By looking deeply into the differences that exist between Keynes and the New Classical, Shaw contends price flexibility is a major point of divergence.  The new classical assumed that markets would always clear and achieve a Walrasian equilibrium condition, in which there existed a single price vector for all markets that satisfied equilibrium in all markets.  In Walrasian theory, there is an auction type environment where all perfect information prevails, products are homogenous and transaction costs are zero.  This is what Hicks refers to as flexprice markets; because whatever market follows these conditions will likely have flexible prices. 

The characteristics of financial markets most closely resemble these requirements of flexprice markets.  There is extreme competition, well informed specialized traders act as auctioneers, information is equally available amongst all market participants, and there is a large quantity of homogeneous products (from the perspective of the buyer) being sold in near cost-loss conditions.  And indeed, these markets do have extremely flexible prices and wages.   However, not all markets have these pleasant conditions .  Most markets in the world are not financial, but rather have trading in goods and services.  Shaw runs through at least seven different sources of price inflexibility that can arise under conditions where information is not perfect, products are not homogenous, and transaction costs are relatively high.  Hicks refers to these as fixprice market.

The first source Shaw mentioned arises when there is no auctioneer as there was in the Walrasian auction system, such that there would be tremendous costs for all the market participants to have to congregate in one place to bid on prices.  The price tag is cheaper to use from the perspective of the buyer and seller, because fixing a price for a time period would also reduce search costs for buyers (the price wont change by the second, so they wont have to keep checking on it), and this allows time for the buyers to become aware of the price and respond accordingly.  Adjustment are never instantaneous, so it is infact a good thing that prices are fixed because it allows the market time to adjust. Also, this lowers the associated costs of changing price tags, up until it becomes profitable to do so.  Because the cost of a price change will be the same regardless of the extent of a price change, a firm will wait until the benefits of changing the price exceed the cost of the price change; until that point all adjustment will be output adjustments.

Also, in the case of Walrasian equilibrium, perfect information exists because the auctioneer acts to coordinate all market clearing price adjustments.  We see an example for this in the game theory dilemma of private firms in the face of an aggregate money demand shock.  When the supply of money decreases, a firm will experience a drop in the nominal demand for its products.  But at the same time, the cost of its inputs should decrease as well; if the firm decreases its output prices, and input prices fall, then profits should remain unaffected.  But to lower prices means that they will be earning less profit than competitors, so who will be the first one to make the move?  This lack of coordination will delay the impact of nominal changes in demand, and result in price stickiness and changes in real variables.  And of course, these effects are magnified because there is often a lag between intermediate and final goods markets.  The input cost price adjustment will respond more slowly than the nominal demand change, so the firm is not likely to decrease prices because profits would decrease in the short run.

This is a summary of Chapter 7 from Rational Expectations by G. K. Shaw.