I'm sorry, but since it's a concept edition, there is no implementation this time.
If you can predict the tendency before the daily price limit is reached, you may be able to make money with less risk by riding on it, so the work is in progress. You should be able to find the daily price limit with the same logic
The reason why I thought that stocks with a daily price limit could be expected is
--There are a certain number of people who can't resist the temptation of money. --Some people fly insider-like --Data analysis and detection of flying get --Undisclosed information You can smoke sweet honey without getting it
Formally, it's legally shared with gray places, but I haven't got any undisclosed information, so let's believe it's okay, lol
Click here for a blog that summarizes the contents around artificial intelligence: Effort with artificial intelligence 1mm
Let's look at the transition by standardizing the return index of a certain stock with the return index of the Nikkei average as usual. However, I'm worried that I may have to do something quite technical.
In the article [Aiming for a decline after shareholder benefits](inserting a link), the value returned by the prepared function was ** the expected value and variance of the past rate of return **. So, once you analyze it, all you have to do is buy a brand that you are comfortable with near the vesting date. ..
However, this time, the value returned by the prepared function is ** whether the price will reach the daily price limit the next day **. This is a story that I know cannot be dealt with by simple statistics or machine learning. (When I did a similar analysis on the decision tree before, it wasn't very accurate.)
In other words, ** I don't know what to do at the moment ** concludes, lol For the time being, I'll just keep a record of my wisdom.
Look at the cycle of the deviation rate from the Nikkei average. If the stock fits my ideal, when the deviation rate from the Nikkei average gradually rises and meets a certain level (whether it is X-day or the deviation rate is Y% or more), the next There is a high probability that the price will stop on the day of the event. If it's something like that, it fits the image and looks good.
I wonder if it will be a good result to look at the deviation rate from the moving average before the stop height (low).
For the time being, [Implementing] Data analysis and profitable strategy aiming for a decline after shareholder benefits-Concept- Is over, I think I'll try it from here.
Recommended Posts