Schlagworte
Skills
Data Science , Machine Learning , Artificial Intelligence , Algorithmic Trading , Systematic Investing , Systematic Portfolio Management , High-Frequency Trading , Algorithmic Market Making , Image Processing ,Time-series Forecasting , Database (MongodB , SQL , Kdb/Q , ...) , Recommendation Systems , Survival Analyics , Cohort Aalysis , A/B testing , Statistical Hypotheis Testing , ....
python, C++, Pandas, Keras, Scikitlearn, tensorflow, AWS, Docker, Algorithm, CTA, Python packages, matplotlib, statlib, datetime, numpy, Pyspark, Deep Learning, large data, machine learning, Neural Networks, large data sets, ALGORITHMIC TRADING, Programming, VIX, Game Theory, Information Theory, scipy, sklearn, * C++, C#, LINK, Multithread, Event-based(publisher, seaborn, fb prophet, statlib Shared Memory, Lock-free queues, MySql, R, VBA/Excel
python, C++, Pandas, Keras, Scikitlearn, tensorflow, AWS, Docker, Algorithm, CTA, Python packages, matplotlib, statlib, datetime, numpy, Pyspark, Deep Learning, large data, machine learning, Neural Networks, large data sets, ALGORITHMIC TRADING, Programming, VIX, Game Theory, Information Theory, scipy, sklearn, * C++, C#, LINK, Multithread, Event-based(publisher, seaborn, fb prophet, statlib Shared Memory, Lock-free queues, MySql, R, VBA/Excel
Projekthistorie
06/2017
-
03/2018
SENIOR VICE PRESIDENT at AMM E-FX TRADING
BNP PARIBAS
* Contributed nearly 1M$ per year to desk's P&L by researching and developing signals and strategies for hedging and trading of eFx book based
on classification of market and client data (Alpha research, Strategy design and backtest, FX 10- minute alpha from orderbook, price action, client
trades/order flow, CLS/EBS volume, news sentiment, events, other currency pairs, other asset prices, ...) successfully monetizing toxic client flow
using machine learning/predictive modeling, python
* Designed an Event-Driven model designed for post-event (2-20 min) market reaction based on surprise factor and other features with accuracy
score more than 0.65
* Designed a classifier for predicting Min Ask and Max Bid within a 5-10 min interval using market prices and orderbook states for use within the
market making framework, exiting trades passively (accuracy score>0.49)
* Used Ravenpack and Federal Reserve news analysis methodology for creating an online news sentiment score for currencies with 10-minute
Alpha (Incomplete)
* Worked closely with software developers in building the book for customized hedging of the toxic clients (Shared Memory,...)
on classification of market and client data (Alpha research, Strategy design and backtest, FX 10- minute alpha from orderbook, price action, client
trades/order flow, CLS/EBS volume, news sentiment, events, other currency pairs, other asset prices, ...) successfully monetizing toxic client flow
using machine learning/predictive modeling, python
* Designed an Event-Driven model designed for post-event (2-20 min) market reaction based on surprise factor and other features with accuracy
score more than 0.65
* Designed a classifier for predicting Min Ask and Max Bid within a 5-10 min interval using market prices and orderbook states for use within the
market making framework, exiting trades passively (accuracy score>0.49)
* Used Ravenpack and Federal Reserve news analysis methodology for creating an online news sentiment score for currencies with 10-minute
Alpha (Incomplete)
* Worked closely with software developers in building the book for customized hedging of the toxic clients (Shared Memory,...)
07/2016
-
06/2017
SENIOR RESEARCH FELLOW
SINGAPORE MANAGEMENT UNIVERSITY
* Applied Recurrent Neural Network to enhanced technical data (implied vol, sentiment) of Russel1000 in order to train the network in predicting
subsequent price move (Accuracy 0.21 out-of-sample vs. 0.1 random, Long-Short Equity based on this strategy produces 0.6% average daily
return without transaction/shorting costs)
* Enhanced OLMAR (Online Learning Moving Average Reversion) with a predictive model based on volume (interacting factor), momentum,
adjusting for Industry bias and Implementing transaction cost
* Applyed Deep Learning to options data for designing a volatility trading strategy (Not Completed)
* Worked on a CTA using standard systematic macro factors (Momentum, Volatility, Mean-Reversion around trend-line, ...) combined by a
Recurrent Neural Network (portfolio optimization wasn't completed)
subsequent price move (Accuracy 0.21 out-of-sample vs. 0.1 random, Long-Short Equity based on this strategy produces 0.6% average daily
return without transaction/shorting costs)
* Enhanced OLMAR (Online Learning Moving Average Reversion) with a predictive model based on volume (interacting factor), momentum,
adjusting for Industry bias and Implementing transaction cost
* Applyed Deep Learning to options data for designing a volatility trading strategy (Not Completed)
* Worked on a CTA using standard systematic macro factors (Momentum, Volatility, Mean-Reversion around trend-line, ...) combined by a
Recurrent Neural Network (portfolio optimization wasn't completed)
01/2014
-
01/2016
QUANTITATIVE STRATEGIST ON QIS (QUANTITATIVE INVESTMENT STRATEGIES) TEAM
GOLDMAN SACHS ASSET MANAGEMENT
* Contributed to Research on 2 equity factors based on Indication of Interest data& Intraday trade imbalance (alpha signals extracted were
implemented within the portfolio optimization signal set, with Sharpe ratios of 0.6 and 1.4)
* Researched and Implemented an orthogonal 6-factor model for a fast trading model enhancing the current trade execution system with a daily
up/down forecast signal (open to close)
* Implemented accounts based on the GTAA (Global Tactical Asset Allocation) model of the Macro team, enhancing and implementing the
accounts in the codebase in SECDB
* Implemented and migrated Equity Alpha Factors (Industry-Specific Data, Intraday Market Data, Value, Quality, Momentum, Sentiment,
Profitability, ...) of the Quantitative Equity team into SECDB codebase
* Conducted and completed Research projects using large data sets for investigating Alpha in certain market datasets using standard criteria for
verification of Alpha/Factor Interaction/Factor Turnover (IC, IR, U-chart,5-1 Decile Long-Short ,63-day Gamma, Residual regression, Turnover
heat map, ...)
