Model Lab
Wavelet + LSTM + ARIMA Fusion Framework
Reduce order-flow noise, separate trend and volatility, and improve signal stability.
Paper · PDFRead paper
Institutional Digital Asset Research
Verifiable research process, disciplined risk control, and institutional-grade crypto strategy access.
PureGamma Research builds cross-market, cross-horizon digital asset portfolios through structured signal research, controlled execution, and transparent review materials for qualified partners.
2017
Operating history
Multi-Horizon
Research engine
Monthly
Risk review cadence
About PG Research
We operate across digital assets, derivatives, and structured portfolios, with research and engineering as core advantages.
Market Microstructure
We monitor order flow, depth, funding, and on-chain behavior to detect structural dislocations and liquidity pulses.
Hedging Framework
We manage drawdowns with basis trades, cross-asset hedges, and volatility strategies to build layered defenses.
Research-Driven Iteration
Experimentation, post-trade review, and deployment form a closed loop to sustain long-term edge.
Institutional Trust Stack
A clearer due-diligence path for partners who need process evidence, operational boundaries, and risk governance before allocation or API integration.
GovernanceRisk-first mandate
VerificationMonthly review trail
AccessSMA / API pathways
DisclosureSuitability boundary
Strategy System
From multi-scale signal fusion to portfolio construction, we target stable risk-adjusted returns.
Closed-Loop Research System
Observe, model, allocate, and review.
-
Signal Generation
Multi-Scale CWT + LSTM/ARIMAWavelet decomposition isolates noise while statistical models and deep networks capture trends and sentiment shifts.
-
Signal Fusion
Dynamic Weights + RSI ThresholdsSignals are blended across horizons, with momentum filters reducing false triggers.
-
Portfolio Management
Cointegration Filters + Layered HedgesCross-asset portfolios are monitored for cointegration drift and dynamically re-hedged.
-
Risk Governance
Stress Filters + Post-Trade ReviewDrawdown, exposure, and execution evidence feed directly into the next research cycle.
Feedback loop
Risk review updates model assumptions, portfolio constraints, and the next validation agenda.
Research Desk
A sustainable content system for market views, model notes, and institutional updates.
Tail Risk Research
Pre-asymptotic inference, epistemic fat tails, and model fragility.
Inspired by Nassim Nicholas Taleb's work on uncertainty, this desk treats market evidence as incomplete before convergence, studies how model error compounds in fat-tailed domains, and reviews where normal assumptions understate downside exposure.
Pre-asymptotic review
Epistemic tail exposure
Convexity and ruin filters
Featured paper · 10-page PDF The Fat-Tail MSTR Effect Open paper reader
If the embedded reader is unavailable, open the PDF directly.
Research archive
Focus
Format / access
Risk Review
Attribution and Drawdown Notes
Package monthly net value, drawdown context, hedge behavior, and execution quality into a repeatable partner update.
Monthly · RequestRequest sample
Framework
Team, Strategy, and Research Framework
Background, process, portfolio construction, and partner access pathways.
Deck · DocSendOpen materials
Repository
Machine Learning for Structured Data
Feature engineering, model selection, and interpretability for structured datasets.
Code · GitHubView repository
Research Methodology
Investment is a repeatable process: observe, hypothesize, validate, review, and execute.
Observation & Hypothesis
Build testable hypotheses around order flow, on-chain behavior, and macro variables.
Validation & Evaluation
Backtests and scenario stress tests assess robustness and limits.
Execution & Review
Real-time monitoring with risk thresholds and structured post-trade reviews.
Strategic Partners
Open Partner Directory
Net Value Curve
Latest NAV
2.5073
Cumulative
+150.72%
Max Drawdown
-9.42%
Monthly performance pack with calculation basis, fee assumptions, and benchmark context.
Maximum drawdown, exposure bands, hedge contribution, and stress-period commentary.
Exchange/broker reporting, SMA/API account boundaries, and post-trade review notes where applicable.
Past performance is not indicative of future results. Digital asset strategies may lose substantial value.
Institutional Access
Research Consultation
For funds, founders, and infrastructure partners seeking a structured discussion on strategy research, portfolio construction, diligence, or partnership fit.
Institutional Contact
For SMA/API partnerships, diligence materials, consultation scheduling, or research collaboration, send a complete message and the team will respond shortly.
BP Partnerships: haoran@pgresearch.org
LP / Consultation: chris@pgresearch.org
Independent Research Support
Support the Research Desk
For readers who value the public research and want to make a voluntary contribution. Support does not create an advisory, management, or client relationship.
This website is for informational purposes only and does not constitute investment advice or an offer.
Digital asset trading involves substantial risk; past performance is not indicative of future results.