Or at least, that’s how I would like things to turn out. I don’t really know.
Adventures are exciting precisely because they are unpredictable. The best parts of an adventure are usually those elements which were least foreseeable at the outset.
The well prepared adventurer understands this, but he nonetheless packs a map and charts a path. He then maintains good humor when circumstances inevitably force him to abandon that path and improvise.
It is in that same spirit that I lay out my plan of attack for this blog:
Our first few forays into the financial data stream will use daily historical data for S&P 500 equities. Using this data we will:
- Systematically assay the data’s statistical properties
- Attempt to isolate individual equity signals from broader market signals
- Transform the signals so they are more receptive to analytic methods, and
- Derive a model from first principles which attempts to explain our observations.
We will then add historical options data, and attempt to determine whether any identified market inefficiencies can be translated into viable trading strategies.
Throughout this journey, I will declaim with great passion against the main villain of this blog: spurious signals. I will denounce methodologies that have a tendency to produce spurious signals. I will then use handwaving to justify my use of those same methodologies when I determine them to be the best available option.
Readers should expect occasional detours to address other questions of interest. These may include:
- Under what conditions of performance is investing in an Alpha generating hedge fund a good idea?
- How have the statistical properties of the market evolved over time?
- Is there a firm basis for attributing observed changes in market dynamics?
While I can’t guarantee definitive answers, I do promise an adventure.
But first, a few introductory posts are in order.