Spoiler: it's highly, highly sensitive to parameters. You might choose to lag the time windows for several months to account for a delay in the effect of new policies, or lead to account for anticipation, or use as the underlying "invest or not" determination a time series reflecting laws, who holds the Executive, who holds the Legislative, some combination... but tweak any of these parameters and the outcome will flip wildly.
Of course this makes it very easy to reverse engineer a particular outcome if your audience is not familiar with the effect.
http://www.demonstrations.wolfram.com/StockMarketReturnsByPa...
Spoiler: it's highly, highly sensitive to parameters. You might choose to lag the time windows for several months to account for a delay in the effect of new policies, or lead to account for anticipation, or use as the underlying "invest or not" determination a time series reflecting laws, who holds the Executive, who holds the Legislative, some combination... but tweak any of these parameters and the outcome will flip wildly.
Of course this makes it very easy to reverse engineer a particular outcome if your audience is not familiar with the effect.