deribit-options-chain

Deribit Options Chain (BTC / ETH / SOL)

Pull the full options chain from Deribit, normalize fields, and convert coin-denominated prices to USD for downstream skills.

---

Step 1: Pull via deribit-data

Use deribit-data to call /get_book_summary_by_currency?currency=BTC&kind=option — one request returns all live option instruments.

import requests, pandas as pd
r = requests.get("https://www.deribit.com/api/v2/public/get_book_summary_by_currency",
                 params={"currency": "BTC", "kind": "option"}, timeout=15)
r.raise_for_status()
chain = pd.DataFrame(r.json()["result"])

If the request fails, stop and surface the error. Do not fall back to cached data without telling the user the data is stale.

---

Step 2: Parse Instrument Name

Each instrument_name decodes to (coin, expiry, strike, type):

def parse_name(name):
    parts = name.split('-')  # e.g. ['BTC', '26DEC25', '120000', 'C']
    coin, exp_str, strike_str, opt_type = parts
    expiry = pd.to_datetime(exp_str, format='%d%b%y')
    return coin, expiry, int(strike_str), opt_type

Add columns expiry, strike, option_type, DTE (calendar days), T (years).

---

Step 3: Convert to USD

Deribit quotes premiums in the underlying. For downstream BS comparison or USD P&L:

chain['underlying'] = chain['underlying_price']     # spot at the quote time
chain['usd_mark']   = chain['mark_price']  * chain['underlying']
chain['usd_bid']    = chain['bid_price']   * chain['underlying']
chain['usd_ask']    = chain['ask_price']   * chain['underlying']
chain['usd_mid']    = (chain['usd_bid'] + chain['usd_ask']) / 2
chain['spread_pct'] = (chain['ask_price'] - chain['bid_price']) / chain['mark_price'].replace(0, float('nan'))
chain['moneyness']  = chain['strike'] / chain['underlying']

Critical: mark_iv from Deribit is already a decimal (0.65 = 65%). Do not re-divide by 100.

---

Step 4: Quality Filters

Filter	Why
Drop state == 'closed'	After-expiry instruments still show up
Drop bid_price == 0 or ask_price == 0	No two-sided market
Drop spread_pct > 0.30	Illiquid, mid is noise
Drop usd_mid < 5	Rounding noise dominates
Drop open_interest < 5 AND volume == 0	No interest

For 0DTE BTC chains, the filters are still strict because near-expiry deep OTM options trade for $0.00 to $0.05 BTC bids only. Don't relax filters silently — instead report "after filtering: N contracts survive out of M".

---

Step 5: Group by Expiry

Standard Deribit expiry cadence:

• Daily: most days (since 2023)
• Weekly: every Friday 08:00 UTC
• Monthly: last Friday of the month
• Quarterly: last Friday of Mar/Jun/Sep/Dec
• 0DTE: today, if listed

Top-of-mind expiries the user usually wants:

1. Next Friday (weekly)
2. Last Friday of current month (monthly)
3. Next quarterly (often most liquid)

If the user just says "the BTC chain", show 1, 2, 3 — and tell them what you picked.

---

Step 6: ATM Strip

For each expiry, identify the 5 strikes closest to spot (call + put). These are the most liquid and used for ATM IV, term-structure work, and most retail strategies.

def atm_strip(df, expiry, spot, n_strikes=5):
    ex = df[df['expiry'] == expiry].copy()
    ex['dist'] = (ex['strike'] - spot).abs()
    strikes = ex.sort_values('dist')['strike'].unique()[:n_strikes]
    return ex[ex['strike'].isin(strikes)].sort_values(['strike','option_type'])

---

Step 7: Greeks Pass-Through

Unlike yfinance, Deribit's /ticker returns Greeks for each instrument. For full chain Greeks, you'd need one /ticker call per instrument (rate-limited). Alternatives:

1. Recompute via inverse-options-pricing skill (faster, requires IV and spot).
2. Only fetch Greeks for the ATM strip via /ticker.

Caveat: Deribit Greeks are in coin units, not USD units. See inverse-options-pricing for conversion math. Do not mix them with equity-style Greeks without converting.

---

Step 8: Respond / Hand Off

Return the structured chain plus:

1. Spot and as-of timestamp.
2. Expiries picked + filter survival counts.
3. ATM strip preview.

Hand off to:

• inverse-options-pricing for theoretical pricing
• iv-surface (with crypto caveat — see that skill's notes) for fitting
• gex-max-pain for positioning
• options-payoff for trade visualization (use usd_mid as premium for inverse-quote conversion)