Hi Dev.to. I'm Radames Belfort, a financial researcher (ex-BlackRock) now building quantitative frameworks in Brazil.
Most people look at price charts. I prefer looking at the raw data feed. Today (Dec 18, 2025), the market provided a perfect dataset for studying "Market Microstructure" anomalies.
The Data Problem We have two correlated assets: BTC and ETH.
BTC_Price is consolidating.
ETH_Price is trending down (~$2833).
Sentiment_Index = 17 (Extreme Fear).
The Algorithmic Signal When querying the exchange APIs (e.g., via CCXT in Python) for fundingRate, we see a divergence:
BTC_Funding > 0 (Positive)
ETH_Funding <= 0 (Neutral/Negative)
Code Logic & Interpretation In a standard mean-reversion bot, this signal often triggers a "Wait" state. Why? Because the correlation is breaking. The "Cost of Carry" (Funding) for BTC implies bullishness/complacency, while the price action of ETH implies capitulation.
Writing code that simply executes on RSI < 30 is dangerous in this environment. A robust risk engine needs to ingest OrderBook_Depth and Funding_Rate as primary constraints. If liquidity (depth) thins out while funding diverges, the probability of slippage increases exponentially.
I am currently working on standardizing these metrics for a new investment firm launching in 2026. Data hygiene is the first step to alpha.
Top comments (0)