{
  "id": 544733,
  "title": "Correlations between symbols",
  "url": "/competitions/jane-street-real-time-market-data-forecasting/discussion/544733",
  "author_name": "",
  "post_date": "2024-11-06T18:23:38.334650600Z",
  "votes": 7,
  "comment_count": 6,
  "views": 0,
  "content": "<p>I've ran a simple correlation between the symbols through time (for responder_6) and it's not as high as I expected. There are some groups with high-ish correlation but most symbols are moving independently. Check for yourselves:<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F22052344%2Fb597d5a9e37c7db3251f56c825a20d96%2Fcorr.png?generation=1730917410127810&amp;alt=media\" alt=\"\"></p>",
  "messages": [
    {
      "id": "3038235",
      "postDate": "11/06/2024 18:23:38",
      "content": "<p>I've ran a simple correlation between the symbols through time (for responder_6) and it's not as high as I expected. There are some groups with high-ish correlation but most symbols are moving independently. Check for yourselves:<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F22052344%2Fb597d5a9e37c7db3251f56c825a20d96%2Fcorr.png?generation=1730917410127810&amp;alt=media\" alt=\"\"></p>",
      "rawMarkdown": "I've ran a simple correlation between the symbols through time (for responder_6) and it's not as high as I expected. There are some groups with high-ish correlation but most symbols are moving independently. Check for yourselves:\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F22052344%2Fb597d5a9e37c7db3251f56c825a20d96%2Fcorr.png?generation=1730917410127810&alt=media)",
      "votes": null
    },
    {
      "id": "3038249",
      "postDate": "11/06/2024 18:53:52",
      "content": "<p>what timeframe has been used in the analysis?</p>",
      "rawMarkdown": "what timeframe has been used in the analysis?",
      "votes": null
    },
    {
      "id": "3038329",
      "postDate": "11/06/2024 20:37:18",
      "content": "<p>Last 3 partitions</p>",
      "rawMarkdown": "Last 3 partitions",
      "votes": null
    },
    {
      "id": "3038364",
      "postDate": "11/06/2024 21:56:30",
      "content": "<p><a href=\"https://www.kaggle.com/natanlabarrere\" target=\"_blank\">@natanlabarrere</a> <br>\nPlease use <strong>hexbin style plot</strong> to make it more readable, you can peruse this below-<br>\n<a href=\"https://seaborn.pydata.org/examples/hexbin_marginals.html\" target=\"_blank\">https://seaborn.pydata.org/examples/hexbin_marginals.html</a></p>",
      "rawMarkdown": "natanlabarrere \nPlease use **hexbin style plot** to make it more readable, you can peruse this below-\nhttps://seaborn.pydata.org/examples/hexbin_marginals.html",
      "votes": null
    },
    {
      "id": "3038489",
      "postDate": "11/07/2024 03:21:24",
      "content": "<p>Financial theory says that a well-diversified portfolio can reduce risk if assets are not perfectly correlated and respond independently to different economic events. <a href=\"https://en.wikipedia.org/wiki/Asset_allocation\" target=\"_blank\">https://en.wikipedia.org/wiki/Asset_allocation</a></p>",
      "rawMarkdown": "Financial theory says that a well-diversified portfolio can reduce risk if assets are not perfectly correlated and respond independently to different economic events. [https://en.wikipedia.org/wiki/Asset_allocation](https://en.wikipedia.org/wiki/Asset_allocation)",
      "votes": null
    },
    {
      "id": "3038748",
      "postDate": "11/07/2024 10:03:18",
      "content": "<p>curious about the normalization (/implementation) of the correlation. I wouldn't call 67% high-ish but high :)</p>",
      "rawMarkdown": "curious about the normalization (/implementation) of the correlation. I wouldn't call 67% high-ish but high :)",
      "votes": null
    },
    {
      "id": "3038783",
      "postDate": "11/07/2024 10:47:30",
      "content": "<p>Nothing sophisticated, I just made x = unique date/time id and y = responder_6 for each symbol_id, and ran Pearson's corr. You're right about 67% but there are very few like that</p>",
      "rawMarkdown": "Nothing sophisticated, I just made x = unique date/time id and y = responder_6 for each symbol_id, and ran Pearson's corr. You're right about 67% but there are very few like that",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3038249,
      "author_name": "casras",
      "author_url": "",
      "post_date": "11/06/2024 18:53:52",
      "content": "<p>what timeframe has been used in the analysis?</p>",
      "votes": null,
      "replies": [
        {
          "id": 3038329,
          "author_name": "natanlabarrere",
          "author_url": "",
          "post_date": "11/06/2024 20:37:18",
          "content": "<p>Last 3 partitions</p>",
          "votes": null,
          "replies": [
            {
              "id": 3038364,
              "author_name": "ravi20076",
              "author_url": "",
              "post_date": "11/06/2024 21:56:30",
              "content": "<p><a href=\"https://www.kaggle.com/natanlabarrere\" target=\"_blank\">@natanlabarrere</a> <br>\nPlease use <strong>hexbin style plot</strong> to make it more readable, you can peruse this below-<br>\n<a href=\"https://seaborn.pydata.org/examples/hexbin_marginals.html\" target=\"_blank\">https://seaborn.pydata.org/examples/hexbin_marginals.html</a></p>",
              "votes": null,
              "replies": []
            }
          ]
        }
      ]
    },
    {
      "id": 3038489,
      "author_name": "johnfariasgiraldo",
      "author_url": "",
      "post_date": "11/07/2024 03:21:24",
      "content": "<p>Financial theory says that a well-diversified portfolio can reduce risk if assets are not perfectly correlated and respond independently to different economic events. <a href=\"https://en.wikipedia.org/wiki/Asset_allocation\" target=\"_blank\">https://en.wikipedia.org/wiki/Asset_allocation</a></p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3038748,
      "author_name": "brtvandenbroeck",
      "author_url": "",
      "post_date": "11/07/2024 10:03:18",
      "content": "<p>curious about the normalization (/implementation) of the correlation. I wouldn't call 67% high-ish but high :)</p>",
      "votes": null,
      "replies": [
        {
          "id": 3038783,
          "author_name": "natanlabarrere",
          "author_url": "",
          "post_date": "11/07/2024 10:47:30",
          "content": "<p>Nothing sophisticated, I just made x = unique date/time id and y = responder_6 for each symbol_id, and ran Pearson's corr. You're right about 67% but there are very few like that</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3038235": "I've ran a simple correlation between the symbols through time (for responder_6) and it's not as high as I expected. There are some groups with high-ish correlation but most symbols are moving independently. Check for yourselves:\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F22052344%2Fb597d5a9e37c7db3251f56c825a20d96%2Fcorr.png?generation=1730917410127810&alt=media)",
    "3038249": "what timeframe has been used in the analysis?",
    "3038329": "Last 3 partitions",
    "3038364": "natanlabarrere \nPlease use **hexbin style plot** to make it more readable, you can peruse this below-\nhttps://seaborn.pydata.org/examples/hexbin_marginals.html",
    "3038489": "Financial theory says that a well-diversified portfolio can reduce risk if assets are not perfectly correlated and respond independently to different economic events. [https://en.wikipedia.org/wiki/Asset_allocation](https://en.wikipedia.org/wiki/Asset_allocation)",
    "3038748": "curious about the normalization (/implementation) of the correlation. I wouldn't call 67% high-ish but high :)",
    "3038783": "Nothing sophisticated, I just made x = unique date/time id and y = responder_6 for each symbol_id, and ran Pearson's corr. You're right about 67% but there are very few like that"
  },
  "source": "meta"
}