BDB has sixty second EUR-USD binary options although they cap the investment at $100/per trade...then again you could probably make 200 trades in an hour if you were so inclined.
For use on the BDB 60 sec platform, here is a very simple EUR-USD signal system based on closing price only.
Three conditions must be met
1. An 8 min high has been reached (closing prices)
2. The two subsequent 1 min closes are higher than than the previous close.
3. The close at min 10 is 5 pips or more than the close at min 8
60.16% of the time this will result in a downward correction over the next min (again based on closing price)
eg.
Year Date Time OPEN HIGH LOW CLOSE
2009 01/02/09 4:35 1.3860 1.3861 1.3859 1.3859
2009 01/02/09 4:36 1.3858 1.3859 1.3853 1.3858
2009 01/02/09 4:37 1.3858 1.3859 1.3856 1.3858
2009 01/02/09 4:38 1.3858 1.3858 1.3857 1.3857
2009 01/02/09 4:39 1.3855 1.3857 1.3855 1.3855
2009 01/02/09 4:40 1.3855 1.3857 1.3854 1.3855
2009 01/02/09 4:41 1.3855 1.3860 1.3855 1.3858
2009 01/02/09 4:42 1.3858 1.3868 1.3857 1.3866 <--8 min high (close)
2009 01/02/09 4:43 1.3865 1.3872 1.3865 1.3870 <--9 min high (close)
2009 01/02/09 4:44 1.3871 1.3875 1.3869 1.3875 <--10 min high (close), >.0005
2009 01/02/09 4:45 1.3876 1.3879 1.3871 1.3872 <--downward correction
This method was developed analyzing a data set of 946,922 minute by minute EUR-USD data points. ~2.75 years of data. The frequency of occurrence is 1.2% which means you can find ~6 signals per 8 hr day based on 1 min closing prices. By the way the converse is also true...if you find an 8 min low followed by a 9 min low followed by a 10 min low and the change between min 8 and min 10 is 5 pips (.0005) or greater there is an upward correction 61.11% of the time in the subsequent minute. Although the frequency of this situation is only .8% meaning you can find 4 signals per 8 hr day. Still, between the two, this very simple system can give you 50 signals per 40 hr week.
Total downward (put) signals in data set: 12,081. 1 min downward corrections: 7,268 (60.16%)
Total upward (call) signals in data set: 8,138. 1 min upward correction 4,973 (61.11%)
Note: Data (from supplier) was rounded to the nearest pip. I have absolutely no reason to believe this rounding affects the underlying analysis. Signals which resulted in no change in price were eliminated from the analysis. Also note that condition #2 is not critical assuming no big downward spike in min 9.
Im certain there are subsets of this data set that will yield higher percentage signals than the ones I have given here....I'm just trying to make this as simple as possible. Metaphorically speaking I am pointing to a method to find "blacksand data". Valuable enough by itself but containing much more valuable material when filtered.
My conjecture is that looking at tick data rather than 1 min closing price will give you 20-30 signals per day. Possibly 40-50.
My second conjecture is that this system (or some small variant) will be applicable to other forex pairs, indices, commodities or stock prices. In other words any asset on the BDB platform.
My third conjecture is that this signal system is scalable. What I mean by that is that instead of looking at 1 min data, you can look at 5 min or 10 min data and look for the same patterns. My observation is that chart patterns are similar to fractals in that small patterns replicate to form larger patterns isomorphic in shape but larger in size. The scaled signal system could then be used on the BDB option builder platform.
m99
Followup: I just tested this on 7 years of 5 min DJIA data, the put signals are at 57%-58% which is mildly useful, but you only get one signal per day. The call signals are at 54-55% which is not useful. So scalablity is suspect without additional conditions. Anybody have any 1 minute DJIA data?
|