.819431
2 Pitcairn   0.638622 4.60517
3 United States Minor Outlying Islands 0.638622 4.60517
4 Bouvet Island   0.607135 4.378113
5 Faroe Islands   0.598477 4.315683
6 Saint Lucia   0.497713 3.589059
7 Tuvalu   0.489656 3.530959
8 Iran; Islamic Republic of   0.443198 3.195948
9 French Southern Territories   0.418584 3.01845
10 Gabon   0.416688 3.004782

Snapshot – United Kingdom and the Olympics

 A specific example of the market working predictively, is the United Kingdom, below:


 The chart above contrasts the price and P/E with Google Trends.   Players had been buying in 2011 and early 2012 in anticipation of more Olympic news.  The chart shows two things, (a) players did sell quickly around 1 July in anticipation of web popularity falling, however, (b) seemed to be unable to shift more stock afterwards.  Thus the P/E ratio climbed while popularity tailed off. 

Snapshot – USA and the Presidential Election

Again, players had bought, where possible (the USA is not a very liquid share), in anticipation of the election. 


Prices correctly spiked and held during the run of increased election news.

Snapshot – China and Leadership Change

China is a sobering reminder that our market players don’t always get it right. Anticipating a fall in news, the price drops in March and April, only to rise again, close to its historic level.  None of this is reflected in a significant change in news volumes.


Looking Ahead

The world is a busy place.  ExtZy players get their news from similar sources and are unlikely to have private knowledge.  Even if they had private knowledge of major events, it would be immodest to suggest that they run to ExtZy to trade.  Other markets where the players had moderate predictive success as a group in determining new levels were Spain (banking crisis), Greece (economic crisis fatigue) and Egypt (regime change).  The Countries Market does seem to be able to determine when news will rise to a new level and ‘hold’, or drop to a new level.  We look forward to looking back in 2014.


Prediction Market: A Study Of Future Country Interest And The Market Behavior

Xueyi Jiang

2011 August

University College London

ExtZy 2013 The World In Review

Xueyi Jiang

2012 March