Basis (usually referring to the perpetual price minus the index price) is the most direct “divergence reading” in perpetual markets. However, many discussions oversimplify basis as “if it’s expensive, it’ll fall; if it’s cheap, it’ll rise,” compressing a complex structure into a slogan. The more accurate approach is to treat basis as the pricing result of risk and constraints, then ask what forces shape it, whether it’s sustainable, and whether leverage and liquidation risk are layered in. The goal of this lesson is to provide a repeatable decomposition framework so readers can quickly judge whether “the divergence is a trend tax, compensatory premium, or a sign of systemic fragility” when combined with funding rates and OI congestion.
1. Definition and Observation Criteria for Basis: Aligning “What Are We Comparing?”
Before discussing basis, it’s crucial to clarify the objects of comparison:
- Index price represents a multi-source spot composite benchmark
- Perpetual price usually refers to the latest trade or mid-price
- Mark price is used for risk control and may not match the latest trade
Thus, “basis” has at least two common criteria: perpetual vs. index, perpetual vs. mark price. The differences can be significant during high-pressure moments. When researching basis, fix the criteria and timestamp to avoid mixing reference frames and causing misjudgment.
2. First Layer of Decomposition: Liquidity Premium and Execution Cost
High-frequency trading in perpetual markets doesn’t mean every price level is equally “executable.” When the order book thins, spreads widen, or concentrated order walls appear, prices become more sensitive to shocks. The resulting divergence often belongs to liquidity compensation: the market demands a higher premium for instant execution and large turnover.
This type of basis is characterized by: divergence accompanied by short-term shocks that don’t necessarily change mid-term trend direction. If basis expands in pulses but quickly mean-reverts and trading volume doesn’t remain elevated, it’s more a “market event” than a “structural supply-demand shift.”
3. Second Layer of Decomposition: Spot Availability, Borrowing Friction, and Arbitrage Channel Tightness
For certain assets, the connection between perpetuals and spot isn’t frictionless. Spot unavailability, slow transfers, staking lock-ups, and high borrowing costs all restrict arbitrage. When arbitrage is limited, the external force to “pull contracts back to the index” weakens, making perpetuals more likely to maintain non-zero basis over time.
Typical signals of this basis include: abnormal ratio between perpetual and spot trading volumes; persistent cross-exchange price spreads; long-term basis in one direction with “stickiness.” Here, basis is less about short-term bubbles and more an explicit channel cost. Beware: when the channel suddenly recovers (e.g., improved withdrawals, lower lending rates), basis can rapidly snap back and volatility surges.
4. Third Layer of Decomposition: Risk Appetite and Narrative-Driven Trend Phases
When the market enters strong trend narratives (macro easing, rising risk appetite, or sector-level catalysts), perpetuals often become the faster expression channel: price swings intensify, participation rises, leading to greater divergence from the index. In this phase, positive basis isn’t necessarily “irrational,” but more likely a risk premium trend capital is willing to pay: long holders accept higher costs for exposure.
To judge if basis is driven by a trend phase, focus on three things: whether spot volume surges in sync; whether risk appetite proxies resonate; whether perpetual basis expansion matches news/event schedules.
5. Structural Crowding: When Basis and Leverage Heat Up Together
The most trading-relevant state isn’t “whether basis exists,” but whether basis keeps expanding in a high-leverage and high-OI environment. When divergence grows alongside derivative leverage buildup, the system enters a more fragile zone: if reverse shocks or liquidity pullbacks occur, mean reversion can happen much more violently.
The key here isn’t predicting reversal timing but confirming risk radius expansion: price swings become more sensitive to identical news; passive liquidations become more likely; order books are prone to gaps.
6. Reading Basis as a Three-Stage Timeline: Initiation—Intensification—Exit
To reduce the impulse to “bet whenever you see divergence,” break basis changes into a simple timeline:
- Initiation: Divergence appears from nothing, often tied to events, spot supply bottlenecks, or shifts in risk appetite.
- Intensification: Divergence expands and lasts longer, often with OI growth, restricted arbitrage, and self-reinforcing trends.
- Exit: Divergence converges, possibly due to channel repair, leverage unwinding, or narrative fading.
At each stage, “reasonable trading insights” are completely different: initiation emphasizes evidence chain; intensification emphasizes risk control; exit emphasizes execution liquidity and speed of snap-back.
7. Common Misjudgments and Correction Points
Misjudgment 1: Shorting perpetuals just because of positive basis. Without channel evidence and a risk control plan, this usually incurs huge costs during trend continuation.
Misjudgment 2: Treating negative basis as “discount buy.” In panic phases, negative basis reflects liquidity breaks and risk appetite collapse—it’s not a safety cushion.
Misjudgment 3: Treating absolute basis thresholds as constants. Across stages, coins, and exchanges, basis centers can shift structurally; “historical threshold experience” often fails.
Summary
The core idea of Lesson 3 is: basis doesn’t signal simple expensive or cheap—it signals the market’s composite valuation of liquidity, availability, risk appetite, and arbitrage channels. First layer reads liquidity premium; second layer reads spot and lending friction; third layer reads trend and narrative; fourth layer aligns basis with leverage, OI, and order book depth to judge if systemic fragility is heating up. Only by decomposing basis to these levels can subsequent chapters form a closed loop with funding rates and crowding trade models—instead of replacing complete reasoning with a single number.
