Result

Hi,

thanks for posting but the strategy do not load on the trader.  The error shows that there is no parabolic SAR (HTF)

)))

Put it in the custom indi folder.

))

Real-Test-invest
IBFX-MT4-Demo2
Login: 7054442
Investor : eha8peh (read only password)
Server: ibfx.com:443

Put it on the test. While the costs. Do not know how long it'll be tested. See as long as necessary.

http://www.fastbrokers.com
http://www.vantagefx.com
Els also suitable brokers rapid execution.

Good to find someone who will fix the indicator will correct the error if it is.

Faults of HTF come up pretty quickly, let it make ~10 trades, cross-reference with FSB and you've got an idea whether it's ok or not.

Hi Kostefan,

HTF bug -- the higher time frame is based on the number of bars, not the time of the bar. The cycle depends on the first bar of data in your data set.
Problem -- the Parabolic SAR HTF values will change most times you open Forex Strategy Trader. Real time trades may not match Forex Strategy Builder backtest trades on same time period (eg, Aug - Sept).
How to see bug:
1) run backtest in FSB, take screenshot of chart
2) open your data file
3) delete first 2 lines
4) run backtest in FSB again
5) compare time of bar for indicator values
ex: Run 1 -- 1:00, 1:05, 1:10, 1:15, 1:20, ... n
Run 2 -- 1:02, 1:07, 1:12, 1:17, 1:22 ... n

WTF fix -- higher time frames indicator values should be on time of the bar, not number of bars
How to verify fix:
1) run backtest in FSB, take screenshot of chart, or equity balance
2) open your data file
3) delete first 2 lines
4) run backtest in FSB again
5) compare time of bar for indicator values
ex: Run 1 -- 1:00, 1:05, 1:10, 1:15, 1:20, ... n
Run 2 -- 1:05, 1:10, 1:15, 1:20, ... n

This was the main problem with the HTF indicators. Users noticed the indicator values changed when they tested the same time period with a different data download. Or, the indicator values in FSB backtest at the end of the week did not match the values they saw during the week in live trading on FST. This happens if you open FST at different times of day varying around the 5-minute cycle -- ex, Monday at 9:03, Tuesday at 9:01, Wed at 9:09. It also appears if you lose a couple of bars of data, or if your data provider varies in giving last day's data at 16:00, 16:01, or 16:02, etc. Lots of problems going over holidays or half-days.

It is possible, on 1-minute, with 5-minute higher time frame, the time and bar sizes are small enough that this difference is not significant. But it is not good to depend on that for testing, for good testing it is best to eliminate inconsistency and unreliability even if it's small. If not you are hoping some random variability will balance out, but this will have unexpected results at sme time in the future since most time-based data have distributions with fat tails.

I know the feeling! And the emotion after that feeling is even more familiar. Long story short - demo it live.

Some time ago I went straight live with a strat similar to this, fortunately it won nearly 50% of trades, therefore I ended in a marginal loss when I found out that it was a buggy indi behind my retirement plan.

Real-Test-invest
IBFX-MT4-Demo2
Login: 7054442
Investor : eha8peh (read only password)
Server: ibfx.com:443

+500 $

now started to VPS