Transformer models and regime detection in crypto markets
Why attention-based networks spot a shift in volatility regime earlier than recurrent models.
Research
Insight from inside the market
Regular analysis from the Nimwold research desk on machine learning in markets, digital-asset infrastructure and the evolving UK rulebook.
Latest analysis
The UK cryptoasset regime: what changes for retail investors
How the FCA's framework for trading platforms and promotions is shaping a more transparent market.
Reading UK financial news with language models
Sentiment models scan headlines and social feeds to raise early alerts before prices react.
Bitcoin after the halving: what on-chain data says now
Wallet flows and miner behaviour point to the likely paths for the next phase of the cycle.
Algorithmic discipline in a falling market
Rules-based models stick to the plan when fear takes over human judgement.
Layer-two networks and the institutional money flow
Arbitrum, Optimism and zkSync are attracting a growing share of professional liquidity.
Reinforcement learning and smarter position sizing
Models that scale exposure to market volatility and to the risk cap you set.
Sterling stablecoins and the future of UK payments
Pound-backed tokens could reshape settlement and cross-border transfers.
Spotting market manipulation without human review
Clustering algorithms flag wash trading ahead of sharp price swings.
Keeping crypto records for HMRC: a practical checklist
What to log on every trade so that Capital Gains Tax reporting stays straightforward.