返回投资研究台 2026-07-26

research thread Brief

  • 研究线程标识: [source-id-redacted]
  • Shared Agent Route session: [source-id-redacted]
  • Root topic: KB 刷新追踪: 棉花 · auto-spawn 1
  • Current 研究记录序号: 10/10
  • Current topic: USD/CNH 隐含波动率反弹对预建 Call Spread 成本结构的压力测试
  • Current question: 若 1M Implied Volatility 从当前的 3.8% 低位显著回升,USD/CNH call spread 的 Vega 风险是否会突破原定的 32%/48% 预算上限?
  • Current stance: stress-test

Completed Cards

既有研究记录

  • Analyst: 消费行业分析师
  • Stance: initial
  • Topic: 我否定 8.0-9.5 million bales 作为 2026/27 全球棉花缺口基准,除非后续天气造成至少 3.31 million bales 的额外净损失。
  • Question: Investigate the topic "我否定 8.0-9.5 million bales 作为 2026/27 全球棉花缺口基准,除非后续天气造成至少 3.31 million bales 的额外净损失。" and make the strongest evidence-based judgment you can.
  • Summary: 截至 2026-07-26,USDA 7 月 2026/27 棉花平衡表对应全球产需缺口约 4.69 million bales,而非 8.0-9.5 million bales。天气损失已经部分进入基准:印度单产、澳大利亚水资源、巴基斯坦单产和美国弃收率均已反映压力。若要达到 8.0-9.5 million bales,还需新增 3.31-4.81 million bales 净收紧。需求端纺用量稳健但受化纤竞争限制,难以独立推动缺口扩张。下一步应由 agri-analyst 验证美国弃收、印度季风、澳大利亚水资源和巴基斯坦天气/虫害的额外减产空间。
  • Handoff: agri-analyst | 2026/27 棉花天气减产的额外空间 | 美国弃收率、印度季风/单产、澳大利亚水资源和巴基斯坦天气/虫害能否共同带来至少 3.31 million bales 的额外净减产?

既有研究记录

  • Analyst: 农业行业分析师
  • Stance: stress-test
  • Topic: 2026/27 棉花天气减产的额外空间
  • Question: 美国弃收率、印度季风/单产、澳大利亚水资源和巴基斯坦天气/虫害能否共同带来至少 3.31 million bales 的额外净减产?
  • Summary: 自下而上对四产区做压力测试:概率加权额外净减产约 +1.2 至 +2.5 million bales,对应缺口 5.9-7.2M,够不着 8.0。+3.31M 门槛只在四国同时向坏且澳大利亚未对冲的叠加尾部才成立。基准情景够不着的最大原因是澳大利亚——ERS 记 3.0M,但 FAS 因北墨累-达令水改善预测 4.9M,更可能是约 −0.4M 上行对冲。美国双向(干旱监测改善、TAMU 提示弃收可能更低),巴基斯坦已在 40 年低基数、增量空间有限。印度(92% 季风+红铃虫)是最可信的单一增量 +0.8 至 +1.4M,但独木难支。支持前序研究对基准的否定,同时确认天气尾部真实可量化。
  • Handoff: china-macro | 中国储备棉与新疆产量对全球棉花天气缺口的缓冲 | 中国国家储备棉轮出与新疆稳产、滑准税进口调节能否吸收 3-4 million bales 的全球天气减产,从而抑制缺口向棉价的传导?

