Implementation of logistic regression into technical analysis
| Authors | |
|---|---|
| Year of publication | 2013 |
| Type | Article in Proceedings |
| Conference | Proceedings of the 31st International Conference Mathematical Methods in Economics 2013 |
| MU Faculty or unit | |
| Citation | |
| Field | Economy |
| Keywords | logistic regression; technical analysis; moving averages; automated trading |
| Description | Most of the investment strategies based on technical analysis are based on the principle that your strategy should be as easy as it is possible; in order to simplify your decision making. Goal of this paper is to use more sophisticated methods to combine the signals from indicators of technical analysis to create advanced form of investing strategy, which will be sustainable in long run. This can be ensured by self-correcting mechanisms build-in the strategy itself. Econometrical methods will be used to determine whether some kind of indicator has it relevance on the chosen type of asset. All input variables will be time series of dummy variables showing whether the indicator is suggesting taking a long position or not and of course their lags. Explained variable will be the successful trade (the price movement upwards is greater then spread and commissions). For this kind of purposes logistic regression seems to be essential, which is widely used in credit scoring. Basically the problem whether to invest or not is the same issue as whether to give a customer a loan or not. The only difference will be in the type of data. Credit scoring use mostly panel data, however it will be handled solely with time series in this paper. |
| Related projects: |