Gorazd Atanasovski
  • Home
  • Research
  • Software
  • Axioms
  • QuantAI
Gorazd Atanasovski
Axiomatic Research & State-Space Execution
Synthesis of continuous-time filtration manifolds. Engineering absolute market dominance through rigorous measure-theoretic isomorphisms and high-dimensional state-space dynamics.
SYSTEM TOPOLOGYISOMORPHIC
ALGO TRADING SOCIETYFOUNDER & LEAD
DATA TERMINALBLOOMBERG INTEGRATED

Theoretical Research

Rigorous formalization of microstructural point processes, continuous-time martingale representations, and advanced measure theory. Heuristic approximations are systematically eradicated in favor of exact analytical bounds and fractional stochastic derivations.

Initialize Research

Analytical Manifold

High-dimensional filtration environments abandoning standard algorithmic design. Featuring GORAZD: defined not as empirical machine learning descent, but as the exact analytical evaluation of infinite-dimensional Fréchet derivatives within a Sobolev space.

Deploy Infrastructure

Measure-Theoretic Isomorphisms

The operational nexus. We synthesize Radon-Nikodym derivatives, spectral filtrations of non-stationary stochastic processes, and rigorous optimal control deployments into asymmetric, probability-weighted execution mappings.

Access Telemetry

Core Competencies & Theoretical Architecture

I. Advanced Stochastic Frameworks & Rough Volatility

Formulation of exact analytical boundaries for martingale pricing within non-Markovian environments. Derivation of fractional stochastic processes, eradicating the need for heuristic smoothing.

Definition 1 (Volatility Formulation):

The volatility process is strictly formulated as a fractional Brownian motion defined by its covariance structure, yielding exact pathwise regularity.

\[ \mathbb{E}[B_H(t)B_H(s)] = \frac{1}{2} \left( |t|^{2H} + |s|^{2H} - |t-s|^{2H} \right) \]

The Hurst parameter is strictly bounded to preserve roughness:

\[ H \in \left(0, \frac{1}{2}\right) \]

II. Measure-Theoretic Filtration & State Space Dynamics

Abandoning empirical machine learning in favor of exact analytical solutions to Stochastic Partial Differential Equations. Application of Girsanov transformations for formal probability measure shifts.

Non-Linear Paradigm:

Optimal control functions are bounded by unassailable analytical inequalities. We formalize the state dynamics via the rigorously derived Backward Stochastic Differential Equation:

\[ -dY_t = f(t, X_t, Y_t, Z_t)dt - Z_t dW_t \]

III. Topological Data Structures

Rejecting standard data engineering workflows for pure measure-theoretic data filtration. Construction of high-dimensional statistical architectures utilizing Banach and Hilbert space embeddings.

Ingestion Axiom:

Data ingestion is formalized strictly as a continuous mapping of observable phenomena onto a rigorously defined Polish space, ensuring absolute Borel measurability.

IV. Microstructural Limit Order Book Asymptotics

Eliminating reliance on standard broker routing. Modeling optimal order execution dynamics via marked point processes and Hawkes process formulations.

Theorem 1 (Execution Optimality):

Optimal execution routing is formulated as a rigorously proven solution to a stochastic optimal control problem. The value function is the unique viscosity solution to the Hamilton-Jacobi-Bellman equation under strict martingale constraints:

\[ \partial_t v + \sup_{u \in \mathcal{U}} \left\{ \mathcal{L}^u v + f(t,x,u) \right\} = 0 \]