Research
Working Papers
Presentations
BIS-World Bank-Banxico Tenth Public Investors Conference (upcoming), Tenth Annual Young Scholars Finance Consortium at Texas A&M Mays, Sovereign Bond Market Conference 2026 on Sovereign Bond Markets in Geopolitical Storms, Northern Finance Association (NFA), Midwest Finance Association (MFA), 2026 Consortium on Asset Management, EUROFIDAI-ESSEC Paris December Finance Meeting (upcoming), 7th Annual RCF-ECGI Conference (upcoming), 2026 LAPE-FINEST Spring Workshop, Notre Dame Mendoza College of Business Seminar, Colorado Finance Summit (2025), ESADE Spring Workshop (2025), and BI Norwegian Business School Seminar (2024)
Abstract
We evaluate whether the U.S. Treasury could lower borrowing costs by managing the debt’s market value within its “regular and predictable” framework. Currently, the Treasury allocates 97.8% of borrowing surprises to bills and issues the duration supply announced a quarter earlier with 96.1% correlation, regardless of changes in swap spreads or term premia. Using auction-level demand elasticities, we estimate a 10% reallocation to shorter duration would have saved over $550 million annually at the 2021 peak. Exploiting the staggered adoption of ESA 2010 in Europe, which extended market-value reporting, we find Core sovereigns lowered term premia by 53 basis points.
Presentations
Midwest Finance Association (MFA) and Inquire
Abstract
I recover cross-country term-premium differentials from two traded prices, the long-tenor cross-currency basis and the swap-spread differential, with no expectations model. The quantity is each sovereign’s own-currency term premium on its own bond, unhedged, distinct from a hedged covered-parity deviation. U.S. 10-year Treasury term premia sat 28 basis points below German Bund and 37 below Japanese JGB term premia before 2008, and 56 and 83 above since 2013. The recovered differentials behave as term risk premia: duration supply prices them at −1.05 percentage points per unit of log U.S. free float, and duration driven policy surprises move them in identified windows.
Presentations
Northern Finance Association (NFA), and Midwest Finance Association (MFA)
Abstract
At long maturities, deviations from covered interest parity (CIP) are measured against interest-rate swap (IRS) curves, and the resulting cross-currency basis has been large and persistent since the global financial crisis. We show that the quoted basis is a relative price between two curves and should not be read directly as an arbitrage return: at the 2022 benchmark transition, the same five-year swap was quoted with a basis of −24 basis points under one floating-rate convention and +6 basis points under another. We argue that relative term-funding costs are an important component of the long-dated cross-currency price. The five-year forward premium is closely aligned with cross-currency differences in same-issuer term-funding rates. In the common monthly sample, the mean absolute residual is 24.9 basis points using IRS rates and 10.9 basis points using same-issuer bond yields. The same-issuer residual is approximately invariant to the benchmark transition. Across seventeen independently quoted currency pairs, a currency-level representation accounts for 99.95 percent of the cross-sectional sum of squares at five years. Funded returns at observed dealer borrowing costs are negative in every period and tenor.
Presentations
2024 Annual Meetings of the American Finance Association (AFA), 2024 European Finance Association (EFA), 2024 Northern Finance Association Annual Conference (NFA), Midwest Finance Association Conference (MFA), Young Scholars Nordic Finance Network Workshop (NFN), Alpine Finance Summit, Finance and Accounting Annual Research Symposium, Financial Engineering and Banking Society, BI Workshop on Uncertainty, NYU Stern School of Business, Miami Herbert Business School, Frankfurt School of Finance and Management, and BI Norwegian Business School
Abstract
Interest rate swaps should theoretically generate zero expected excess returns under risk-neutral pricing, yet systematically deviate from this principle. Fixed-rate receivers earn 2.74% annual excess returns (“Hairy premiums”) in U.S. dollar 10-year swaps, with similar magnitudes across G10 currencies. Applying novel triangular extraction methodology from cross-currency swaps, we decompose realized U.S. relative premiums into expected and unexpected components. Expected premiums explain 67.1-77.1% of realized variance and significantly differ from zero, demonstrating markets expect these deviations ex-ante. We attribute expected premiums to market incompleteness: absent long-term government floating-rate notes and binding shorting constraints prevent static replication, forcing costly dynamic hedging. Forecast errors reveal systematic monetary policy misinterpretation.
