Identifying Systemically Important Banks in Payment Systems
1.
Latsis SymposiumEconomics on the moveZürich, 13 September 2012Identifying SystemicallyImportant Banks in PaymentSystemsKimmo Soramäki, Founder and CEO, FNASamantha Cook, Chief Scientist, FNA
2.
Interbank Payment Systems• Provide the backbone of all economic transactions• Banks settle claims arising from customers transfers, own securities/FX trades and liquidity management• Target 2 settled 839 trillion in 2010 Example system across this• 130 out of 290 billion lent to presentation: A pays B one Greece consist of Target 2 unit and C two, B pays A one, balances and C pays B one
3.
Systemic Risk in Payment Systems• Credit risk has been virtually eliminated by system design (real-time gross settlement)• Liquidity risk remains – “Congestion” – “Liquidity Dislocation”• Trigger may be – Operational/IT event – Liquidity event – Solvency event• Time scale = intraday
6.
Common centrality metrics Degree: number of links Closeness: distance to other nodes via shortest paths Betweenness: number of shortest paths going through the node Eigenvector: nodes that are linked by other important nodes are more central, probability of a random process
7.
Eigenvector Centrality• Eigenvector centrality (EVC) vector, v, is given by – vP=v , where P is the transition matrix of an adjacency matrix M – v is also the eigenvector of the largest eigenvalue of P (or M)• Problem: EVC can be (meaningfully) calculated only for “Giant Strongly Connected Component” (GSCC)• Solution: PageRank
8.
Pagerank• Solves the problem with a “Damping factor” which is used to modify the adjacency matrix (M) – Gi,j= Si,j• Effectively allowing the random process out of dead-ends (dangling nodes), but at the cost of introducing error• Effect of – Centrality of each node is 1/N – Eigenvector Centrality – Commonly is used
10.
Network processes• Centrality depends on network process – Trajectory : geodesic paths, paths, trails or walks – Transmission : parallel/serial duplication or transfer Source: Borgatti (2004)
11.
Distance to Sink• Markov chains are well-suited to model transfers along walks• Example: From B 1 To A From C 2 From A To B From C 1 From A To C From B
12.
SinkRank SinkRanks on unweighed• SinkRank averages distances networks to sink into a single metric for the sink• We need an assumption on the distribution of liquidity in the network – Asssume uniform -> unweighted average – Estimate distribution -> PageRank weighted average – Use real distribution -> Average using real distribution as weights
15.
Experiments• Design issues – Real vs artificial networks? – Real vs simulated failures? – How to measure disruption?• Approach taken 1. Create artificial data with close resemblance to the US Fedwire system (Soramäki et al 2007) 2. Simulate failure of a bank: the bank can only receive but not send any payments for the whole day 3. Measure “liquidity dislocation” and “congestion” by non-failing banks 4. Correlate them (the “disruption”) with SinkRank of the failing bank
16.
Data generation processBased on extending Barabasi–Albert model of growth andpreferential attachment
17.
Distance from Sink vs Disruption Relationship between Failure Distance and Disruption when the most central bank fails Highest disruption to banks whose liquidity is absorbed first (low distance to sink)
18.
SinkRank vs Disruption Relationship between SinkRank and Disruption Highest disruption by banks who absorb liquidity quickly from the system (low SinkRank)
http://www.economics-ejournal.org/economics/discussionpapers/2012-43