hey guys,
I am trying to make a spread vs volatility index that compares crossrates based on their spread and volatility. The idea behind this is to have some sort of ranking for my crossrates that tells me that certain pairs are "cheaper" to trade because the spread in relation to their volatility is lower then that of other pairs.
I've tried to calculate that index with the following formula:
In a daily chart with an atr 14, take the lowest atr (That is supposed to be the lowest volatility, or worstcase volatility of the given timeframe). Then:
ATR / spread = "volatility spread index"
so for example, lets take the euro, we would have:
0,00013(oanda spread of 1,3 at this time) / 0,012 = 92,3....
But for other pairs this formula does not work so well:
xagjpy has an atr(14) of 34, and a spread of 0,0032. That would mean an index of 10625 which would "outperform" every other crosspair by far... so well that it is unrealistic, and the chart shows the same. So I guess my "method" is fundamentally flawed.
Anyone got an idea what my error is? Or maybe someone got a better method?
thanks!
I am trying to make a spread vs volatility index that compares crossrates based on their spread and volatility. The idea behind this is to have some sort of ranking for my crossrates that tells me that certain pairs are "cheaper" to trade because the spread in relation to their volatility is lower then that of other pairs.
I've tried to calculate that index with the following formula:
In a daily chart with an atr 14, take the lowest atr (That is supposed to be the lowest volatility, or worstcase volatility of the given timeframe). Then:
ATR / spread = "volatility spread index"
so for example, lets take the euro, we would have:
0,00013(oanda spread of 1,3 at this time) / 0,012 = 92,3....
But for other pairs this formula does not work so well:
xagjpy has an atr(14) of 34, and a spread of 0,0032. That would mean an index of 10625 which would "outperform" every other crosspair by far... so well that it is unrealistic, and the chart shows the same. So I guess my "method" is fundamentally flawed.
Anyone got an idea what my error is? Or maybe someone got a better method?
thanks!