众所周知,在A股市场中存在日历效应,最有名的应该就属“周四效应”,即在周四这天A股市场下跌的期望要大于其他时候。那么在可转债市场中,是否也能得出类似的结论呢?今天我们来用数据进行一个解析。
(数据包括了2018-01-01至2022-08-10所有的可转债日线数据)
首先我们先看看最简单的分类,把所有数据按照周一到周五进行分组统计
星期 | 胜率 | 平均涨幅 | 涨幅中位数
--------+--------+------------+--------------
周一 | 52.48% | 0.23% | 0.04%
周二 | 48.34% | 0.06% | -0.01%
周三 | 50.53% | 0.16% | 0.01%
周四 | 49.65% | 0.1% | 0.0%
周五 | 49.98% | 0.05% | 0.0%
有朋友看完可能觉得周一胜率高,周二胜率低,但是这个结论很奇怪且与A股中的周内效应也不一致,于是我又将数据按照年份进行拆分,进行统计如下:
年份 | 星期 | 胜率 | 平均涨幅 | 涨幅中位数
--------+--------+--------+------------+--------------
|2018 | 周一 | 48.16% | 0.04% | -0.01%
|2018| 周二 | 49.03% | 0.02% | 0.0%
|2018 | 周三 | 43.77% | -0.08% | -0.05%
|2018 | 周四 | 44.44% | -0.05% | -0.04%
|2018 | 周五 | 48.51% | 0.03% | 0.0%
|2019 | 周一 | 50.77% | 0.17% | 0.01%
|2019 | 周二 | 54.71% | 0.15% | 0.05%
|2019 | 周三 | 48.89% | 0.09% | 0.0%
|2019 | 周四 | 48.95% | 0.02% | 0.0%
|2019 | 周五 | 55.76% | 0.16% | 0.06%
|2020 | 周一 | 51.32% | 0.31% | 0.03%
|2020 | 周二 | 52.91% | 0.35% | 0.05%
|2020 | 周三 | 46.49% | 0.1% | -0.04%
|2020 | 周四 | 47.16% | 0.04% | -0.04%
|2020 | 周五 | 44.68% | -0.09% | -0.09%
|2021 | 周一 | 56.19% | 0.37% | 0.12%
|2021 | 周二 | 46.36% | 0.0% | -0.05%
|2021 | 周三 | 51.16% | 0.14% | 0.02%
|2021 | 周四 | 52.1% | 0.17% | 0.03%
|2021 | 周五 | 51.21% | 0.11% | 0.02%
|2022 | 周一 | 50.84% | 0.04% | 0.01%
|2022 | 周二 | 41.55% | -0.25% | -0.2%
|2022 | 周三 | 57.22% | 0.38% | 0.15%
|2022 | 周四 | 50.67% | 0.14% | 0.01%
|2022 | 周五 | 50.65% | 0.06% | 0.01%
到这里我们发现,刚刚的结论已不复存在。在不同的年份,可转债星期N的表现都不相同,没有规律。
文章本来到这里就该结束了,结论就是可转债没有明显的周内效应。
但就这么结束的话,各位读者老爷势必不答应。遂继续深挖。这次我把正股的涨跌幅一并进行统计:
年份 |星期 |转债胜率|平均涨幅|中位数 |正股胜率|平均涨幅|中位数
-------+-----+---------+----------+--------+---------+----------+---------
|2018 | 周一 | 48.16% | 0.04% | -0.01% | 50.27% | 0.22% | 0.06%
|2018 | 周二 | 49.03% | 0.02% | 0.0% | 52.98% | 0.08% | 0.19%
|2018 | 周三 | 43.77% | -0.08% | -0.05% | 38.02% | -0.38% | -0.38%
|2018 | 周四 | 44.44% | -0.05% | -0.04% | 42.65% | -0.43% | -0.2%
|2018 | 周五 | 48.51% | 0.03% | 0.0% | 45.81% | -0.15% | -0.1%
|2019 | 周一 | 50.77% | 0.17% | 0.01% | 53.75% | 0.28% | 0.26%
|2019 | 周二 | 54.71% | 0.15% | 0.05% | 52.02% | 0.26% | 0.16%
|2019 | 周三 | 48.89% | 0.09% | 0.0% | 43.53% | 0.01% | -0.16%
