Whatever we might think about the Red Sox, they aren't cheap. But in the baseball landscape, you have two currencies, players and greenbacks.
The Sox have a variety of intriguing bargaining chips (relatively low-salaried players under extended obligation) including Clay Buchholz, Daniel Bard, Ryan Westmoreland, Casey Kelly, Ryan Kalish, and more. If we use the commodity business as a model, then the Sox have a certain 'book value', budgeted salary plus the "proven reserves" model used in mining.
Of course, the proven reserves aren't so proven, but if they are assigned a 'value', trading them for high-salaried players depletes your future in both salary obligation AND in projected value. In other words, mortgaging your farm system to acquire 'expensive' stars makes you pay twice. For the Yankees, the bottomless money pit, spending 29 percent of free agent dollars in the past couple of seasons (per an ESPN article), that's simply business as usual.
For the Red Sox, they need to work smarter not more expensively, and trading the combination of a Buchholz, Kelly, and change for a Roy Halladay isn't as efficient as spending on free agents and developing players.
Clearly, the Red Sox management didn't exactly fall of the turnip truck. You never know what disinformation campaign exists to drive up the price (Boras style) in the Winner's Curse. Everyone can point to the Beckett-Lowell trade as bringing the Sox one title, a gamble worth taking.
But we've gone through the six year productivity of some of the best pitchers in history during the Santana discussions a couple of years ago. Historically, over that timeframe, some of the best pitchers in history (from Clemens to Pedro) have averaged about 15 wins. In the past two seasons, Johan Santana has won 29 games for the Mets, and one of the players rumored to be on the block for him, Jon Lester has won 31 games...at a fraction of the cost. Caveat emptor.