dS = μS dt + σS dW    VaR_α = −inf{x : P(L > x) ≤ 1−α}    Σ = E[(r−μ)(r−μ)ᵀ]    ΔP ≈ δΔS + ½γ(ΔS)² + θΔt    HHI = Σᵢ sᵢ² E[R_p] = Σ wᵢE[Rᵢ]    σ_p² = wᵀΣw    SR = (μ−r_f)/σ    CVaR_α = E[L | L > VaR_α]    β = Cov(rᵢ,r_m)/Var(r_m)    N_eff = (Σwᵢ)²/Σwᵢ² ∂V/∂t + ½σ²S²∂²V/∂S² + rS∂V/∂S − rV = 0    SaR(α) = Q_α(slippage)    P(default) = Φ(−DD)    λ(t) = λ₀ exp(αN_t) max_w {wᵀμ − γwᵀΣw}    r_t = μ + σ_t ε_t    σ²_t = ω + αr²_{t−1} + βσ²_{t−1}    ESaR(α) = E[S | S > SaR(α)]    CR₁ = max(wᵢ)/Σwᵢ ρ(X,Y) = Cov(X,Y)/(σ_X σ_Y)    HF = (C × LT)/D    IF ≥ TSaR(α) × (1+κ)    L = Σ max(0, Dᵢ − Cᵢ × LT)    Δ = ∂V/∂S    Γ = ∂²V/∂S² f(r) = (2π)^{−n/2}|Σ|^{−½} exp(−½(r−μ)ᵀΣ⁻¹(r−μ))    κ = E[(X−μ)⁴]/σ⁴ − 3    ES = −(1/(1−α))∫_α¹ VaR_u du R(t) = ln(S_t/S_{t−1})    E[max(S−K,0)] = SΦ(d₁) − Ke^{−rT}Φ(d₂)    MDD = max_{t}(M_t − S_t)/M_t    ν = ∂V/∂σ P(cascade) = 1 − ∏(1−pᵢ)    funding = (mark−index)/8h    slippage(q) = ∫₀ᵦ (P(x)−P₀)dx / q    OI_cap = f(σ,depth,IF) w* = Σ⁻¹μ / 1ᵀΣ⁻¹μ    TE = σ(r_p − r_b)    IR = α/TE    corr_stress = ρ₀ + (1−ρ₀)(1−e^{−λΔVol})    RR = E[recovery|default] dS = μS dt + σS dW    VaR_α = −inf{x : P(L > x) ≤ 1−α}    Σ = E[(r−μ)(r−μ)ᵀ]    ΔP ≈ δΔS + ½γ(ΔS)² + θΔt    HHI = Σᵢ sᵢ² E[R_p] = Σ wᵢE[Rᵢ]    σ_p² = wᵀΣw    SR = (μ−r_f)/σ    CVaR_α = E[L | L > VaR_α]    β = Cov(rᵢ,r_m)/Var(r_m) ∂V/∂t + ½σ²S²∂²V/∂S² + rS∂V/∂S − rV = 0    SaR(α) = Q_α(slippage)    P(default) = Φ(−DD)    λ(t) = λ₀ exp(αN_t) max_w {wᵀμ − γwᵀΣw}    σ²_t = ω + αr²_{t−1} + βσ²_{t−1}    ESaR(α) = E[S | S > SaR(α)]    CR₁ = max(wᵢ)/Σwᵢ ρ(X,Y) = Cov(X,Y)/(σ_X σ_Y)    HF = (C × LT)/D    IF ≥ TSaR(α) × (1+κ)    Δ = ∂V/∂S    Γ = ∂²V/∂S² f(r) = (2π)^{−n/2}|Σ|^{−½} exp(−½(r−μ)ᵀΣ⁻¹(r−μ))    κ = E[(X−μ)⁴]/σ⁴ − 3    MDD = max(M_t − S_t)/M_t V(S,t) = e^{−r(T−t)} E_Q[payoff]    d₁ = [ln(S/K)+(r+σ²/2)T]/(σ√T)    LGD = 1 − RR    PD × LGD × EAD = EL
Home About Research
Quantitative finance, computational physics, and open tools. Projects and research by Otar Sepper.
Current Projects
DeFi Risk Dashboard
Live
Cross-protocol lending risk metrics for Aave V3 and Compound V3. Scenario-based stress tests, liquidation heatmaps, and Liquidity-at-Risk analysis measuring whether DEX markets can absorb liquidations. Updates every 15 minutes.
risk.sepperlabs.com →
CalcMonte
Live
Financial calculator combining loan amortization analytics with Monte Carlo investment simulation. Single-file app, no sign-up, runs entirely in your browser.
calcmonte.com → | github →
Bayesian Incentives Optimizer
Open Source
Bayesian optimization framework for DeFi protocol incentive allocation. Regime detection, elasticity estimation, and data-driven reallocation strategies.
github →
Contact
Email contact@sepperlabs.com
About