|2019 | 周四 | 48.95% | 0.02% | 0.0% | 45.17% | -0.14% | -0.1%
|2019 | 周五 | 55.76% | 0.16% | 0.06% | 48.81% | 0.06% | 0.0%
|2020 | 周一 | 51.32% | 0.31% | 0.03% | 53.55% | 0.37% | 0.25%
|2020 | 周二 | 52.91% | 0.35% | 0.05% | 52.51% | 0.33% | 0.17%
|2020 | 周三 | 46.49% | 0.1% | -0.04% | 43.13% | -0.12% | -0.27%
|2020 | 周四 | 47.16% | 0.04% | -0.04% | 46.6% | -0.06% | -0.12%
|2020 | 周五 | 44.68% | -0.09% | -0.09% | 44.32% | -0.17% | -0.19%
|2021 | 周一 | 56.19% | 0.37% | 0.12% | 52.77% | 0.37% | 0.19%
|2021 | 周二 | 46.36% | 0.0% | -0.05% | 46.09% | -0.05% | -0.1%
|2021 | 周三 | 51.16% | 0.14% | 0.02% | 48.61% | 0.14% | 0.0%
|2021 | 周四 | 52.1% | 0.17% | 0.03% | 46.09% | 0.05% | -0.11%
|2021 | 周五 | 51.21% | 0.11% | 0.02% | 47.5% | 0.09% | 0.0%
|2022 | 周一 | 50.84% | 0.04% | 0.01% | 51.41% | 0.17% | 0.13%
|2022 | 周二 | 41.55% | -0.25% | -0.2% | 42.68% | -0.54% | -0.42%
|2022 | 周三 | 57.22% | 0.38% | 0.15% | 52.51% | 0.37% | 0.17%
|2022 | 周四 | 50.67% | 0.14% | 0.01% | 46.96% | -0.06% | -0.1%
|2022 | 周五 | 50.65% | 0.06% | 0.01% | 47.19% | 0.03% | -0.05%
完成后数据的内容已经有些多了,看起来不太容易了,但没关系,我们距离真正的结论已经很近了。
从这份表格中,我们能看出:
这些相对有规律结论仅仅是针对正股的。有聪明的朋友肯定已经想到利用轮动的办法来获取超额,接下来我将展示最终的表格及结论:
年份 |星期 |转债胜率|平均涨幅|中位数 |正股胜率|平均涨幅|中位数 |超额胜率|平均涨幅|中位数
-------+-----+---------+----------+--------+---------+----------+--------+---------+----------+---------
|2018 | 周一 | 48.16% | 0.04% | -0.01% | 50.27% | 0.22% | 0.06% | 46.62% | -0.17% | -0.12%
|2018 | 周二 | 49.03% | 0.02% | 0.0% | 52.98% | 0.08% | 0.19% | 45.72% | -0.06% | -0.14%
|2018 | 周三 | 43.77% | -0.08% | -0.05% | 38.02% | -0.38% | -0.38% | 59.61% | 0.3% | 0.32%
|2018 | 周四 | 44.44% | -0.05% | -0.04% | 42.65% | -0.43% | -0.2% | 55.91% | 0.38% | 0.21%
|2018 | 周五 | 48.51% | 0.03% | 0.0% | 45.81% | -0.15% | -0.1% | 52.81% | 0.18% | 0.1%
|2019 | 周一 | 50.77% | 0.17% | 0.01% | 53.75% | 0.28% | 0.26% | 46.05% | -0.11% | -0.13%
|2019 | 周二 | 54.71% | 0.15% | 0.05% | 52.02% | 0.26% | 0.16% | 48.63% | -0.11% | -0.04%
|2019 | 周三 | 48.89% | 0.09% | 0.0% | 43.53% | 0.01% | -0.16% | 54.72% | 0.08% | 0.13%
|2019 | 周四 | 48.95% | 0.02% | 0.0% | 45.17% | -0.14% | -0.1% | 53.55% | 0.15% | 0.11%
