ETuser wrote
Hi Art.
What is the reason that your system does not give the calendar biases for the bonds?
Thanks in advance,
ETUser
__________________________
Good question Et.
the biases work in the indices because they are the most psychologically driven of markets. separate studies of mine, which can be found throughout this thread and elsewhere, verify that the common perceptions about monthly over-and-under performers tend to be correct. you would have done well throughout history to "sell in may and go away." i found that the duration of staying "away" can be 6 months--which is convenient in allowing long and short time blocks to be equal and i love symmetry. (be long november through april, short may through october).
you also see biases within months. this was inspired by the phenomenon of "window dressing"--the tendency stocks have to rise in and around month ends. the reason--it allows fund and portfolio managers to buy the hot stocks in order to advertise that they in fact owned them (never mind that they owned them too late t do any good for their clients). it seemed logical that if there were an over-performing timeframe, there should be an underperforming one also. that resulted in the following pronounced bias--be long from the 22d of the month through the month end up through and including the following 6th. be short from the 7th through the 21st. window dressing drives this--it's psychological. same with being confident in the first month of the year and dubious as the summer malaise sets in. psychology. you don't see it exist to such a degree in the non-stock-related markets.