Open interest is the total number of outstanding futures contracts that have not been settled. Every futures trade has a buyer and a seller. When a new buyer opens a long and a new seller opens a short, open interest increases by one contract. When an existing long closes against an existing short, OI decreases by one. OI measures the total amount of money committed to directional bets in the futures market, and its changes relative to price movement reveal the nature and strength of the current trend.
The Four Quadrants of OI and Price
The relationship between OI changes and price changes creates a 2x2 matrix that every derivatives trader should internalize. Each quadrant tells a different story about who is entering and exiting the market.
Rising OI with rising price means new longs are entering the market. Fresh capital is flowing in on the buy side, which supports the uptrend. This is the strongest bullish configuration because both price and participation are increasing. The trend has conviction behind it and is likely to continue until something changes the flow dynamic.
Rising OI with falling price means new shorts are entering. Capital is flowing in on the sell side, putting downward pressure on price. This is aggressively bearish because sellers are not just taking profits on existing longs, they are opening new positions to bet on further decline. The downtrend has conviction.
Falling OI with rising price is a short squeeze or short covering rally. Existing short holders are closing their positions by buying back contracts, which pushes price up but reduces total market participation. This is a weak rally because no new long capital is entering. The price increase is driven by forced buying from shorts exiting, not genuine bullish conviction. These moves tend to be sharp but short-lived.
Falling OI with falling price is long liquidation or long capitulation. Existing long holders are closing positions, either voluntarily or through forced liquidation. Price falls as these longs sell, and OI declines because contracts are being closed rather than opened. This is the exhaust phase of a downtrend, and it often marks the final washout before a bottom forms.
Why OI Matters More Than Volume
Volume tells you how much trading activity happened in a period. But volume does not distinguish between new positions being opened and existing positions being closed. A million dollars of volume could mean a million dollars of new money entering the market, or it could mean existing positions shuffling between participants without any new capital commitment.
OI makes this distinction. When volume is high and OI is rising, new money is entering. When volume is high and OI is falling, existing participants are exiting. This difference is fundamental for understanding whether a move has staying power or is running on fumes. A breakout accompanied by rising OI is far more likely to follow through than a breakout on flat or falling OI.
How QuantForge Collects OI Data
We source open interest data from two providers. Binance provides OI through their REST API, but with a significant limitation: the historical data only goes back approximately 30 days. For backtesting purposes, 30 days is insufficient. We supplement with Coinglass, which provides approximately six months of OI history at 4-hour intervals on their Hobbyist plan.
The API server runs a cron job every 4 hours that fetches the latest OI readings from Binance and stores them in the open_interest table. Each record contains the symbol, timestamp, and the total open interest value in USD notional terms. A unique constraint on the combination of exchange, symbol, and timestamp prevents duplicate entries. Over time, the database accumulates a growing history that extends beyond what either API provides in a single request.
For the backtest engine, OI data is loaded alongside price candles and made available to strategies through the standard data dictionary. The engine slices the data with strict no-lookahead guarantees, ensuring each bar only sees OI values that existed at that point in time. Strategies access OI through a simple key lookup, treating it as another time series alongside price and volume.
Our OI Momentum Strategy
We built an oi_momentum strategy that uses the rate of change (ROC) of open interest over a 14-period lookback window. When OI rises by more than 2.4 percent over 14 hourly periods, the market is seeing significant new capital commitment. The strategy then checks price direction to determine which quadrant of the OI-price matrix applies and generates signals accordingly.
The sweep optimization found Sharpe 3.34 on the best symbol. However, validation across three sub-periods showed that the edge was inconsistent. Only INJ showed positive results in 2 of 3 periods. The strategy did not earn a ROBUST verdict on any symbol.
The failure taught us that OI momentum alone has too many false positives. A 2.4 percent OI spike could mean aggressive new longs entering, aggressive new shorts entering, or simply a large hedging transaction from an institutional desk. Without additional context from funding rates and positioning data, the OI signal is ambiguous.
OI Spikes and Leverage Buildup
The most actionable OI signal is not the direction of OI change but the magnitude of the buildup. When OI rises sharply relative to its recent history, the market is becoming increasingly leveraged. Highly leveraged markets are fragile because a price move in either direction will trigger margin calls and forced liquidations, which amplify the move further.
Our leverage_composite strategy uses OI threshold detection as one of its three components. When OI momentum exceeds the threshold, it contributes one point to the composite score. Combined with extreme funding and crowded positioning, elevated OI signals that the conditions are ripe for a leveraged unwind.
The threshold calibration is important. We tested values from 0.01 to 0.05 (1 percent to 5 percent OI change over 14 periods). Values below 0.02 generated too many signals, most of which were noise from normal market activity. Values above 0.04 generated too few signals, missing genuine leverage buildup events. The sweet spot at 0.024 captures meaningful OI spikes while filtering out routine fluctuations.
Practical Considerations
OI data has several quirks that traders should understand. First, OI is reported in contract terms, not dollar terms, on some exchanges. Binance reports in USDT notional, which is what we use. Second, OI can decrease when an exchange liquidates positions, which creates a specific OI-price pattern (falling OI, falling price) that signals forced selling rather than voluntary exit. Third, OI on different exchanges can diverge significantly, so single-exchange OI may not capture the full picture of market leverage.
For our deployment, we use Binance OI as the primary data source because Binance dominates perpetual futures volume, accounting for roughly 50 to 60 percent of total crypto derivatives trading. Coinglass provides aggregated OI across multiple exchanges for a more comprehensive view, which we use in backtesting. The gap between single-exchange and aggregated OI is something to monitor but has not materially affected our strategy performance in testing.