|2019 | 周五 | 55.76% | 0.16% | 0.06% | 48.81% | 0.06% | 0.0% | 52.57% | 0.1% | 0.08%
|2020 | 周一 | 51.32% | 0.31% | 0.03% | 53.55% | 0.37% | 0.25% | 45.83% | -0.06% | -0.15%
|2020 | 周二 | 52.91% | 0.35% | 0.05% | 52.51% | 0.33% | 0.17% | 47.7% | 0.02% | -0.08%
|2020 | 周三 | 46.49% | 0.1% | -0.04% | 43.13% | -0.12% | -0.27% | 55.09% | 0.22% | 0.17%
|2020 | 周四 | 47.16% | 0.04% | -0.04% | 46.6% | -0.06% | -0.12% | 51.83% | 0.1% | 0.06%
|2020 | 周五 | 44.68% | -0.09% | -0.09% | 44.32% | -0.17% | -0.19% | 51.27% | 0.08% | 0.04%
|2021 | 周一 | 56.19% | 0.37% | 0.12% | 52.77% | 0.37% | 0.19% | 49.48% | -0.0% | -0.01%
|2021 | 周二 | 46.36% | 0.0% | -0.05% | 46.09% | -0.05% | -0.1% | 51.61% | 0.05% | 0.05%
|2021 | 周三 | 51.16% | 0.14% | 0.02% | 48.61% | 0.14% | 0.0% | 50.78% | 0.0% | 0.02%
|2021 | 周四 | 52.1% | 0.17% | 0.03% | 46.09% | 0.05% | -0.11% | 55.25% | 0.13% | 0.16%
|2021 | 周五 | 51.21% | 0.11% | 0.02% | 47.5% | 0.09% | 0.0% | 52.23% | 0.02% | 0.08%
|2022 | 周一 | 50.84% | 0.04% | 0.01% | 51.41% | 0.17% | 0.13% | 46.71% | -0.13% | -0.14%
|2022 | 周二 | 41.55% | -0.25% | -0.2% | 42.68% | -0.54% | -0.42% | 53.0% | 0.29% | 0.13%
|2022 | 周三 | 57.22% | 0.38% | 0.15% | 52.51% | 0.37% | 0.17% | 51.11% | 0.01% | 0.04%
|2022 | 周四 | 50.67% | 0.14% | 0.01% | 46.96% | -0.06% | -0.1% | 54.31% | 0.21% | 0.18%
|2022 | 周五 | 50.65% | 0.06% | 0.01% | 47.19% | 0.03% | -0.05% | 51.42% | 0.03% | 0.04%
这一次加入了超额涨幅的统计,即转债的涨幅-正股涨幅,举例转债A当天的涨幅为9%,正股A当天涨幅为6%,那么其当天的超额涨幅为9% - 6% = 3%。在超额涨幅这一栏中,情况已经变得非常简单了:
这个规律在18年到今年8月10日,按年为单位来统计,都是有效的。
另有一点要注意,如果进行实践,应该再加上牛熊市的判断,单纯的每周轮动,收益无法覆盖换仓成本。
从统计上看,有明显规律,至于内在的逻辑原因,大家可以集思广益,一起讨论。
最后放上一张总结整体超额,按星期统计的表格作为结尾:
星期 |转债胜率|平均涨幅|中位数 |正股胜率|平均涨幅|中位数 |超额胜率|平均涨幅|中位数
------+---------+----------+--------+---------+----------+--------+---------+----------+---------
周一 | 52.48% | 0.23% | 0.04% | 52.66% | 0.3% | 0.19% | 47.27% | -0.07% | -0.1%
周二 | 48.34% | 0.06% | -0.01% | 48.27% | -0.01% | 0.0% | 50.11% | 0.06% | 0.0%
周三 | 50.53% | 0.16% | 0.01% | 46.69% | 0.07% | -0.05% | 53.07% | 0.09% | 0.1%
周四 | 49.65% | 0.1% | 0.0% | 46.04% | -0.06% | -0.12% | 54.03% | 0.16% | 0.13%
周五 | 49.98% | 0.05% | 0.0% | 46.74% | -0.0% | -0.07% | 51.91% | 0.06% | 0.07%
(数据包括了2018-01-01至2022-08-10所有的可转债日线数据)