既有研究记录

  • Analyst: 全球宏观分析师
  • Stance: stress-test
  • Topic: 中国储备棉与新疆产量对全球棉花天气缺口的缓冲
  • Question: 中国国家储备棉轮出与新疆稳产、滑准税进口调节能否吸收 3-4 million bales 的全球天气减产,从而抑制缺口向棉价的传导?
  • Summary: 截至 2026-07-26,我部分否定“完全吸收”命题:中国缓冲真实存在,但更像近月价格减震器,不是 3-4M 全球减产的完全吸收器。USDA 表中中国 2026/27 年末库存为 35.487M bales,约占全球 49.8%;但可轮出国家储备规模未正式披露。新疆稳产降低中国恐慌进口概率,但 FAS 预计 MY26/27 中国产量 7.2MMT、下降 4.5%。2026 年 TRQ 89.4 万吨、滑准税加工贸易配额 30 万吨,多数属于基准进口制度。若储备棉按约 8,000 吨/日投放,抵消 3.31M bales 需约 90 个销售日。
  • Handoff: asset-allocator | 中国缓冲后的棉花价格传导与商品配置含义 | 若中国储备棉轮出压住近月现货但把冲击推迟到补库周期,ICE 棉花、纺织链利润和商品组合应如何定价 3-4M 天气尾部?

既有研究记录

  • Analyst: 首席策略师
  • Stance: synthesize
  • Topic: 中国缓冲改变的是棉价传导时点而非天气尾部本身,最优表达是远月ICE凸性加纺织链利润分化,而不是追近月现货。
  • Question: Investigate the topic "中国缓冲改变的是棉价传导时点而非天气尾部本身,最优表达是远月ICE凸性加纺织链利润分化,而不是追近月现货。" and make the strongest evidence-based judgment you can.
  • Summary: 截至2026-07-26,我综合前序观点:中国储备棉轮出能压住近月国内现货和普通纱成本,但不能消除ICE远月天气尾部。USDA 7月基准显示2026/27全球库存消费比为58.4%;若额外天气损失3.0-4.0 million bales,一比一测算会降至55.9%-55.1%。储备投放约8,000吨/日,对冲3.0-4.0 million bales需约82-109个销售日,因此更像缓冲而非完全吸收。配置上应少做近月beta,多用Mar/May-2027 ICE期权或日历结构,并在A股纺织链中偏向有储备棉获取能力的国内纺企、低配常规进口纱敞口。
  • Handoff: asset-allocator | 中国缓冲后的棉花远月期权与跨商品风险预算 | 在中国储备棉压住近月但保留远月天气尾部的情景下,ICE Mar/May-2027 棉花期权相对其他天气敏感商品应获得多少组合风险预算?

既有研究记录

  • Analyst: 首席策略师
  • Stance: synthesize
  • Topic: 中国缓冲把棉花变成风险调整后最优的单一天气尾部表达(便宜波动率+免费下行地板),但一场超级 El Niño 使整个天气篮子相关,故棉花只应拿受限 3–5% NAV sleeve 的约 23%(约 0.9% NAV),而非整笔交易。
  • Question: Investigate the topic "中国缓冲把棉花变成风险调整后最优的单一天气尾部表达(便宜波动率+免费下行地板),但一场超级 El Niño 使整个天气篮子相关,故棉花只应拿受限 3–5% NAV sleeve 的约 23%(约 0.9% NAV),而非整笔交易。" and make the strongest evidence-based judgment you can.
  • Summary: 综合前四卡后回答定量问题:在天气尾部多头凸性 sleeve 中,ICE Mar/May-2027 棉花期权应中等超配、非主导——约占 sleeve 的 20–25%、折合约 NAV 的 0.8–1.0% 权利金风险敞口。关键新事实是 NOAA CPC 预示 2026-27 冬季极强/超级 El Niño(96–98% 持续、63–81% 极强),它把天气尾部变成跨商品共同因子,使棉花、咖啡、可可、白糖的尾部正相关、无法互相分散,因此必须按篮子(相关 VaR)封顶在 3–5% NAV。棉花凭借便宜 IV(约 20%)、中国储备缓冲提供的免费下行地板、以及新疆带来的特异性,拿到打分卡最高分(4.00/17.5≈23%)。同时以此叠加否定 th_p_3339c1e9:El Niño 之夏对美国粮食偏利,玉米远月天气溢价更可能流失。
  • Handoff: derivatives-strategist | ICE Mar/May-2027 棉花期权的具体结构、偏斜与日历套利设计 | 在约 0.9% NAV 的棉花凸性预算内,应如何具体构建 CTH27/CTK27 期权结构——看涨价差、看涨阶梯还是日历价差?应利用远月相对近月的期限结构与看涨偏斜到什么程度,且如何让被中国储备缓冲压住的下行为凸性融资、同时把 El Niño 相关篮子风险控制在 sleeve 上限内?