SepperLabs is the research and engineering practice of Otar Sepper, formerly of Gauntlet. Current work spans quantitative finance, DeFi protocol risk, systematic trading, and computational tools.

Research interests include order book microstructure, liquidation mechanics, tail-risk modeling, portfolio construction, and the application of statistical physics to financial systems.

Prior work includes quantitative risk frameworks for DeFi lending and derivatives protocols, live systematic trading systems, deep learning pipelines for NLP and recommender systems, and large-scale research initiatives across finance and industry.

Otar holds a Ph.D. in theoretical physics from the University of Arizona, with earlier research on unconventional superconductivity in quasi-one-dimensional conductors.

Areas of Focus
DeFi Protocol Risk
  • Liquidation engine modeling and cascade analysis
  • Solvency stress testing under volatility shocks
  • Margin parameter calibration and capital efficiency
  • Monte Carlo simulation of cross-market contagion
  • Incentive design and structural stability analysis
Quantitative Finance
  • Portfolio construction and factor-based allocation
  • Volatility modeling and regime detection
  • Tail-risk analytics and scenario analysis
  • Statistical arbitrage and systematic signal research
Publications
Quantitative Finance
Slippage-at-Risk (SaR): A Forward-Looking Liquidity Risk Framework for Perpetual Futures Exchanges
O. Sepper  ·  arXiv preprint  ·  2026
Order book liquidity driven stress calibration and liquidation predictor.
arxiv.org/abs/2603.09164 →
Condensed Matter Physics
Possible Restoration of Superconductivity in the Quasi-One-Dimensional Conductor Li₀.₉Mo₆O₁₇ in Feasibly High Pulsed Magnetic Fields, H ≃ 100 T
O. Sepper, A.G. Lebed  ·  arXiv preprint  ·  2014
arxiv.org/abs/1411.0019 →
Quantum Limit in a Magnetic Field for Triplet Superconductivity in a Quasi-One-Dimensional Conductor
O. Sepper, A.G. Lebed  ·  arXiv preprint  ·  2014
arxiv.org/abs/1410.7420 →
Nodeless Versus Nodal Scenarios of Possible Triplet Superconductivity in the Quasi-One-Dimensional Layered Conductor Li₀.₉Mo₆O₁₇
O. Sepper, A.G. Lebed  ·  arXiv preprint  ·  2013
arxiv.org/abs/1310.0059 →
Phenomenological Approach to the Possible Existence of a Triplet Superconducting Phase in the Quasi-One-Dimensional Conductor Li₀.₉Mo₆O₁₇
O. Sepper, A.G. Lebed  ·  arXiv preprint  ·  2012
arxiv.org/abs/1211.1961 →
Four-Fold Anisotropy of the Parallel Upper Critical Magnetic Field in a Pure Layered d-Wave Superconductor at T = 0
A.G. Lebed, O. Sepper  ·  JETP Letters  ·  2020
Springer →
© 2026 SepperLabs