The prediction markets reflect the polls and the national media.

Jason Ruspini:

In general, futures markets are “less futures markets than immediate-past markets”.

It is hard to evaluate the markets fairly given how difficult it is for someone in New Hampshire to participate in them. Naturally then, the [prediction] markets more easily come to reflect the polls and national media – which were wrong in this case. (Here, the left-leaning media fell in love with Obama, while the right-leaning media happily went about trying to bury the Clinton “dynasty”. Consider the tremendously misleading FOX NEWS clip that PoliticalBetting.com linked to over the weekend.)

The legal situation also damages liquidity to the point where these markets are usually manipulated at the margin, as traders with momentum sense a rout and begin to engage in predatory trading, i.e. pushing prices to extremes to force capitulation and/or margin blow-ups.

Dr. Servan-Shrieber’s comment suggests that these market errors might persist even in modern regulatory regimes like Holland, and corroborates the excellent Erikson/Wlezien paper. Leaving aside the question of significance, Erikson and Wlezien have apparently found a market inefficiency, but since markets can take this into account (they are a “meta forecasting tool“), we should expect it to dissipate over time.