This is not true and asserting it is so betrays shocking ignorance of how Pearson product moment correlation is calculated.Quote from bhardy307:
Perfect negative correlation would produce a zero net result
You can prove this pretty easily in Excel. In the first column draw simulated returns from a normal distribution:
in a1 through a1000 --> = norm.inv(rand(),0.001,0.01)
In the second column enter this formula:
in b1 through b1000 --> = 0.002 - 0.1*a1
then calculate the correlation:
in d1 --> = correl(a1:a1000,b1:b1000)
You will get a correlation of -1, or perfect negative
correlation. Yet both have the same expected mean
return (0.0001). Trading the two together produces
a smoother rising equity curve than trading either
alone.
