Skip to content

Reward Inflation Bugs

id: LS09H
title: Reward Inflation Bugs 
baseSeverity: H
category: tokenomics
language: solidity
blockchain: [ethereum]
impact: Excessive or infinite reward generation causing economic collapse
status: draft
complexity: medium
attack_vector: external
mitigation_difficulty: medium
versions: [">=0.4.0", "<0.8.21"]
cwe: CWE-682
swc: SWC-135

📝 Description

  • Reward inflation bugs occur when token rewards (for staking, farming, or liquidity provision) are over-allocated, unbounded, or miscalculated due to logic flaws or missing constraints.
  • This leads to:
  • Exponential or incorrect reward issuance,
  • Devaluation of the native token,
  • Potential collapse of the economic model or drain of reward pools.
  • Often the cause lies in improperly updating accRewardPerShare, incorrect time tracking, or failing to cap multipliers and emissions.

🚨 Vulnerable Code

contract InflationBug {
    mapping(address => uint256) public staked;
    mapping(address => uint256) public rewardDebt;
    uint256 public accRewardPerShare;
    uint256 public lastRewardTime;
    uint256 public rewardRate = 10 ether; // per second

    function updatePool() public {
        uint256 timeElapsed = block.timestamp - lastRewardTime;
        accRewardPerShare += timeElapsed * rewardRate; // ❌ no scaling or bounds
        lastRewardTime = block.timestamp;
    }

    function pendingReward(address user) public view returns (uint256) {
        return staked[user] * accRewardPerShare - rewardDebt[user];
    }
}

🧪 Exploit Scenario

Step-by-step exploit process:

  1. Contract has rewardRate = 10 ether, no upper limit, and is idle for 10 days.
  2. A user stakes right before updatePool() is called.
  3. timeElapsed is large, resulting in massive inflation of accRewardPerShare.
  4. The user claims rewards worth thousands of tokens for a minimal stake.
  5. Reward pool is drained, and token price crashes due to excess emission.

Assumptions:

  • No upper bound or time capping logic in reward multiplier.
  • accRewardPerShare is not scaled or protected from overflow.
  • Pool can sit idle for long periods.

✅ Fixed Code

function updatePool() public {
    uint256 currentTime = block.timestamp;
    if (currentTime > lastRewardTime) {
        uint256 timeElapsed = currentTime - lastRewardTime;

        // ✅ Cap the max multiplier to prevent excessive reward jumps
        if (timeElapsed > 86400) {
            timeElapsed = 86400; // max 1 day
        }

        accRewardPerShare += (timeElapsed * rewardRate) / 1e18;
        lastRewardTime = currentTime;
    }
}

🧭 Contextual Severity

- context: "Default"
  severity: H
  reasoning: "Inflated token supply harms protocol economics and trust."
- context: "DeFi protocol with governance token rewards"
  severity: C
  reasoning: "Governance control and tokenomics may be irreparably harmed."
- context: "Centralized reward distributor with capped pool"
  severity: M
  reasoning: "Limits and manual oversight may mitigate ongoing inflation."

🛡️ Prevention

Primary Defenses

  • Always scale accRewardPerShare to prevent overflows (e.g., divide by 1e18).
  • Cap time-based multipliers to prevent abuse from idle pools.
  • Use safeRewardRate, maxEmissionRate, or MAX_DURATION constants.

Additional Safeguards

  • Emit UpdatePool logs with timestamps for off-chain tracking.
  • Audit the reward formula for linearity and expected inflation under edge cases.
  • Use automated tests to simulate long idle durations and reward cliffs.

Detection Methods

  • Slither: custom detectors for accRewardPerShare growth and unchecked multipliers.
  • Formal modeling or fuzzing reward logic under long time intervals.
  • Unit test suites simulating multi-day inactivity and claim bursts.

🕰️ Historical Exploits

  • Name: Mango Markets Exploit
  • Date: 2022-10-11
  • Loss: Approximately $114 million
  • Post-mortem: Link to post-mortem

📚 Further Reading

✅ Vulnerability Report

id: LS09H
title: Reward Inflation Bugs 
severity: H
score:
impact: 5         
exploitability: 4 
reachability: 4   
complexity: 3     
detectability: 3  
finalScore: 4.2

📄 Justifications & Analysis

  • Impact: Can destroy protocol tokenomics by overpaying users or draining reward pools.
  • Exploitability: Easily triggered after periods of low usage or deployment delay.
  • Reachability: Appears in nearly all staking, farming, or yield protocols.
  • Complexity: Moderate — attacker just needs timing or minimal capital.
  • Detectability: Static tools may miss it; simulation or test coverage often required.