既有研究记录

  • Analyst: 全球宏观分析师
  • Stance: support
  • Topic: ICE Mar/May-2027 棉花期权的具体结构、偏斜与日历套利设计
  • Question: Investigate the topic "ICE Mar/May-2027 棉花期权的具体结构、偏斜与日历套利设计" and make the strongest evidence-based judgment you can.
  • Summary: 在约 0.9% NAV 棉花凸性预算内,推荐核心为有限风险的 CTH27 看涨价差(示例 90/110),由单腿 68 put 卖出(中国储备地板)融资掉约 30-35% 毛权利金,外加小型 CTZ26/CTK27 看涨日历卫星。明确否定裸露看涨阶梯 1×2 作为核心——超级 El Niño 是右尾供给冲击,卖第二腿看涨即做空要持有的尾巴。对期限结构(持 Mar/May 而非 Dec)与看涨偏斜(短腿≥108¢)均适度利用。储备看跌是全篮唯一卖出看跌腿,卖出名义≤0.5% NAV,另留约15%作备用弹药,以相关 VaR(ρ≈0.6-0.8)将软商品篮封在≤5% NAV。整体以约0.9% NAV买入最高约4.8% NAV右尾毛收益。参考:CTH27 平值 IV 23.69%,Mar-27 约83.2¢。
  • Handoff: fx-strategist | 中国储备地板与 CNY 政策对棉花看跌卖出融资腿的支撑力度及断裂情景 | 本结构的凸性融资完全依赖卖出的 CTH27 68 put,而其地板由中国储备轮出支撑;在何种 CNY 走势与储备管理动机变化下该地板会破裂,使卖出看跌与 El Niño 相关篮子尾部同向受损?请给出 USD/CNH 情景阈值与对冲建议,并对照 th_p_2da11237。

既有研究记录

  • Analyst: 全球宏观分析师
  • Stance: stress-test
  • Topic: 中国储备地板不是无条件存在:USD/CNH 低于 7.20 时仍可小规模融资,7.20-7.30 必须降风险,7.30-7.35 叠加储备动机转向进口替代时应回补或封顶 CTH27 68 put。
  • Question: Investigate the topic "中国储备地板不是无条件存在:USD/CNH 低于 7.20 时仍可小规模融资,7.20-7.30 必须降风险,7.30-7.35 叠加储备动机转向进口替代时应回补或封顶 CTH27 68 put。" and make the strongest evidence-based judgment you can.
  • Summary: 截至 2026-07-27,本报告对 CTH27 68 put 融资腿做压力测试:USD/CNH 低于 7.20 时,中国储备公式与 PBOC 稳汇率目标仍能支撑地板;7.20-7.30 为黄色区,应停止新增裸卖 put 并削减 delta;7.30-7.35 叠加储备轮出放量、折价扩大或进口替代动机时,地板应视为破裂。当前 USD/CNY 约 6.77、CTH27 约 81.64¢,68 put 仅约 16.7% OTM。建议用 68/60 put spread、USD/CNH call spread 或 HKEX USD/CNH futures 对冲,并不再把其他 El Niño 软商品 call 当作分散化。
  • Handoff: asset-allocator | FX 断裂触发后的天气凸性 sleeve 再配置 | 若 USD/CNH 触发 7.20-7.30 黄色区或 7.30-7.35 断裂区,应如何在 3-5% NAV 天气凸性 sleeve 内重新分配棉花、咖啡、可可、白糖与现金/美元对冲预算?

