With an Eye to the Future...

This blog is interested in examining some of the undesirable notions that exist within finance, namely, bubbles and panics. On top of this however it has been hoping to explore the opportunities that are available in testing virtual economies to draw lessons about the real world.

Eyjolfur Guomundsson, the head economist of Eve Online has repeatedly stated that if he was to retire he would have enough data for ten years research. This data is detailed, accurate and complete to an astounding detail in comparison to real world data. UDP (Universal Domestic Product) in Eve is never estimated or revised but simply known. On top of quantitative work he is interested in exploring more subjective ideas. Senior managers who play the game often comment on the skills required to co-ordinate hundreds of players of different nationalities and abilities to complete massive projects. For example, building a single “Titan” starship will take six months if there are no delays. Most corporations within the game also employ accounting methodology to examine how best to invest funds, namely through the analysis of projected cash flows.

Yanis Varoufakis, head economist of Valve gaming company, explained his motivation in researching the virtual economy came from a frustration with modern econometrics. He was frustrated with the inability of scientific rigour to disprove absurd statements such as “Christmas is explained by a prior increase in the demand for toys”. His reasoning behind the inability to run experiments on the macro-economy or even the micro-economy and “assumptions” of ceteris paribus were simply wishful thinking. In Yanis’s announcement on joining Valve he neatly summed up his opportunity by saying:
“In sharp contrast to our incapacity to perform truly scientific tests in “normal” economic settings, Valve’s digital economies are a marvellous test-bed for meaningful experimentation. Not only do we have a full-information set (making sampling superfluous) but, more importantly, we can change the economy’s underlying values, rules and settings, and then sit back to observe how the community responds, how relative prices change and the new behavioural patterns that evolve. An economist’s paradise indeed…”

We must though be cautious before we burn all the data sets of the real economy and begin employing everyone in the virtual world. Economies like that of Eve make reasonable testing places because of relatively large populations(400,000), with relatively complicated economies and where goals are broadly similar to that of the real world. We can quickly see how this breaks down when applied to other games. The best known online game (and one which the author has “invested” a few thousand hours) is World of Warcraft, a world with 12,000,000 players. The above preconditions do not hold in this seemingly tantalising data set. For one, the world is in fact split into hundreds of servers so that many players operate in an economy of a few hundred rather than millions. While trade and work exists in World of Warcraft it is definitely a secondary feature to the more popular combative element of the game. Motivations also differ wildly; raw materials are valued far more than the finished products they produce, a system where in the real world we would see a baker lose money by turning wheat to bread.  

Another issue is that most games developers are not interested in economists, and in fact are mistrustful of economist’s aims. Developers invest time and millions of dollars in creating games and are unlikely to be pleased to see economists treat their creation as a plaything. In response to this we can see some economists begin to develop their own games in competition and to provide the best testing beds possible. However, what economists may offer in scholarly thought they seem to lack in ability to design games which attract and maintain large populations. Until the goals of developers and economists are aligned, it is unlikely to become common practice for them to work together.

I've tried to bring a balanced view to testing virtual economies and though I am truly hopeful for the future we must proceed with a word of caution. Just as we can struggle to develop accurate pictures due to the complexity of the real world so too we can struggle because of the simplicity of the virtual world. There are some great opportunities available for virtual economies to support research but at this stage there are no calls for virtual research to usurp real world investigations. As games and developer/economists relationships develop we should see a helpful boost to economics and finance as subjects but not a redefining.

Read Yanis’s blog here:

http://tinyurl.com/YanisVValve

Bubbles In Space

Bubbles come in diverse forms and as such we don’t have a universal definition to follow. The economic definition, and one that I broadly agree with, states that Ex post asset prices are viewed as having been over-valued and not based on fundamentals. Between the South Sea, Wall Street and the Dot-Com bubbles economists have an unfortunately large amount of data to study. We can also find evidence of bubbles in virtual economies, in this case Eve Online.

The asset being traded in the image below is Helium Isotopes. In Eve players use Isotopes as fuel for stations which are used to operate extraction or production facilities. Helium is just one type of isotope but the other three isotopes endured price bubbles similar to the one below over the same period.

The price moved from around 500 ISK (Interstellar Credits) to approximately 1300 ISK over two and a half months before collapsing to 400 ISK over the next two months. These are wild fluctuations with the price doubling and then more than halving over a period of a little under 6 months. The motivation seems to have come from a reaction to announcements of new ships being added to the game. People reacted to this by raising expectations of the value of the materials used in production which is a rational step to take. However, Eve has an active array of investors and speculators who engage in trading. It seems that these players attempted to ride the wave of the bubble and many of the more established traders did indeed profit. Eventually though as understanding about the true impact of the new technology (new ships) grew people could reset their expectations and the price returned to reflect the fundamentals.

One of the common queries that surround bubbles is whether we should or are capable of bursting bubbles. Eyjolfur Guomundsson, head economist with CCP the company behind Eve Online, is attributed with taking action against another in-game asset experiencing a bubble. It is possible to buy a subscription for a month using the in-game currency. CCP was worried that this price was rising too fast and was being affected by speculation and so they offered a sale on the cost of subscription with real currency, effectively reducing the value of the in-game subscription. This method has been attempted repeatedly now with increasingly marginal impact and price appears to be much stickier than Eco 101 would suggest.

With America in 1929 there were attempts to head off the bubble and reduce the build-up of problems. Like most questions worth asking in economics and finance though we don’t have a conclusive answer. Perhaps through testing virtual environments we can see whether Friedman and Schwartz and their views on monetary policy were accurate in explaining the great depression.

For those interested in the wider Eve economy there is a 45 minute presentation delivered at the last Eve-Con from Eyjolfur Guomundsson describing the changes in 2012.