首先我们先看看最简单的分类,把所有数据按照周一到周五进行分组统计
星期 | 胜率 | 平均涨幅 | 涨幅中位数
--------+--------+------------+--------------
周一 | 52.48% | 0.23% | 0.04%
周二 | 48.34% | 0.06% | -0.01%
周三 | 50.53% | 0.16% | 0.01%
周四 | 49.65% | 0.1% | 0.0%
周五 | 49.98% | 0.05% | 0.0%
有朋友看完可能觉得周一胜率高,周二胜率低,但是这个结论很奇怪且与A股中的周内效应也不一致,于是我又将数据按照年份进行拆分,进行统计如下:
年份 | 星期 | 胜率 | 平均涨幅 | 涨幅中位数
--------+--------+--------+------------+--------------
|2018 | 周一 | 48.16% | 0.04% | -0.01%
|2018| 周二 | 49.03% | 0.02% | 0.0%
|2018 | 周三 | 43.77% | -0.08% | -0.05%
|2018 | 周四 | 44.44% | -0.05% | -0.04%
|2018 | 周五 | 48.51% | 0.03% | 0.0%
|2019 | 周一 | 50.77% | 0.17% | 0.01%
|2019 | 周二 | 54.71% | 0.15% | 0.05%
|2019 | 周三 | 48.89% | 0.09% | 0.0%
|2019 | 周四 | 48.95% | 0.02% | 0.0%
|2019 | 周五 | 55.76% | 0.16% | 0.06%
|2020 | 周一 | 51.32% | 0.31% | 0.03%
|2020 | 周二 | 52.91% | 0.35% | 0.05%
|2020 | 周三 | 46.49% | 0.1% | -0.04%
|2020 | 周四 | 47.16% | 0.04% | -0.04%
|2020 | 周五 | 44.68% | -0.09% | -0.09%
|2021 | 周一 | 56.19% | 0.37% | 0.12%
|2021 | 周二 | 46.36% | 0.0% | -0.05%
|2021 | 周三 | 51.16% | 0.14% | 0.02%
|2021 | 周四 | 52.1% | 0.17% | 0.03%
|2021 | 周五 | 51.21% | 0.11% | 0.02%
|2022 | 周一 | 50.84% | 0.04% | 0.01%
|2022 | 周二 | 41.55% | -0.25% | -0.2%
|2022 | 周三 | 57.22% | 0.38% | 0.15%
|2022 | 周四 | 50.67% | 0.14% | 0.01%
|2022 | 周五 | 50.65% | 0.06% | 0.01%
到这里我们发现,刚刚的结论已不复存在。在不同的年份,可转债星期N的表现都不相同,没有规律。
文章本来到这里就该结束了,结论就是可转债没有明显的周内效应。
但就这么结束的话,各位读者老爷势必不答应。遂继续深挖。这次我把正股的涨跌幅一并进行统计:
年份 |星期 |转债胜率|平均涨幅|中位数 |正股胜率|平均涨幅|中位数
-------+-----+---------+----------+--------+---------+----------+---------
|2018 | 周一 | 48.16% | 0.04% | -0.01% | 50.27% | 0.22% | 0.06%
|2018 | 周二 | 49.03% | 0.02% | 0.0% | 52.98% | 0.08% | 0.19%
|2018 | 周三 | 43.77% | -0.08% | -0.05% | 38.02% | -0.38% | -0.38%
|2018 | 周四 | 44.44% | -0.05% | -0.04% | 42.65% | -0.43% | -0.2%
|2018 | 周五 | 48.51% | 0.03% | 0.0% | 45.81% | -0.15% | -0.1%
|2019 | 周一 | 50.77% | 0.17% | 0.01% | 53.75% | 0.28% | 0.26%
|2019 | 周二 | 54.71% | 0.15% | 0.05% | 52.02% | 0.26% | 0.16%
|2019 | 周三 | 48.89% | 0.09% | 0.0% | 43.53% | 0.01% | -0.16%
|2019 | 周四 | 48.95% | 0.02% | 0.0% | 45.17% | -0.14% | -0.1%
|2019 | 周五 | 55.76% | 0.16% | 0.06% | 48.81% | 0.06% | 0.0%
