Open Interest indicator development

Open Interest Indicator Development

Open Interest (OI) is the total number of open futures contract positions. Unlike volume (which counts each trade), OI shows how many positions remain open. OI growth = new money entering market. OI decline = positions closing. This is one of the few genuinely leading indicators.

Open Interest Interpretation

High OI + sharp move = risk of large liquidation cascade. This is how harshest dumps happen: accumulated shorts or longs liquidate in chain reaction.

Getting OI Data

Binance Futures API

import httpx
from decimal import Decimal
import asyncio

class OpenInterestCollector:
    BINANCE_FUTURES = "https://fapi.binance.com"
    BYBIT_API = "https://api.bybit.com"

    async def get_binance_oi(self, symbol: str) -> dict:
        """Current OI and historical data"""
        async with httpx.AsyncClient() as client:
            # Current OI
            current = await client.get(
                f"{self.BINANCE_FUTURES}/fapi/v1/openInterest",
                params={"symbol": symbol}
            )
            current_data = current.json()

            # Historical OI (up to 500 points by 5-minute intervals)
            history = await client.get(
                f"{self.BINANCE_FUTURES}/futures/data/openInterestHist",
                params={
                    "symbol": symbol,
                    "period": "5m",
                    "limit": 500
                }
            )
            history_data = history.json()

        return {
            "current_oi": float(current_data["openInterest"]),
            "current_oi_usd": float(current_data["openInterest"]) * self.get_price(symbol),
            "history": [
                {
                    "time": h["timestamp"],
                    "oi": float(h["sumOpenInterest"]),
                    "oi_usd": float(h["sumOpenInterestValue"])
                }
                for h in history_data
            ]
        }

    async def get_aggregated_oi(self, symbol: str) -> dict:
        """Aggregate OI from all exchanges for single view"""
        tasks = [
            self.get_binance_oi(symbol),
            self.get_bybit_oi(symbol),
            self.get_okx_oi(symbol),
        ]
        results = await asyncio.gather(*tasks, return_exceptions=True)

        total_oi_usd = sum(
            r.get("current_oi_usd", 0)
            for r in results
            if not isinstance(r, Exception)
        )

        return {
            "total_oi_usd": total_oi_usd,
            "by_exchange": {
                "binance": results[0].get("current_oi_usd", 0) if not isinstance(results[0], Exception) else 0,
                "bybit": results[1].get("current_oi_usd", 0) if not isinstance(results[1], Exception) else 0,
                "okx": results[2].get("current_oi_usd", 0) if not isinstance(results[2], Exception) else 0,
            }
        }

OI Change Rate Indicator

OI change rate is often more important than absolute value:

def calculate_oi_change_rate(oi_history: list[dict], window: int = 12) -> list[dict]:
    """
    OI Change Rate: % change in OI over last N periods.
    Abnormal growth → potential liquidation event.
    """
    result = []
    for i in range(window, len(oi_history)):
        current_oi = oi_history[i]['oi_usd']
        past_oi = oi_history[i - window]['oi_usd']
        change_rate = (current_oi - past_oi) / past_oi * 100

        # Anomaly flag: change > 2 standard deviations
        all_changes = [
            (oi_history[j]['oi_usd'] - oi_history[j-window]['oi_usd']) / oi_history[j-window]['oi_usd'] * 100
            for j in range(window, i+1)
        ]
        mean_change = sum(all_changes) / len(all_changes)
        std_change = (sum((x - mean_change)**2 for x in all_changes) / len(all_changes)) ** 0.5
        z_score = (change_rate - mean_change) / std_change if std_change > 0 else 0

        result.append({
            'time': oi_history[i]['time'],
            'oi': current_oi,
            'change_rate': change_rate,
            'z_score': z_score,
            'is_anomaly': abs(z_score) > 2.0
        })

    return result

Liquidation Heatmap

Combining OI with price levels allows building liquidation heatmap — map of levels where liquidation orders concentrate:

def estimate_liquidation_levels(
    open_positions: list[dict],  # position data (avg leverage, side)
    current_price: float
) -> dict:
    """
    Estimate liquidation price levels.
    Data source: exchanges sometimes publish aggregated position level data.
    """
    long_liquidations = {}
    short_liquidations = {}

    for position in open_positions:
        if position['side'] == 'long':
            liq_price = position['avg_entry'] * (1 - 1/position['leverage'] + 0.005)
            bucket = round(liq_price / 100) * 100  # group by $100
            long_liquidations[bucket] = long_liquidations.get(bucket, 0) + position['size_usd']
        else:
            liq_price = position['avg_entry'] * (1 + 1/position['leverage'] - 0.005)
            bucket = round(liq_price / 100) * 100
            short_liquidations[bucket] = short_liquidations.get(bucket, 0) + position['size_usd']

    return {
        "long_liquidation_levels": long_liquidations,   # below current price
        "short_liquidation_levels": short_liquidations,  # above current price
        "largest_long_liq": max(long_liquidations.items(), key=lambda x: x[1]) if long_liquidations else None,
        "largest_short_liq": max(short_liquidations.items(), key=lambda x: x[1]) if short_liquidations else None
    }

OI-Price Correlation: Practical Patterns

Long squeeze: OI high, mostly longs → sharp price drop → longs liquidate → more drops. Signal: before move OI abnormally high, Funding Rate positive.

Short squeeze: opposite. Many shorts, price spikes up, shorts liquidate — rocket up.

OI accumulation in sideways: OI grows with flat price → market accumulating positions → will move (direction unknown beforehand, but explosive).

Monitoring OI together with Funding Rate and price gives much more complete picture of market state than technical analysis alone. Professional traders use this data as key inputs for decisions.