The recent run-up in the price of silver, in response to the Fed unleashing the dogs of QE3, has reignited interest in silver as an investment. Can the run-up continue? Let’s look at some phase space reconstructions of the silver price, which I will use for a discussion which should not be construed as investment advice.
For today’s charts I am using the Friday PM London fixes since 2007, which can be obtained here. This is approximately the week-end closing price for silver over the past 5-1/2 years.
Note the trough in late 2008 and the spike last year. I’m sure we all remember those.
Reconstructing a phase space is one method of illustrating the dynamics of a complex system. I have previously described it here and here. I actually experimented with several different lags before choosing one of six weeks, as depicted below. Choosing 12 weeks or 24 weeks does not significantly change the results.
As related previously, dynamic systems display different forms of stability. Exponential stability leads to point attractors, in which the function tends to evolve towards a fixed point (called a point attractor) or, alternatively, a cycle. Lyapunov stability is a little weaker, in that the systems tends to remain confined to some region of phase space (a Lyapunov-stable area, or LSA), but tends to wander around in that region.
Many natural systems are characterized by episodes of stability, punctuated by brief periods of rapid change. The episodes of stability are each commonly considered to represent a metastable equilibrium, and represent periods of time in which the dynamics of the system are dominated by negative feedback. The episodes of rapid change occur when the system is driven by positive feedback.
I think (but have not proven) that the mechanism for positive feedback in stock or commodity prices is momentum trading (or possibly trend following), and the mechanism for negative feedback is mean-reversion trading.
There are three areas of Lyapunov stability in the above graph–one centred around the $13 level, one at about $17 level and the last ranging from $27 to $34.
The implication here is that the recent rise in the price of silver only represents motion within a region of stability, not an excursion which we could expect to lead to a price collapse.
Going forward, the only regions of stability are either roughly at the current price or substantially lower. There is no sign yet of any regions of stability at a significantly higher price than today. It will be hard work tracing out such a new LSA. Consequently, I would not argue that silver should be shorted at these levels, but acknowledge that a shorting opportunity could occur on any rapid price increase from the current level.