标题： 学习基于模式匹配的零成本投资组合选择 显示英文标题

标题： Learning zero-cost portfolio selection with pattern matching

摘要： 我们考虑并扩展了Györfi{\it 等}的对抗性基于代理的学习方法，以适用于零成本投资组合选择的情况，该情况通过从共同基金分离定理导出的二次近似实现。 该算法应用于每日采样的序列开盘-最高-最低-收盘数据以及来自约翰内斯堡证券交易所（JSE）的序列日内5分钟柱状数据。 考虑了算法的统计测试。 将算法直接与先前文献中的标准纽约证券交易所测试案例进行比较。 学习算法用于选择由模式匹配过去动态生成的代理（或专家）的参数，使用简单的最近邻搜索算法。 表明使用共同基金分离定理的解析解存在速度优势。 认为性能的预期损失不会削弱其在日内量化交易中的潜在应用，并且当考虑交易成本和滑点时，策略在未杠杆的情况下仍可能保持盈利。 本文展示了可以在集体层面上系统地利用JSE上的金融时间序列模式，但这并不意味着个体资产时间序列本身具有可预测性。 显示英文摘要

摘要： We consider and extend the adversarial agent-based learning approach of Gy{\"o}rfi {\it et al} to the situation of zero-cost portfolio selection implemented with a quadratic approximation derived from the mutual fund separation theorems. The algorithm is applied to daily sampled sequential Open-High-Low-Close data and sequential intraday 5-minute bar-data from the Johannesburg Stock Exchange (JSE). Statistical tests of the algorithms are considered. The algorithms are directly compared to standard NYSE test cases from prior literature. The learning algorithm is used to select parameters for agents (or experts) generated by pattern matching past dynamics using a simple nearest-neighbour search algorithm. It is shown that there is a speed advantage associated with using an analytic solution of the mutual fund separation theorems. It is argued that the expected loss in performance does not undermine the potential application to intraday quantitative trading and that when transactions costs and slippage are considered the strategies can still remain profitable when unleveraged. The paper demonstrates that patterns in financial time-series on the JSE can be systematically exploited in collective but that this does not imply predictability of the individual asset time-series themselves.