Transformer models and volatility regime shifts in crypto
How attention-based architectures spot the move from calm to turbulent markets earlier than classic volatility indicators, and where they still fall short.
Research
Reading from inside the market
Regular analysis from the Nyfcom research desk on machine learning in markets, the Canadian regulatory and tax landscape, and digital-asset infrastructure.
Latest articles
Crypto trading in Canada: what retail investors should check first
Registration, custody and disclosure: the questions to ask any platform before you fund an account, and where Canadians can verify the answers.
Bilingual NLP: reading English and French market news in real time
Why sentiment models trained only on English miss signals from Québec and European French-language sources, and how we close the gap.
Bitcoin after the halving: on-chain signals worth watching
Miner flows, long-term holder supply and exchange reserves: the network indicators that frame the current cycle.
Algorithmic discipline in falling markets
Rules-based exits, position caps and cool-down periods: how automation takes panic out of a drawdown.
Layer-2 networks and institutional capital
Arbitrum, Optimism and zkSync are cutting costs for large participants. What that means for liquidity, fees and settlement times.
Reinforcement learning and position sizing
Teaching an agent to size trades by expected risk rather than conviction, and the guardrails it still needs.
Crypto and your taxes: how the CRA treats digital assets
Capital gain or business income? An overview of how the Canada Revenue Agency looks at crypto dispositions, and why good records matter.
Detecting market manipulation without human supervision
Anomaly detection that flags spoofing, wash trading and pump-and-dump patterns before they can distort a signal.
Assessing risk in DeFi protocols
Smart-contract audits, oracle design and governance concentration: a practical checklist for evaluating DeFi exposure.