ATMA Discussion Forums
ATMA Discussion Forums for Affiliates and Members
Accepted AnswerNikhil DograOfflineOne can fiddle around time-frames to see what suits one's personality the best for trading & also recognizing the instruments personality. For Indexes using less than 1H or more than 1H is ideal. Personally for me in majority instances 1H for indexes causes lot of whipsaws however the same timeframe works better in other instruments like Stocks/Commodities. Ultimately i would conclude that one should select time frames based on the personality of the instrument traded rather than being rigid about using 1H only! [Slightly offtopic but , its one of the reasons why i don't like Mechanical Trading Systems , they are just too rigid, moreover Barclay's Index for Discretionary & System Trading also proves that Discretionary has a slight edge over Systems via the nature of drawdowns,returns etc]
Accepted AnswerVignesh EswarOfflineMr. Sancheti - if you are looking for fully completed intraday bars of the same time frame - then each trading day - 09:15 to 15:30 - has 375 minutes. that break down nicely into 15, 25 minute bars and 5, 75 minute bars. You may consider substituting these timeframes for 15/30 minute and for 60/90 minute analysis. This would solve your - 1 in 7 candles are defective in the hourly bars - problem. But like Mr. Dogra mentions, it all boils down to what you are comfortable with and the personality of the instrument.
Accepted AnswerBhooshan ShanbhagOffline
In my view, the longest timeframe that you can use for intraday charts is 15 minutes. Anything longer than that will have one candle per day completing before the stipulated time. If you want shorter tha 15 minutes, 5 minutes is another suitable option. 10 minutes candles will also show the last candle of the day as incomplete. Moreover, 10 minute candles will not close at the hour, but five miutes after that, as opening time is 9.15. Personally, if I want to use intraday charts, I would prefer 5 minutes.
Also one has to understand one other important factor, which is about moving averages. 15 minute charts with 25 period moving average gives results much delayed that 5 minute charts with 75 periods moving average, the later showing much better performance than the former. Mathematically, both averages are for the identical period, but as the numebr of data points is more with 5 minute chart, the normalisation of data works better. If someone uses 1 minute candles with 375 periods moving average, performance will improve further.