There Might Be Hope for Speculators! from Martin Lindkvist
Even if the random-walk model best approximates reality in the short run, long speculators could still profit from long-run price changes if average upward price changes exceeded average downward price changes. If there is an upward bias, do speculators profit solely and simply because they bear the risks that hedgers transfer to them, or do they profit because they can forecast prices successfully? The 'risk premium' concept is a common point of departure in the literature of futures price behavior.
Whether or not here is a risk premium and, if so, whether it leads to a reasonable expectation of profits, there may be enough observable biases in futures prices to give hope to a speculator attempting to forecast prices.
R. Teweles, F. Jones: The Futures Game - Who Wins, Who Loses, Why? McGraw Hill, 1987