|2020 | 周一 | 51.32% | 0.31% | 0.03% | 53.55% | 0.37% | 0.25%
|2020 | 周二 | 52.91% | 0.35% | 0.05% | 52.51% | 0.33% | 0.17%
|2020 | 周三 | 46.49% | 0.1% | -0.04% | 43.13% | -0.12% | -0.27%
|2020 | 周四 | 47.16% | 0.04% | -0.04% | 46.6% | -0.06% | -0.12%
|2020 | 周五 | 44.68% | -0.09% | -0.09% | 44.32% | -0.17% | -0.19%
|2021 | 周一 | 56.19% | 0.37% | 0.12% | 52.77% | 0.37% | 0.19%
|2021 | 周二 | 46.36% | 0.0% | -0.05% | 46.09% | -0.05% | -0.1%
|2021 | 周三 | 51.16% | 0.14% | 0.02% | 48.61% | 0.14% | 0.0%
|2021 | 周四 | 52.1% | 0.17% | 0.03% | 46.09% | 0.05% | -0.11%
|2021 | 周五 | 51.21% | 0.11% | 0.02% | 47.5% | 0.09% | 0.0%
|2022 | 周一 | 50.84% | 0.04% | 0.01% | 51.41% | 0.17% | 0.13%
|2022 | 周二 | 41.55% | -0.25% | -0.2% | 42.68% | -0.54% | -0.42%
|2022 | 周三 | 57.22% | 0.38% | 0.15% | 52.51% | 0.37% | 0.17%
|2022 | 周四 | 50.67% | 0.14% | 0.01% | 46.96% | -0.06% | -0.1%
|2022 | 周五 | 50.65% | 0.06% | 0.01% | 47.19% | 0.03% | -0.05%
完成后数据的内容已经有些多了,看起来不太容易了,但没关系,我们距离真正的结论已经很近了。
从这份表格中,我们能看出:
- 可转债对应的正股走势符合传统A股市场的周四效应,且在周四和周五的表现都不好,胜率历年都小于50%;
- 相反的,正股在周一的表现相当好,胜率一直>50%,且平均涨幅均大于0。
这些相对有规律结论仅仅是针对正股的。有聪明的朋友肯定已经想到利用轮动的办法来获取超额,接下来我将展示最终的表格及结论:
年份 |星期 |转债胜率|平均涨幅|中位数 |正股胜率|平均涨幅|中位数 |超额胜率|平均涨幅|中位数
-------+-----+---------+----------+--------+---------+----------+--------+---------+----------+---------
|2018 | 周一 | 48.16% | 0.04% | -0.01% | 50.27% | 0.22% | 0.06% | 46.62% | -0.17% | -0.12%
|2018 | 周二 | 49.03% | 0.02% | 0.0% | 52.98% | 0.08% | 0.19% | 45.72% | -0.06% | -0.14%
|2018 | 周三 | 43.77% | -0.08% | -0.05% | 38.02% | -0.38% | -0.38% | 59.61% | 0.3% | 0.32%
|2018 | 周四 | 44.44% | -0.05% | -0.04% | 42.65% | -0.43% | -0.2% | 55.91% | 0.38% | 0.21%
|2018 | 周五 | 48.51% | 0.03% | 0.0% | 45.81% | -0.15% | -0.1% | 52.81% | 0.18% | 0.1%
|2019 | 周一 | 50.77% | 0.17% | 0.01% | 53.75% | 0.28% | 0.26% | 46.05% | -0.11% | -0.13%
|2019 | 周二 | 54.71% | 0.15% | 0.05% | 52.02% | 0.26% | 0.16% | 48.63% | -0.11% | -0.04%
|2019 | 周三 | 48.89% | 0.09% | 0.0% | 43.53% | 0.01% | -0.16% | 54.72% | 0.08% | 0.13%
|2019 | 周四 | 48.95% | 0.02% | 0.0% | 45.17% | -0.14% | -0.1% | 53.55% | 0.15% | 0.11%
|2019 | 周五 | 55.76% | 0.16% | 0.06% | 48.81% | 0.06% | 0.0% | 52.57% | 0.1% | 0.08%
