返回题库

冰山订单逻辑

Iceberg Order Logic

专题
Systems & Architecture / 系统与架构
难度
L2
来源
MyntBit

题目详情

冰山订单在量化金融和算法交易中用于执行大额交易而不扰动市场。它隐藏真实委托量，仅向公开订单簿暴露较小的"峰值"数量，成交后从隐藏储备中补充可见部分。

任务：实现 IcebergOrder 类，维护订单的可见量和隐藏量。当可见量被成交消耗后，自动从隐藏储备补充至峰值大小，直到隐藏量耗尽。

英文原题

Iceberg orders are crucial in quantitative finance and algorithmic trading for executing large trades without moving the market. They hide the true order quantity by exposing only a smaller "peak" size to the public order book, replenishing the visible part from a hidden reserve as trades execute.
Task
Implement an IcebergOrder class that maintains the visible and hidden quantities of an order.

  • IcebergOrder(double total_qty, double peak_size): Initializes the order. The initial visible quanti
解析

问题分析

冰山订单只显示总数量的一小部分（可见量），其余隐藏。当可见量被成交后，自动从隐藏量中补充新的可见量。用于大订单隐藏真实意图，减少市场冲击。

解法

class IcebergOrder {
    int total_qty_, visible_qty_, hidden_qty_, displayed_;
public:
    IcebergOrder(int total, int visible) : total_qty_(total), visible_qty_(visible), hidden_qty_(total - visible), displayed_(std::min(visible, total)) {}
    int displayed() const { return displayed_; }
    void onFill(int qty) {
        total_qty_ -= qty; displayed_ -= qty;
        if (displayed_ <= 0 && hidden_qty_ > 0) {
            int refill = std::min(visible_qty_, hidden_qty_);
            displayed_ += refill; hidden_qty_ -= refill;
        }
    }
    int remaining() const { return total_qty_; }
};

复杂度与边界

  • 时间复杂度：所有操作 O(1)
  • 边界条件：(1) 最后一次补充可能小于 visible_qty (2) 全部成交后 remaining=0 (3) 市价单可能一次吃掉多片

英文解析

Analysis

Iceberg orders display only a small portion (visible quantity) of the total order, keeping the remainder hidden. When the visible quantity is filled, a new visible slice is automatically replenished from the hidden reserve. Used to conceal true trading intent and reduce market impact for large orders.

Solution

class IcebergOrder {
    int total_qty_, visible_qty_, hidden_qty_, displayed_;
public:
    IcebergOrder(int total, int visible) : total_qty_(total), visible_qty_(visible), hidden_qty_(total - visible), displayed_(std::min(visible, total)) {}
    int displayed() const { return displayed_; }
    void onFill(int qty) {
        total_qty_ -= qty; displayed_ -= qty;
        if (displayed_ <= 0 && hidden_qty_ > 0) {
            int refill = std::min(visible_qty_, hidden_qty_);
            displayed_ += refill; hidden_qty_ -= refill;
        }
    }
    int remaining() const { return total_qty_; }
};

Complexity & Edge Cases

  • Time complexity: All operations O(1)
  • Edge cases: (1) Final replenishment may be smaller than visible_qty (2) All filled: remaining=0 (3) visible_qty >= total: behaves as regular order

Key Considerations

  1. Display quantity: Visible quantity must be small relative to total order size to minimize market impact; typical display is 5-10% of total quantity
  2. Refresh mechanism: When displayed quantity fills, iceberg refreshes from hidden reserve; refresh must not reveal total remaining quantity in book data
  3. Priority reset: Some exchanges reset price-time priority when iceberg refreshes — effectively reinserting at back of queue at same price level
  4. Detection risk: Iceberg patterns are detectable through repeated fills at same price level; sophisticated participants can infer hidden reserve size from fill frequency