disappearing crossovers?

[dottom]

Quote from ArchAngel:


That is a coding issue - it's one thing to update a stochastic as the current bar changes and quite another to generate an actual "signal" intra-bar. The generation of an actual signal should only occur at bar completion.

If a signal alert was generated prior to the bar completing, that would be a coding error.

The original poster was asking about "disappearing crossovers" which is most likely that the charting software is updating indicators & price in real-time, hence you see a crossover b/c price temporarily moves in that direction, but then retreats before the bar is complete. That's why I said it's not a coding error b/c the observation of a "disappearing crossover" is due to observing realtime charting.

 

[dis]

Quote from SocalTrader:
We show you those MA crosses that are statistically confirmed.

Gee.. and I thought MA crosses were deterministic in nature

 

[Lobster]

If there is one piece of advice I can give an aspiring trader about TA, especially MAs, it would have to be:

Develop your system by backtesting on data of days with even dates only (or every third trading day or something like that). Then when you finally have a profitable system, run it on the OTHER days only and see if it is still anything as profitable as on the days you used for developing it. If not (which will most likely be the case), forget that system and move on.

TA is all good, but you can't just apply TA to the market and money will be generated automatically. That's like getting an MBA and thinking money will start flowing in now that you know how to run a business.

 

[SocalTrader]

If you are dealing with daily candles only perhaps you are right. The exchanges make an official statement of the closing values usually delayed by no more than 15 minutes. However these things do change between the close and the official announcement of the closing price.

Now imagine the same problem for one minute candles. Do you really believe that different data provider ticker plants will have identical candle sticks especially at the edges? That is one reason why moving average crosses are not deterministic.

Our confirmations are most noticeable when the current price is near a slow moving average like the 200 day. When this happens it is easy to see a lot of noise and easy to believe that one person's chart will show a cross-over while anothers may not. Or you could see it cross over and back very quickly.

The original subject was disappearing crossovers. It is common to see a bad print that don't appear on later charts as data is cleaned up. Our statistical analysis is designed so that no one bad print causes you to pull the trigger on a trade.

 

[LongShot]

your ma's will change if you have one of those time checker clocks installed like atomic clock etc. they change your computer time and hence your charts. otherwise sierracharts does NOT have any disappearing crossovers. (and neither does QT)