Post on 27-Jan-2015
description
Financial Network Analytics
Kimmo Soramäkikimmo@soramaki.net
Presentation at European Central Bank
17 August 2011
Katsushika Hokusai. The great wave off Kanagawa ~1830
Eratosthenes' map of the known world, c. 194 BC
Go is a board game that originated in China more than 2,000 years ago.
Data deluge
• Data tsunami - digital information is doubling every 1.2 years
• Regulatory response to recent financial crisis was to strengthen macro-prudential supervision with mandates for more regulatory data
• The challenge will be to understand and analyze the data (public, commercial and regulatory)
• Data enables “Analytics based policy”, i.e. the application of computer technology, operational research, and statistics to solve regulatory problems
Network maps
• Erastothenes created the first known map of the world and is considered the father of geography and the concepts of latitude and longitude (i.e. ways to measure it)
• Recent financial crisis brought to light the need to look at links between financial institutions
• Network are a natural way to present financiallinkage data. Mapping of the financial systemhas only begun
• Maps reduce uncertainty about interconnections
Soramaki, Bech et al (2006)
Garrett, Mahadeva and Svirydenska (2011)
Intelligence
• Pattern recognition is hard for computers -> the best Go programs only manage to reach an intermediate amateur level
• Financial crisis are different and rare
• The patterns to be recognized must be frequent enough for computers to learn
• A solution is to augment human intelligence (in contrast to AI)
• Intelligence amplification (William Ross Ashby 1956)
Financial Network Analytics Platform
Objectives
• Provide a tool for exploration, analysis and visualization of regulatory financial data
• Provide a extendible platform for custom functionality, agent based models and other simulation models
• Make advances in research available to policy
Situation around March 2010
Payment System Simulator
Furfine algorithm FNA 1.1
• Because analyzing financial microdata is difficult– Files are large, content is difficult to understand– Data is incomplete and messy– Data is confidential– Most analysis packages are opaque– Error and trial
Paradigm
Data validation
Analysis and modeling Visualization
PlatformEasy to extend functionality, integration to corporate IT
Intuitive to use, flexible GUI, dashboard
ModulesModels for financial
stability, (agent-based) simulation,
visualization, …
Can be developed by third parties,
saves development time
SharingSame tool is shared
among several regulators leads to
robust models
Knowledge transfer over organizations
and time
Platform + modules + sharing
Current modules
NET – Network analysis, network metrics,Network operations
VIZ-HIGHCHARTS – Common chart visualizations
VIZ-PROTOVIS – Interactive network visualizations
DATA – Tools for working with large CSV files
MATCH – Algorithm for identifying overnight loans from payment data
FLOW – Interbank payment system simulator
14
The network model
A1Source=V1Dest=V2Props…
A2Source=V2Dest=V3Props…
A3Source=V2Dest=V1Props…
Network 1 – N1Prop1=AbcProp2=Def…
V2Prop1=20Prop2=AbProp3=Bc
V3Prop1=20Prop2=AbProp3=Bc
V1Prop1=20Prop2=AbProp3=Bc
N2 Prop1=Aaa; Prop2=Bbb;…..
N3 Prop1=Aaa; Prop2=Bbb;…..
Time Series
15
Explain screen
Screen elementsAccess via browser in intranet, internet
or desktop
Operation based on commands
Result panel shows command output
Submit commandEach command has
different parameters
And creates files (charts, data, etc)
Files and database connections are in
file panels
‘Visualize’ screen shows created
charts and layouts
Switch between ‘point-and-click’ and command line view
16
Tabs allow multiple visualisations open
All visualisations are html documents that
work also outside FNA
17
Dashboard (concept)The dashboard can combine multiple views to the data on a single screen
It can be available e.g. on the
intranet and updated daily
18
All commands can be submitted using command syntax
Command line
All commands submitted (also from point-and-click) are
shown in history
History provides an easy way to make
new scripts for research or for the
dashboard
19
Command line
Scripts can be run from the scripts panel or as regular jobs by
the server
www.fna.fi
(as from 29 August 2011)
Contact uskimmo@soramaki.net
FNA is available at
• Performance – Client server architecture allows use of high performance servers, computer
clusters and cloud computing– High-performance graph engine (neo4j.org) – Fast client application (Google web toolkit e.g. as in gmail.com)
• Security– Sensitivity and confidentiality of data creates addition constraints for analysis– Data is stored on server where it can be protected better (vs analysts desktops)– Each user accesses FNA with her own account– SSL encryption of traffic– Logging and analytics, audit trail
• Integration to corporate IT – Integration to databases possible– FNA accounts can be managed centrally by IT (integration to LDAP systems)– Can run on most application servers– Modular structure allows easier updates
Technical details