Head of the team of global strategists covering macro, equities, bonds, currencies, and commodities.

Developing machine learning models on financial markets in R (caret) and Python (scikit-learn) with the goal of creating useful trading strategies. Focus on data, data quality and feature engineering. Current projects focus on Chinese equities. Written interface in Python to interact with Bloomberg's API.Developed and managing a dynamic asset allocation strategy (Stronghold). The underlying model focuses on managing the tail-risk. The strategy comes in two versions (USD and EUR) and were commercialised in 2017.Enhanced the existing equity factor model with new factors and better normalisation of data on a global universe of 2,000 stocks. The bank's primary engine in equity research.Other models include breakout models on liquid futures (cross-asset) and 13F filing analysis.

Global equity model (S&P 1200 universe) based on the factors: quality, value, momentum and beta driving cost-effective equity research. Model output was used by sales traders and the top picks were live traded.Pair trading strategies using standard co-integration method. Valuation models based on modelling P/B ratios on single stocks using ensemble learning methods such as random forests models (the feature space was other fundamental ratios). CROIC (cash return on invested capital) models which essentially fitted EV/IC ratios across global equities based on various fundamental metrics used as features.Fully automated earnings season script (in R) scanning the entire earnings release calendar including key earnings estimates. Another script (in R) analysed the entire earnings season across industries and sectors with output automated into PDF.Created several R packages used internally in the organisation. The most important R package was wrapper functions to the Rblpapi (on CRAN) package which provided better data output (object type) and also more easily handled arguments for the various Bloomberg requests. Another R package contained wrapper functions for the ggplot2 (on CRAN) package which provided implementation of ggplot objects.General statistical studies of value, momentum, quality, volatility/beta, and other risk factors. Backtesting of dynamic asset allocation models based on mean-variance, minimum variance, risk-parity (the classic implementation, entropy implementation, cluster methods), and drawdown-based statistics.

Responsible for equity research writing daily market comments, trade ideas, sector/industry analyses and the bank's Quarterly Outlook. Regular guest on CNBC and Bloomberg TV. Client meetings with HNWI and seminar in across key markets in Europe.Developed simple trading strategies (rule-based) using simple statistical indicators in R. Developed deep value equity strategy in R with the model output used by the Trading Advisory Team.

Financial technology startup financed by business angels. The company offered financial analytics on public listed companies through a web-application delivered to investors and banks across the Nordic region.Product development scoping our financial reporting and metrics logic. Developed proprietary valuation metrics. Technology sales to banks ending with a system integration in Bankdata.

Devloped a substantial database, that contained large amounts of accounting and non-accounting data on all major publicly traded Danish companies which was used in stories on publicly listed companies.Thorough analysis of property transactions that eventually revealed one the largest property scandals in the history of Denmark. Based on several articles covering the scandal, the two journalists were later nominated for The Cavling Award, the most prestigious journalist award in Denmark.