3. Time-varying Market Segmentation: The Twin D’s and a Tale of Two Tails
(This paper subsumes the previous paper “Currency Crashes and Sovereign Defaults: Two Markets and One Tail Risk”)
Presentations
AFA Annual Meeting, EEA Annual Meeting, Bank of England: 8th International Conference on Sovereign Bond Markets, Anderson School of Management Brownbag Seminar, and SDU
Abstract
This paper documents striking market segmentation between sovereign credit default swaps (CDS) and far out-of-the-money (FOM) foreign exchange (FX) options markets. Despite pricing exposure to the same fundamental tail risks, sovereign default and currency crash, the “Twin Ds”, these markets exhibit persistent pricing discrepancies that cannot be explained by differences in contract specifications or risk exposures. We propose that this segmentation reflects institutional separation between market participants: credit specialists and FX traders operate in distinct spheres with limited information sharing during normal times. Using two cross-market trading strategies, one exploiting USD-denominated CDS versus FOM FX options spreads, and another using quanto CDS versus FOM FX options, we show that these pricing discrepancies yield substantial abnormal returns with Sharpe ratios exceeding 7.2. Crucially, both strategies generate consistent results despite differing in their treatment of depreciation size upon default, confirming that our findings reflect genuine market segmentation rather than risk compensation for variable depreciation magnitude. The abnormal returns collapse to zero during sovereign crises, when information sharing intensifies and pricing converges across previously segmented markets. Our analysis further reveals that information flows unidirectionally from CDS to FX options markets, supporting our institutional explanation. We exploit regulatory changes and cross-country variation in market structure to strengthen identification and rule out alternative explanations based on rational risk pricing. These findings have implications for understanding market structure effects on tail risk pricing and for detecting potentially destabilizing market fragmentation.
Presentations
EEA Annual Meeting, AFA Annual Meeting (Poster), Auckland Center’s Derivative Markets Conference, International Risk Management Conference, Virtual Derivatives Workshop, and Anderson School of Management Brownbag Seminar
Abstract
I show that a no-arbitrage consistent but costly collateral rental yield explains about two-thirds of the apparent standard Covered Interest Rate (CIP) violations. I proxy this yield with the difference between the risk-free and overnight index swap rates between the cross bilateral currencies and apply it to both short and long-term CIP violations. Further, the results suggest an important direct collateral channel through which costly collateralization explains CIP violations, independent of previously documented global risks and intermediary frictions.
1. Constraints of Collateralization and Regulation and the Law of One Price
Macro Financial Modeling Project Fellowship Award from the Becker Friedman Institute at the University of Chicago
Summary
I research collateralisation and bank regulatory changes in the post-2008 crisis period and their ability to explain large and persistent deviations from the law of one price observed in a number of arbitrage opportunities measured by cash-derivative bases across asset classes. The study attributes the inability to narrow these arbitrage opportunities to costs of participation which arise due to intermediaries facing collateral, leverage and liquidity constraints, resulting in limits-to-arbitrage. These constraints inhibit regulated dealers/investors’ ability to arbitrage bases across various asset classes and over time. Moreover, I go a step further and measure the magnitude of the effects of each binding constraint on the deviations from law of one price across time and in the cross section across several asset classes. This allows me to classify asset classes by the magnitude of impact received from the binding constraints. The asset class bases I look at are: Bonds-CDS, Treasuries-Interest Rate Swaps, Covered interest rate parity, Inflation rate swaps-TIPS, and Stock Index - Stock index swaps bases.
Publications
Georgievska A., Georgievska, Lj., Stojanovic, A. and Todorovic, N. (2011), “Country Debt Default Probabilities in Emerging Markets: Were Credit Rating Agencies Wrong?” in Robert W. Kolb (ed.), Sovereign Debt: From Safety to Default (by invitation), Wiley, ISBN 978-0-470-92239-2.
Georgievska A., Georgievska, Lj., Stojanovic, A. and Todorovic, N. (2008), “Sovereign rescheduling probabilities in emerging markets: a comparison with credit rating agencies’ ratings”, Journal of Applied Statistics, 35(9), p. 1031-1051.