|2020 | 周一 | 51.32% | 0.31% | 0.03% | 53.55% | 0.37% | 0.25% | 45.83% | -0.06% | -0.15%
|2020 | 周二 | 52.91% | 0.35% | 0.05% | 52.51% | 0.33% | 0.17% | 47.7% | 0.02% | -0.08%
|2020 | 周三 | 46.49% | 0.1% | -0.04% | 43.13% | -0.12% | -0.27% | 55.09% | 0.22% | 0.17%
|2020 | 周四 | 47.16% | 0.04% | -0.04% | 46.6% | -0.06% | -0.12% | 51.83% | 0.1% | 0.06%
|2020 | 周五 | 44.68% | -0.09% | -0.09% | 44.32% | -0.17% | -0.19% | 51.27% | 0.08% | 0.04%
|2021 | 周一 | 56.19% | 0.37% | 0.12% | 52.77% | 0.37% | 0.19% | 49.48% | -0.0% | -0.01%
|2021 | 周二 | 46.36% | 0.0% | -0.05% | 46.09% | -0.05% | -0.1% | 51.61% | 0.05% | 0.05%
|2021 | 周三 | 51.16% | 0.14% | 0.02% | 48.61% | 0.14% | 0.0% | 50.78% | 0.0% | 0.02%
|2021 | 周四 | 52.1% | 0.17% | 0.03% | 46.09% | 0.05% | -0.11% | 55.25% | 0.13% | 0.16%
|2021 | 周五 | 51.21% | 0.11% | 0.02% | 47.5% | 0.09% | 0.0% | 52.23% | 0.02% | 0.08%
|2022 | 周一 | 50.84% | 0.04% | 0.01% | 51.41% | 0.17% | 0.13% | 46.71% | -0.13% | -0.14%
|2022 | 周二 | 41.55% | -0.25% | -0.2% | 42.68% | -0.54% | -0.42% | 53.0% | 0.29% | 0.13%
|2022 | 周三 | 57.22% | 0.38% | 0.15% | 52.51% | 0.37% | 0.17% | 51.11% | 0.01% | 0.04%
|2022 | 周四 | 50.67% | 0.14% | 0.01% | 46.96% | -0.06% | -0.1% | 54.31% | 0.21% | 0.18%
|2022 | 周五 | 50.65% | 0.06% | 0.01% | 47.19% | 0.03% | -0.05% | 51.42% | 0.03% | 0.04%
这一次加入了超额涨幅的统计,即转债的涨幅-正股涨幅,举例转债A当天的涨幅为9%,正股A当天涨幅为6%,那么其当天的超额涨幅为9% - 6% = 3%。在超额涨幅这一栏中,情况已经变得非常简单了:
- 周一转债跑输正股,不管从胜率还是平均涨幅都跑输;
- 周三到周五转债跑赢正股,不管从胜率还是平均涨幅都跑赢。
这个规律在18年到今年8月10日,按年为单位来统计,都是有效的。
另有一点要注意,如果进行实践,应该再加上牛熊市的判断,单纯的每周轮动,收益无法覆盖换仓成本。
从统计上看,有明显规律,至于内在的逻辑原因,大家可以集思广益,一起讨论。
最后放上一张总结整体超额,按星期统计的表格作为结尾:
星期 |转债胜率|平均涨幅|中位数 |正股胜率|平均涨幅|中位数 |超额胜率|平均涨幅|中位数
------+---------+----------+--------+---------+----------+--------+---------+----------+---------
周一 | 52.48% | 0.23% | 0.04% | 52.66% | 0.3% | 0.19% | 47.27% | -0.07% | -0.1%
周二 | 48.34% | 0.06% | -0.01% | 48.27% | -0.01% | 0.0% | 50.11% | 0.06% | 0.0%
周三 | 50.53% | 0.16% | 0.01% | 46.69% | 0.07% | -0.05% | 53.07% | 0.09% | 0.1%
周四 | 49.65% | 0.1% | 0.0% | 46.04% | -0.06% | -0.12% | 54.03% | 0.16% | 0.13%
周五 | 49.98% | 0.05% | 0.0% | 46.74% | -0.0% | -0.07% | 51.91% | 0.06% | 0.07%