* Implemented a Non-linear(L-1) filter with Mann-statistic for trend Filtering/Detection in equity trend
implemented within the portfolio optimization signal set, with Sharpe ratios of 0.6 and 1.4)
* Researched and Implemented an orthogonal 6-factor model for a fast trading model enhancing the current trade execution system with a daily
up/down forecast signal (open to close)
* Implemented accounts based on the GTAA (Global Tactical Asset Allocation) model of the Macro team, enhancing and implementing the
accounts in the codebase in SECDB
* Implemented and migrated Equity Alpha Factors (Industry-Specific Data, Intraday Market Data, Value, Quality, Momentum, Sentiment,
Profitability, ...) of the Quantitative Equity team into SECDB codebase
* Conducted and completed Research projects using large data sets for investigating Alpha in certain market datasets using standard criteria for
verification of Alpha/Factor Interaction/Factor Turnover (IC, IR, U-chart,5-1 Decile Long-Short ,63-day Gamma, Residual regression, Turnover
heat map, ...)
* Implemented a Non-linear(L-1) filter with Mann-statistic for trend Filtering/Detection in equity trend
01/2013
-
12/2013
ALGORITHMIC TRADING QUANT ON FX SPOT TEAM IN INVESTMENT BANKING DIVISION
BARCLAYS CAPITAL
* Designed and Implementated hedging strategies based on decay profiles/trade patterns of streams (calculating riskiness of each currency
position based on toxicity score of the counterparty clients)
* Designed and Implemented methods for Trade Cost Analysis (risk reduction and slippage measures for child trades)
* Designed and Implementated of an optimal execution strategy for large orders based on Dynamic Programming model and Risk-Return Profile of
small trades
* Implemented robust strategies for detection of regime shift in the trend of currencies
* Built a vector autoregression model for predicting intraday bid-ask spread (EBS/Reuters), volume and volatility for multiple pairs/crosses
* Studied distribution of risk-increasing/risk-decreasing trades, for different currency crosses and different clients
* Implemented Variance-Covariance VAR method for our portfolio, testing different skews on currency Pairs for reduction of risk, checking their
effectiveness in terms of speed of risk-reduction vs. P&L
position based on toxicity score of the counterparty clients)
* Designed and Implemented methods for Trade Cost Analysis (risk reduction and slippage measures for child trades)
* Designed and Implementated of an optimal execution strategy for large orders based on Dynamic Programming model and Risk-Return Profile of
small trades
* Implemented robust strategies for detection of regime shift in the trend of currencies
* Built a vector autoregression model for predicting intraday bid-ask spread (EBS/Reuters), volume and volatility for multiple pairs/crosses
* Studied distribution of risk-increasing/risk-decreasing trades, for different currency crosses and different clients
* Implemented Variance-Covariance VAR method for our portfolio, testing different skews on currency Pairs for reduction of risk, checking their
effectiveness in terms of speed of risk-reduction vs. P&L
01/2012
-
03/2012
QUANTITATIVE PORTFOLIO RESEARCH INTERN
FERI TRUST
* Devised quantitative methods for detecting market panic using comparison of current VIX index to implied 30-day forward VIX calculated using
VIX futures
* Analyzed and compared put options with different maturities and strike prices in order to find relatively more expensive ones to short
* Researched possible correlation at times of distress between different asset pairs within a class of assets and recommending pairs with strong
correlation for trade
VIX futures
* Analyzed and compared put options with different maturities and strike prices in order to find relatively more expensive ones to short
* Researched possible correlation at times of distress between different asset pairs within a class of assets and recommending pairs with strong
correlation for trade
09/2007
-
07/2010
RESEARCH ASSISTANT
UNIVERSITY OF MICHIGAN
* Researched applications of Game Theory and Stochastic Control in Resource Allocation over Wireless Networks
09/2003
-
05/2007
RESEARCH ASSISTANT
UNIVERSITY OF ILLINOIS
* Researched on Information Theory and Wireless Communication
Reisebereitschaft
Verfügbar in den Ländern
Deutschland