Crypto event impact with CoinGecko¶

This tutorial shows how to build a public crypto event case with CoinGecko price, market-cap, and volume data.

Crypto event share card

Why this case works¶

Crypto already has a large audience, and event windows naturally invite counterfactual questions: what would price or volume have looked like without the event?

One-command demo¶

python -m tscfbench demo crypto-event

Minimal example¶

from tscfbench.datasets import load_coingecko_market_chart, make_event_impact_case, to_log_returns

btc = load_coingecko_market_chart("bitcoin")
eth = load_coingecko_market_chart("ethereum")
sol = load_coingecko_market_chart("solana")

btc_r = to_log_returns(btc, value_col="price", out_col="value")[["date", "value"]].dropna()
eth_r = to_log_returns(eth, value_col="price", out_col="value")[["date", "value"]].dropna()
sol_r = to_log_returns(sol, value_col="price", out_col="value")[["date", "value"]].dropna()

case = make_event_impact_case(
    btc_r,
    {"eth": eth_r, "sol": sol_r},
    intervention_t="2024-01-11",
)

Notes¶

For markets, returns or spreads are usually more defensible than raw prices.
You can also use market_cap or total_volume as the treated series.