资产配置不是一道选择题：从 60/40 到全天候，再到 L1Asset allocation is not one answer: from 60/40 to All Weather and L1

这不是“哪个组合最好”的排行榜。60/40、永久组合、全天候、风险平价、1/N 和趋势覆盖各自解决的问题不同。真正决定结果的，是风险如何分配、规则如何执行、约束和成本如何处理，以及结果能否被归因和复核。This is not a league table of portfolios. The 60/40, Permanent Portfolio, All Weather, risk parity, 1/N and trend overlays solve different problems. What decides the outcome is how risk is allocated, how rules are implemented, how constraints and costs are handled, and whether results can be attributed and audited.

基准不是策略，是尺子。60/40 足够简单，但会在股债同跌时失效；没有统一基准，所谓“更好配置”无法判断。A benchmark is a yardstick, not a strategy. The 60/40 is simple and breaks when stocks and bonds fall together; without it, “better allocation” cannot be measured.
资金权重不等于风险权重。权益常在组合里承担远超其资金占比的风险，而全天候也可能把风险集中到长债。Capital weights are not risk weights. Equities often carry more risk than their capital share, while All Weather can concentrate risk in duration.
越复杂不等于越好。1/N 在多项测试中击败复杂优化器，说明估计误差和约束落地往往比公式更重要。More complex is not better by default. 1/N beats many optimisers, showing that estimation error and implementation often dominate the formula.
专题已经足够，但必须从文章集合升级成决策地图。现有内容已覆盖基准、替代方案、风险、执行和负结果；缺的是一个统一入口和 L1 这种端到端案例。The series is already sufficient, but it needs to become a decision map. It covers benchmarks, alternatives, risk, implementation and negative results; what was missing was one entry point and an end-to-end case such as L1.

4层决策：基准 / 风险 / 执行 / 治理decision layers: benchmark / risk / execution / governance

10篇研究，一条阅读路径studies in one reading path

221L1 月的影子组合记录months in the L1 shadow record

0真实资金；所有结果均已披露限制real money; every result discloses its limits

先选尺子Choose the yardstick

不要先问哪个组合最好，先问它要和什么比较、替谁承担风险。Do not start with the best portfolio. Start with the comparison and the risk owner.

研究Study 60/40 及其变体The 60/40 and its variants 最常用的基准，也是检验其他配置是否真的更好的第一把尺。The default baseline, and the first test any alternative must beat. 阅读笔记 →Read the note → 研究Study 永久组合The Permanent Portfolio 用简单四等分检验分散化、风险贡献与通胀环境。A simple four-way split tests diversification, risk and inflation. 阅读笔记 →Read the note →

再分风险Allocate the risk

资金权重只是表面。真正的配置问题是风险预算如何落到资产和因子上。Capital weights are cosmetic. The real allocation question is where the risk budget lands.

研究Study 全天候组合All Weather 按经济环境分配风险，但长债仍可能成为隐藏的风险中心。Risk is assigned across environments, but duration can remain the hidden concentration. 阅读笔记 →Read the note → 研究Study 风险平价Risk parity 把资金权重改成风险权重，收益并不会自动提高。Equal risk does not imply equal return. 阅读笔记 →Read the note → 研究Study 1/N 与优化器1/N versus optimisers 估计误差足以让复杂优化输给最简单的等权。Estimation error can make complex optimisation lose to a simple equal weight. 阅读笔记 →Read the note →

最后拼执行Execution decides

再好的目标权重，也要经过再平衡、约束和负债口径才能变成真实组合。A target weight becomes a portfolio only after rebalancing, constraints and the correct objective.

研究Study 再平衡Rebalancing 频率、漂移和成本共同决定规则是否值得执行。Frequency, drift and cost decide whether the rule is worth running. 阅读笔记 →Read the note → 研究Study 约束的代价The cost of constraints 机构约束既保护组合，也在同一风险预算下收费。Institutional constraints protect the portfolio and charge for it. 阅读笔记 →Read the note → 研究Study 负债驱动配置Liability-driven allocation 换一个分母，资产端的排序会完全改变。Change the denominator and the ranking of assets changes. 阅读笔记 →Read the note →

形成闭环Close the loop

最后必须把观点变成权重，把权重变成可归因的组合，并留下不可追溯修改的记录。Turn views into weights, weights into attributable results, and keep a record that cannot be rewritten.

研究Study 趋势跟踪Trend following 它更像危机保险：降低坏日子的损失，而不是稳定制造收益。More like crisis insurance than a return engine. 阅读笔记 →Read the note → 研究Study L1 多资产影子组合The L1 shadow portfolio 把 SAA、TAA、成本、风险归因和版本治理放进同一条流程。SAA, TAA, costs, risk attribution and version control in one process. 阅读笔记 →Read the note →

这个主题够不够？足够支撑一个完整的资产配置专题。它不是缺少更多策略，而是需要一条从“选择基准”到“执行治理”的阅读路径。继续堆叠新策略的边际价值，已经低于把现有研究连接起来、统一口径，并用一个持续更新的组合案例证明研究流程。Is the topic sufficient? Yes. It already supports a complete asset-allocation series. The missing piece was not more strategies, but a reading path from benchmark choice to implementation and governance. Adding another strategy now has less value than connecting the existing work, aligning conventions, and proving the process through one continuously updated portfolio case.

从文章走向完整流程：打开 L1，并查看五分钟英文投资陈述Move from articles to a full process: open L1 and the five-minute investment presentation L1 研究案例 →L1 case study → 五分钟英文陈述 →Five-minute presentation →

免责声明。本专题整理个人公开研究，不构成投资建议。各篇样本期、资产池和成本假设并不完全相同，不能把不同文章中的收益率直接横向排名；比较时必须先对齐口径。L1 为纸面组合，不含真实资金或真实 PnL。Disclaimer. This map organises personal public research and is not investment advice. The studies differ in sample period, universe and cost assumptions, so returns should not be ranked without first aligning definitions. L1 is a paper portfolio with no real money or PnL.