I consider myself to be something of an amateur historian when it comes to the different versions of The Intelligent Investor.
First things first. There are 6 different versions of the book. The first 5 versions have been labeled in a confusing way that has made people over the years think there is one fewer than there really is. Why? Because you'll notice that the second edition is referred to as the "first revised edition" and the third edition is referred to as the "second revised edition" and so on. This has made people think that "second revised edition" means second edition when that actually makes it the third edition.
It is much easier to refer to the 6 different versions by original publishing date
- 1949
- 1954
- 1959
- 1965
- 1973
- 2003
For the most part, you may disregard 1949, 1954, and 1959 versions unless you want to read them out of historical curiosity. The 1965 edition contains all of the content in those previous versions but also includes updates, new material, and new formulas. Largely, the only thing omitted in the 1965 version that is present in previous versions is the outdated examples of individual stock issues in the analysis chapters. They were, of course, replaced with up to date examples for the time.
The 1965 edition is the best edition of the book in my opinion. It contains the most practical advice for passive investors, and the most comprehensive advice for enterprising investors. The famous "Graham Formula" is present in this edition as well. This tome is a true gem that reads very much like an "essentials" version of the much larger text Security Analysis. I personally own 2 copies of the 1965 version and guard them carefully.
The 1973 edition represents a significant departure in both philosophy and approach relative to earlier revisions. Graham eschews the detailed approaches for finding bargain issues found in previous editions in favour of more systematic filtering systems. These were clearly intellectual precursors to modern factor funds. The book, overall, is somewhat disjointed as Graham retained some chapters from earlier versions but inserted entirely new chapters with new machinations (not just updates). There are inconsistencies such as Graham's recommendation of a maximum P/E ratio of 15 for defensive investors in chapter 14 but stating a higher maximum ratio in chapter 4. Again, this stems from the fact that the book is really an amalgamation of old and new essays. This may have been rather expected given that Graham credits his collaboration with Buffett on this edition. The real gem in this revision of the book is chapter 12, which deals with accounting confabulations that can distort EPS numbers.
The redheaded stepchild in the bunch is the 2003 "update" of the 1973 version of the book with Jason Zweig's commentary. I don't even want to waste energy expounding on this nonsense. Reading Graham's elegant writings mixed in with Zweig's rubbish such as the "Vodka and Burrito Portfolio" (yes that's an actual quote from the book) makes this a tragic tarnishing of Graham's work.
Bottom line, stick with the 1965 version if you want the pure Grahamian approach to security analysis. Read the 1973 version if you want a more "modern" filtering system with echoes of Buffett's influence (and the ever so faint foreshadowing of EMH). The 1973 version pretty much screams for a small cap value + quality fund.