Transcript CISC 849 : Applications in Fintech

BIG DATA APPROACH TO ANALYZING
MARKET VOLATILITY
ARUNPRABHU SADAIYAPPAN
Dept of Computer & Information Sciences
University of Delaware
CISC 849 : Applications in Fintech
WHY BIG DATA TECHNOLOGY?
CISC 849 : Applications in Fintech
BIG DATA IN FINANCE SERVICES
• Markets
• Customers
• Channels
• Products
• Regulations
• Competitors
• Suppliers
• Employees
CISC 849 : Applications in Fintech
MARKET LIQUIDITY
Market's ability to facilitate the purchase or sale of
an asset without causing drastic change in the asset's
price
CISC 849 : Applications in Fintech
LIQUIDITY INDICATOR
• VOLUME SYNCHRONISED PROBABILITY OF
INFORMATION TRADING
CISC 849 : Applications in Fintech
TECHNIQUES FOR BIG DATA
ANALYSIS
• EFFICIENT FILE ORGANISATION FOR STORING
TRADING RECORDS
• EFFICIENT ALGORITHM FOR COMPUTING VPIN
• PARALLELIZATION OF COMPUTATIONAL TASKS
CISC 849 : Applications in Fintech
FILE ORGANIZATION
CONVENTIONAL METHOD:
• 67 Months worth of data (Liquid futures trade
information)
• File Format : CSV (Comma separated value)
• File Size
: 140 GB
Processing time for computing VPIN values :
142 seconds
CISC 849 : Applications in Fintech
FILE ORGANIZATION
PROPOSED METHOD:
• Same amount of data (for 67 months)
• File Format : HDF5 (Hierarchical Data Format)
• File Size
: 41 GB
Processing time for computing VPIN values :
0.4 second
CISC 849 : Applications in Fintech
FILE ORGANIZATION
IMPROVEMENT:
• 29% gain in memory storage
• Less time in reading/writing the file
• Better organization for data
Reason behind the Efficiency: HDF5 files store
data in binary form
CISC 849 : Applications in Fintech
ALGORITHM FOR COMPUTING VPIN
Initially, Data arrives at irregular frequency
Volume Bars
Bulk Volume
Classification
Buckets
VPIN
CISC 849 : Applications in Fintech
VOLUME BAR
• SHELL SORT – To order trades in a volume bar
• To compute Median Prices
CISC 849 : Applications in Fintech
SHELL SORT ANALYSIS
In-place Sorting (Space complexity is O(N) )
• BEST CASE (LOWER BOUND):
O(N(log N )^2)
• WORST CASE (UPPER BOUND):
Θ(N^2)
In this paper, the runtime is mentioned as O(log N )
CISC 849 : Applications in Fintech
ALGORITHM FOR COMPUTING VPIN
Time Taken to construct Volume bars with different
nominal prices
CISC 849 : Applications in Fintech
BULK VOLUME CLASSIFICATION
• Trades are classified as “Buyer-initiated” or “Sellerinitiated”
• BVC assigns a fraction of volume as buys
The remainder as sells based on normalized sequential
price change
CISC 849 : Applications in Fintech
BUCKETS
• When forming buckets  Each volume bar is
considered as a single trade with the nominal price
• 30 Volume bars in a buckets (maximum upto 50)
• Most recent buckets with buy and sell volumes 
Kept in Fixed Memory
CISC 849 : Applications in Fintech
BUCKETS
Pseudocode for storing Volume in Buckets
CISC 849 : Applications in Fintech
CALCULATING VPIN
FORMULA:
PSEUDOCODE:
CISC 849 : Applications in Fintech
EXPERIMENTAL RESULTS
Statistics on prices for volume bars and the resulting VPIN:
(Overall trades of ES)
CISC 849 : Applications in Fintech
EXPERIMENTAL RESULTS
CISC 849 : Applications in Fintech
PARALLELIZATION
PREFERRED IMPLEMENTATION: POSIX THREADS
Instrument with large number of trades assigned first
CISC 849 : Applications in Fintech
HEDGE ANALYTICS – SWISS BASED STARTUP
• RISK FACTOR ANALYSIS
• PERFORMANCE ANALYSIS
• NEW CUSTOMER WIN RATE ANALYSIS
 Universal Meta data Search Engine
 Incorporated NoSQL Technology to avoid inconsistency
CISC 849 : Applications in Fintech
THANK YOU!
CISC 849 : Applications in Fintech