既有研究记录

  • Analyst: 首席策略师
  • Stance: synthesize
  • Topic: USD/CNH 上穿 7.20 后,棉花储备地板不再是免费融资腿;上穿 7.30 后,天气 sleeve 的第一任务是保住凸性预算而非追逐软商品 beta。
  • Question: Investigate the topic "USD/CNH 上穿 7.20 后,棉花储备地板不再是免费融资腿;上穿 7.30 后,天气 sleeve 的第一任务是保住凸性预算而非追逐软商品 beta。" and make the strongest evidence-based judgment you can.
  • Summary: 截至 2026-07-27,我建议把 USD/CNH 断裂后的天气凸性 sleeve 从纯软商品篮子改成天气叠加 FX 的期权账本:7.20-7.30 黄色区把总规模封顶在 4.0% NAV,棉花从 23% 降至 15%,现金/美元对冲升至 32%;7.30-7.35 断裂区把总规模压至 3.0% NAV,棉花降至 10%,现金/美元对冲升至 48%。咖啡、可可、白糖只小幅承接,因为 El Nino 与 USD 流动性使软商品尾部相关。
  • Handoff: fx-strategist | USD/CNH 对冲腿的执行成本与触发纪律 | 在 7.20-7.30 与 7.30-7.35 两个触发区间内,USD/CNH call spread、HKEX USD/CNH futures 与美元现金应如何组合,才能以不超过 sleeve 的 32%/48% 预算覆盖棉花储备地板断裂风险?

既有研究记录

  • Analyst: 全球宏观分析师
  • Stance: stress-test
  • Topic: USD/CNH 对冲腿的执行成本与触发纪律
  • Question: 在 7.20-7.30 与 7.30-7.35 两个触发区间内,USD/CNH call spread、HKEX USD/CNH futures 与美元现金应如何组合,才能以不超过 sleeve 的 32%/48% 预算覆盖棉花储备地板断裂风险?
  • Summary: 在 既有研究记录的 32%(黄色区 1.28% NAV)/48%(断裂区 1.44% NAV)现金-美元对冲预算内,正确结构是杠铃:预先建仓的 USD/CNH call spread 作凸性内核、HKEX 期货作触发确认后启用的线性加仓、美元现金/T-bill 作剩余缓冲。关键压力测试发现是:真正约束不是权利金——1M 隐含波动率约 3.8% 处年内低位、凸性便宜——而是 HKEX 期货流动性(Jun-2026 OI 仅约 3,282 张、约 USD 3.28 亿),中等规模基金无法把期货当断裂主对冲。故 call spread 必须在 7.20 之前买好承载凸性,期货只作补充 delta。远期贴水(th_p_d044e1fb 的 −1.5%至−3.5%)使期货持有便宜,但线性、会反转,只该为 delta 加仓供资。现货 6.7717 远低于压力带,阶梯为常备预案。支持并细化 th_p_2da11237(执行价从 6.90/7.20 移到 7.20/7.45)。
  • Handoff: fx-strategist | USD/CNH 隐含波动率反弹对预建 Call Spread 成本结构的压力测试 | 若 1M Implied Volatility 从当前的 3.8% 低位显著回升,USD/CNH call spread 的 Vega 风险是否会突破原定的 32%/48% 预算上限?

Shared Workspace

  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/研究档案 (25896 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (2870 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (10374 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (9482 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (4844 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (11907 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (11054 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/china-cotton-annual-2026.pdf (570 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (5094 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (12151 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (10927 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/usda-cotton-world-trade.pdf (1150982 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (3137 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (12487 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (10907 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (4864 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (10573 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (10351 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (5550 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (10881 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (10207 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (3856 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (11279 bytes)
  • research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (10610 bytes)

Archived Files

Archived files - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/研究档案 (25896 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (2870 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (10374 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (9482 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (4844 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (11907 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (11054 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/china-cotton-annual-2026.pdf (570 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (5094 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (12151 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (10927 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/usda-cotton-world-trade.pdf (1150982 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (3137 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (12487 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (10907 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (4864 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (10573 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (10351 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (5550 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (10881 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (10207 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/meta.json (3856 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.en.md (11279 bytes) - research discussion/dc36a116-efa9-4bd4-bf8b-40ab9e39504d/research note/report.zh.md (10610 bytes)