|
PureBytes Links
Trading Reference Links
|
This thread has stimulated a lot of thinking for me. A few years ago I
experimented with various moving averages of the equity
curve and also looked into analyzing the equity curve as a data stream
in order to come up with some advance idea as to the
performance of the system. However, I have concluded that the most
important issue to understand is why systems stop
working. Markets constantly change and it becomes clear that the reason
a system fails to perform during any given time
period is because the decision rules of the system no longer fit the
market conditions at the time. Of course this seems
elementary, yet it leads to the whole question of self adaptivity and
optimization. In analyzing many different compressions of
the same security for the same period, ranging from tick to daily by 1
minute increments, it is possible to determine some
relationships of noise to identifiable trending behavior as that ratio
moves through the various compressions. On a simple level,
it is clear that the smaller the compression, the more noise, however it
does not follow that every compression yields a logical
progression of that ratio. It is also instructive to analyze different
time frames in each of the compressions to understand how
this ratio evolves.
We know that markets go through periods of congestion and periods of
direction. We also understand that there are
relationships between securities, and various indices. Systems work
when whatever paradigm of market conditions that were
designed into the system exist. When the market has evolved into an new
circumstance, the system stops working and we
begin lose money. Confidence in the system is really based on the idea
the we believe that the market will return to the former
iteration that produced results in the past. Depending on the level of
refinement of the systems decision rules, that may or not
happen. For example, we all know that over enough time and in the right
compressions, moving average systems can produce
good results. But we also know that these systems go through periods of
deep draw downs. Perhaps the reason the simple
systems produce better results over time, is because they contain a more
common market condition and the system is designed
to, by it's very nature, ignore more of the intermediate market
perturbations.
What is clear is we can organize the data that is created by a security
in many different ways and that the act of organizing the
data is significant to the ultimate potential to be systematized.
Perhaps we bring order to the chaos of a data stream by that
very act. We can also change the ratio of noise to trend for any given
time period by altering the compression, however what
we cannot do is predict that ratio in the future. It constantly
changes. This is the challenge when we try to create evolutionary
systems that can adapt to market conditions.
Thanks to all who have contributed to this thread,
Lawrence Price
|