Transformer models and the early detection of volatility regimes
Why attention-based networks recognise a shift in market regime sooner than older recurrent designs.
Journal
Insight from inside the markets
Regular analysis from the Havnorth research desk on machine learning, digital-asset infrastructure and the UK regulatory landscape.
Latest
Recent articles
The UK cryptoasset regime: what changes for retail traders
Promotion rules, authorisation and disclosure: how the evolving UK framework is reshaping the retail crypto market.
Reading the Bank of England with natural-language models
How language models parse MPC minutes and speeches to flag a change in tone before sterling reprices.
Bitcoin beyond the halving: what on-chain data is signalling
Wallet flows, miner behaviour and exchange balances sketch the likely paths for the months ahead.
Why algorithms keep their nerve in falling markets
Rule-based models stay disciplined precisely when fear clouds human judgement.
Layer-two networks and the arrival of institutional money
Arbitrum, Optimism and zkSync are attracting a growing share of professional liquidity.
Reinforcement learning and the art of position sizing
How adaptive models scale exposure to volatility while respecting a fixed risk ceiling.
Sterling stablecoins and the future of settlement
Pound-denominated tokens could change how British businesses and investors move money across borders.
Spotting market manipulation without a human in the loop
Clustering techniques flag wash trading and spoofing before they distort prices.
Weighing up the risks of DeFi protocols
Headline yields come with smart-contract, liquidity and regulatory risks that deserve careful scrutiny.
Trade with deeper understanding
Put these insights to work in your Havnorth account.
Open an account