Probabilistic stock market forecasts. Free, everyday. |
|
Feeds
Want to learn how to read these charts? Check our BLOG
Investolution Probabilistic Forecast Charts are generated daily to predict major stock market indices. Each stock market prediction generates a stochastic distribution of the maximum and minimum percent price change expected in the next 40 trading days.
Join our growing community that uses them free of charge for Option and Stock trading
Click here to send us a message
Blog
The title of today's blog is the most commonly asked question to us. We'll therefore leave the theory and technical details to a future blog and jump right into a real life example. Please keep in mind that this is only one of the many ways we believe our technology might be useful for a trader. Also note that what we present here should not be regarded as financial advise. Use your common sense while trading and take your time to read the disclaimer at the bottom of this page.
If we recap, our prediction models estimate the likelihood of a stock price to reach and exceed certain points (minimum and maximum) within a time period. The chart below is the prediction for Nasdaq on February 12 2008.
The prediction also comes with a plain text description, which reads:
It is predicted that Nasdaq Composite will not close under 2,185 and over 2,392 between the dates February 12, 2008 and April 8, 2008. This prediction method was accurate for 80.0% and 86.0% of the cases (for minimum and maximum predictions, respectively) within an error margin of +-5% in the past.
There is also a more detailed report (example) which shows how each method performed historically and what was their prediction on this specific day. For the sake of keeping it simple, we'll ignore everything for now except:
- The prediction was done on 12 February 2008 and it covers the time period from 12 February to 8 April 2008.
- The maximum predicted change was 3.1% by the rule MEDIAN, which performed best in the past.
- The maximum prediction was accurate 99% of the cases for an error range of +-10%
- The maximum prediction was accurate 86% of the cases for an error range of +-5%
Armed with this information we can now try to see if there are any trading opportunities. One of the most commonly traded stock options is QQQQ Nasdaq Power Shares. At the time of writing, coincidentally, 8 April 2008 was one of the expiration days and the QQQQ price was 43.82.
The model predicted the maximum close to be 3.1% above today's price until 8 April, we therefore expect that the price should not close over 45.17 (43.82 x 1.031) if the prediction was 100%.
We know that our model was not 100% accurate in the past so we have no reason to believe that it will be in the future. But it had a somewhat predictable probability distribution of relative error ranges. This means, if we add these error ranges on top of our predicted maximum change percentage, we can then calculate the probability of a daily close to be above a certain price. Using the 10% error range:
- We can estimate with 99% accuracy that the price will NOT close above 13.1% (3.1 + 10) higher than the current price 43.82, which is 43.82 x 1.131 = 49.56.
- This price is just below 50, therefore we can assume a call option at the price $50 has a chance of less than 1% to get exercised (http://finance.yahoo.com/q/os?s=QQQQ&m=2008-04-18)
- With the same token, we can estimate with 86% accuracy that the price will NOT close above 8.1% (3.1% + 5%) higher than the current price 43.82, which is 43.82 x 1.081 = 47.37.
- This price is below 48, therefore we can assume a call option at the price $48 has a chance of less than 14% to get exercised (http://finance.yahoo.com/q/os?s=QQQQ&m=2008-04-18)
Following the same procedure above, a trader can calculate risks associated with each option (both calls and puts) and determine the optimal strategy for her/his risk level. Needless to say, the prediction does not have to expire exactly on the last day of prediction period. If the option is expiring in 30 trading days and our prediction forecasts it to close under a certain value in 40 trading days, the price will still be under predicted value for the first 30 days. It is also smart to observe how the predicted prices and the option prices are changing for a while before drawing conclusions and creating strategies. Some predictions such as Dow Jones and TSX are more precise than Nasdaq and Nikkei 225.
Automatically calculating the likelihood of a stock price to go beyond option strike prices is one of our active projects, results of which might show up in our daily reports soon. If you find other ways to use the predictions, or just want to share your opinions/ideas please let us know. We are always looking for new ways to improve our technology and share it with our community online.