The biased short-term futures price at Nord Pool: can it really be a risk premium?

ABSTRACT We analyze the forecasting performance of the four-week and six-week futures prices in the Nordic power market (Nord Pool), from 1995 to 2008. We find that short-term futures have been biased forecasts for the subsequent spot prices, ie, the futures price significantly overshoots the spot price four to six weeks later. The forecast error is large and has increased during recent years. Given the size of the forecast error (7.4%-9.3% on a monthly basis), it is hard to explain it away as a risk premium only. Someone who had routinely taken short positions in the four-week or six-week contract and reversed this position at maturity would have made substantial profits. The magnitude of the forecast errors, together with the fact that the errors do not differ significantly across seasons and are uncorrelated to risk indicators, may be taken as evidence of market inefficiency.