Battiston, Stefano, Dr.

ETH-Zentrum
Dr. Stefano Battiston
Systemgestaltung
F 27.2
Kreuzplatz 5
8032 Zürich
SWITZERLAND

Phone: +41 44 632 84 64
Fax: +41 44 632 18 80
E-Mail: 

Recent and News

2011/10/31: THE CDS BOMB: An article on Bloomberg News by Mark Buchanan reviews our Liaisons Dangereuses paper

2011/10/28: WIDE MEDIA COVERAGE AND DEBATE on the Global Network of Corporate Control just accepted on PLoS-ONE PUBLISHED VERSION OPEN ACCESS --->
RESPONSE IN THE MEDIA - FULL LIST ---> arxiv
FOC, the FET-OPEN project supporting the work.
The article on New Scientist: The capitalist network that runs the world
James's blog summarizing the debate

Because we are flooded with emails and questions: here are few important points:

THE BASICS OF THE METHOD TO ESTIMATE CONTROL
In the simplest scenario we have used, we estimate the control of a shareholder as the fraction of its shares weighted by the value of the company. E.g., if A owns 20% of B, then we say A has a control over 20% of the revenue. This is the basic method. We then extend the computation to take into account longer chains, multiple paths and loops. We used also other more complicated scenarios to test the robustness of the results.

WHAT THE PAPER CLAIMS
The basics of the recipe are simple and the results show that there is a group of 1300 companies, the core, that are all connected to each other by relations of direct or indirect ownership, the shares of which remains for 3/4 within the core.
This fact alone indicate a very high level of interconnectedness. Moreover, altogether, this core controls (in the above meaning of "control") 50% of the total TNC operating revenue in the world (By the way, funds are a tiny minority in the core. The fund issue is a minor issue that diverts the attention from the major facts). Depending on where we cut the crop of the cream, we obtain various figures. The one has circulated on the web is that 147 topholders TNC in the core of the network, control nearly 40% of the TNC value. Overall, we observe high concentration of control in a core of actors connected by high level of mutual control. These are the main results of the study.

WHAT THE PAPER DOES NOT CLAIM
Our study does NOT claim that the actors in the core are colluding. NOR it claims that this structure is the result of some intentional design. We actually think that it probably emerges "naturally" as a result of simple mechanisms that are at work in the market. However, "naturally", does not necessarily means "good" for our economies and societies. What we claim is that further studies are needed to investigate the implications of such a structure, because it is very well possible that it is engenders market competition and financial stability. We do not say it certainly engenders. We say that, based on what is known, we should better check.

The FOC 1st review meeting
The paper on knowledge recombination and R&D networks is accepted.
A new working paper on Diversification and Instability

Upcoming Talks

Research Interests

Systemic Risk in Financial Networks

Systemic Risk is the probability that large fraction of the financial system default or go under distress. Today's financial system consists of many financial intermediaries connected in a network of contracts of diverse nature. I am interested in the following topics: financial contagion, default cascades and propagation of financial distress.

Projects

FOC : European FET OPEN Project: "Forecasting Financial Crises"

How does the structure of financial networks impact on systemic risk?
When is a financial system, really, more stable? Are crises avoidable? Is it more desirable to have a tightly-nit architecture or one organized in clusters.

OTC : Swiss National Fund Project: "OTC Derivatives and Systemic Risk in Financial Networks"

What is the role of Over-the-Counter (OTC) contracts in systemic crises? Are OTC just a useful tool for managing risk?

Trust-Based Networks

On-line social networks provide novel ways for people to connect and share digital content over the web. However, this seldom translate in effective tools to build and share knowledge.
This is way we investigate how trust-based social network can be used to tackle societal issues. We have developed an algorithm for personalized ranking TrustWebRank
I am interested in the following topics: Trust, Recommender Systems, Buyers-Sellers in e-Markets, Coalition formation

Publications