Research & Publications
Where academic research and real banking decisions meet
My research asks questions a bank's Treasury actually faces: how to see liquidity stress coming, how assets behave when liquidity dries up, and how to measure the stability of banks and of the system. Most of it applies machine learning.
Constructing early warning indicators for banks using machine learning models
Coskun Tarkocin and Murat Donduran · The North American Journal of Economics and Finance, vol. 69
Uses supervised machine learning on data from 2007 to 2021 to give banks early warning of liquidity stress, backtested on the COVID-19 period. An ensemble model using the RUSBoost algorithm performed best. An initial version was presented at the 7th International Conference on Economics in April 2021.
Liquidity classification of equities under stress using machine learning models: evidence from major world share indices
Coskun Tarkocin and Murat Donduran · Review of Economics and Finance, vol. 21, pp. 1317–1331
Finds that sector, beta, industry and market capitalisation help predict how equities behave during liquidity stress.
Bank stability and systemic risk measurement: application of Z-score variations to the Turkish banking sector
Coskun Tarkocin and Murat Donduran · International Journal of Finance & Banking Studies, vol. 12, no. 1, pp. 63–73
Compares Z-score measures of bank stability in Turkey through regulatory change, the 2018 currency crisis and COVID-19, and uses a leave-one-out Z-score to estimate each bank's contribution to systemic risk.
Earlier work
Comparing yield curve estimation methods: performance for the Turkish market
Coskun Tarkocin · Master's thesis, MA in Economics
Estimates the yield curve for the Turkish bond market with several methods widely used by central banks, and asks which performs best.
Different input-output approaches' impact on data envelopment analysis efficiency results: application to the Turkish commercial banking sector
Coskun Tarkocin and Murat Gencer · Bankacılar Dergisi (Journal of Bankers)
Measures the relative efficiency of 22 Turkish commercial banks between 2003 and 2009 with data envelopment analysis, under three different input-output models.
Interest rate movements during the global financial crisis and the impact on the banking sector
Coskun Tarkocin and Adem Yekeler · International Conference on Economics, Turkish Economic Association (in Turkish)
Looks at how Turkish bank balance sheets changed as interest rates fell from 2002 to 2010, and what record-low rates meant for bank margins and strategy.
Articles
One-page summaries of my papers, research blogs and commentary now have their own page.
Read the articles →Research interests
- Financial stability
- Bank liquidity
- Macroprudential regulation
- Systemic risk
- Early warning indicators
- Stress testing
- Predictive analytics
- Machine learning in finance
Academic engagement
Former Visiting Research Fellow
Qatar Centre for Global Banking & Finance
King's Business School, King's College London