For VCE purposes what observer7 has said will suffice, however I am going to elaborate a bit more on this.
First it depends on what efficient market you are assuming the market to be in, Fama-French proposed 3 stages of market efficiency, weak, semi-strong and strong efficiency hypothesis (
http://en.wikipedia.org/wiki/Efficient-market_hypothesis). If the market was in strong form, then in fact shareholders will be indifferent in investing in Coles or Safeway assuming they are bidding in a perfectly competitive equilibrium market (
http://en.wikipedia.org/wiki/Economic_equilibrium), so infact, little trading will occur between shareholders of Coles/Safeway so the price war will have no effect on shareholders. In this case, it can be said that ALL information (public and private) are already reflected in the share prices, so no new value will be given to shareholders through price alone (although for theoretic purposes we are assuming a perfectly equilibrium market, obviously if market inefficiencies exist, price definitely has value)
However, if we assume weak EMH, then this does not require that prices remain at or near equilibrium, but only that market participants are able to systematically profit from any momentary market inefficiencies. In this case, using technical (
http://en.wikipedia.org/wiki/Technical_analysis) or fundamental analysis (
http://en.wikipedia.org/wiki/Fundamental_analysis) could yield abnormal returns, however previous studies have shown that on a risk-adjusted basis these abnormal returns do not differ by too much of a margin, hence transaction costs could easily wipe out the profit.
So essentially one must make a decision on which stage the market is in, in terms of efficiency, then based on this decision one can then assess what are the benefits/costs to shareholder value. Note simply asking whether it is "good" or "bad" is very broad, rather the point of every company should be shareholder maximisation, this aligns the vested interests of all stakeholders and reduces agency costs, so asking whether a certain action is "good" or "bad" for shareholders may not necessarily mean it will benefit stakeholders in general. Rather, maximising overall benefits will provide the biggest value to ALL stakeholders.