Internal Research Notes
By Philippos Varnava-Marouchos

Bayesian Analysis , Python

Signals of Crisis

Bayesian Inference transforms risk perception into explicit, continuously updated probabilities. A case in the unraveling of the 2008 subprime crisis.

Monte Carlo Simulation , Python

The Coin Toss

Aggregate belief of uncertainty converges into a predictable distribution that uncovers the statistical foundation for events of symmetrical expectation via the Monte Carlo Simulation, which in turn becomes the meeting point between the statistical shape of a movie audience not-knowing and the 2024 Microsoft-Activision Takeover.

All these in an attempt to diagnose whether uncertainty is being priced coherently in events of this type.

Modern Portfolio Theory, Python

At the Fallible Edge of the Efficient Frontier; Market Regimes Across 2021–2023

This study uses eight composite sleeves to examine how the observed return–risk trade-off changed across 2021, 2022 and 2023. The central investment lesson is that an unattractive frontier may accurately describe the limitations of the assets , highlighting the importance of timing in allocations across regimes.

The analysis is grounded in historical price returns and does not take into account reinvested income. The conclusions concern the way in which the opportunity set changed over time, not the predictability or the timing of those changes.