Market analysis, product updates, and trading strategies from the Blockcircle team.
Most liquidity dashboards average the big central banks as if they move markets equally. Here is how I weight the inputs so the composite tracks real transmission instead of an arbitrary sum.
A 30-day rolling Sharpe running from 0.64 down past -5.5 looks like a collapse in returns. Half of it can be the volatility denominator shrinking, and here is the arithmetic that separates the two.
A single full-sample correlation number hides that the relationship drifts and flips sign. Here is how to roll it, pick a sane lookback, and read the series before a hedge breaks.
A traded probability is not a scenario weight until it has survived three haircuts: who was allowed to set it, what it costs to act on it, and how badly the tails are distorted. The screener price is where that work starts, not where it ends.
A hands-on way to separate directional markets from range-bound noise using three measures that each fail differently, then combine them into one honest trend-or-chop label.
Members hold the big employer down the road, the same way everyone does. The heatmap has no state view, so here is the manual check and the two tests that separate ordinary home bias from something informational.
Equal weighting sounds fair, but it lets the 1-hour whipsaw drown out the daily trend. Here is a taper toward the higher timeframes, and the reasoning behind each coefficient so you can tune your own.
A cross-timeframe relative strength grid gives conflicting answers on purpose. Here is how I read daily, weekly, and monthly RS as one picture instead of chasing whatever is green today.
An import is a data migration and its failures are silent. Repainting, intrabar fill assumptions and a commission field left at zero all survive the crossing, and three matched trades tell you which one bit.
Standard deviation punishes big up days as harshly as big down days. Here is how downside deviation and the Sortino ratio fix that, worked through a crypto return series.
A step-by-step walkthrough of regressing a token's excess returns against a benchmark to estimate beta, and why one full-sample number quietly lies to you.