Saturday, July 12, 2014

The Momentum Curve: Expanding the Search for Market Profits

Let's say an explorer discovers gold in the mountains of Alaska.  The initial group of miners prospecting the territory has a pretty wide open field of opportunity.  They don't need sophisticated equipment to see and extract the gold from river beds.

As more prospectors move in, the low-hanging fruit is gone.  Now gold must be extracted from rocks and from deeper in the ground.  This requires special equipment.  The single miner with his pan, combing through the river bed debris, no longer has an "edge" in discovering gold.

Still later, as more of the territory is mined, extracting what remains becomes a more complex task.  Deep drilling into the ground and exploration of more remote mountain areas is required to make the investment of time and effort worthwhile.  Individual miners, picking through areas that have already been explored, have almost no advantage in discovering gold--even though they still recount the stories of big strikes just a few years prior.

A well-mined area means that either you have to find new areas to explore or you have to find new means of exploration.  In the case of natural gas and oil, fracking has been a new mode of exploration.  Drilling in the Arctic would be an example of finding new areas to explore.  Either way, ingenuity is required to find value once others have been searching for a while.

Financial markets have been well-mined for a while.  Excellent traders, portfolio managers, and system developers around the world have been attracted to the gold rush of markets.  While looking for nuggets of profitability in new ways and in new places does not guarantee success, looking for them where others have been searching for years with sophisticated tools inevitably invites failure.

I recently have been posting on the topics of understanding vs. predicting markets; looking at markets in new ways; and using quant processes to aid discretionary expertise.  The common theme is becoming better at the exploration for profits by looking at new things and looking at old things in new ways.

Above is a chart of what I call the Momentum Curve.  It takes every stock in the SPX and gauges whether it is trading above its 3, 5, 10, 20, 50, 100, and 200-day moving averages.  The aggregated data are charted (available through the excellent Index Indicators site), so that you can see how the percentages of stocks shift over time.  Observe from the graphic that--going into Friday's session--we had been undergoing a meaningful short-term correction (most stocks moving below their 3, 5, and 10-day averages) in a strong uptrend (most stocks above their 100 and 200-day averages).  

It turns out that the shape of the Momentum Curve is important in forecasting future stock market returns.  The return profile looks very different depending on where the kinks are in the curve, whether the curve is steep or flat, etc.  I find it useful as a qualitative tool--it provides a quick visualization of where we stand across multiple time frames--and also as a source of quantitative hypotheses regarding curve shape and forward price movement.  

Some of the best market tools don't generate conclusions.  Rather, they suggest hypotheses worth testing.  The first step in finding fresh answers is asking fresh questions.

Further Reading:  The Psychology of Quantitative Analysis