Generally Accepted Accounting Principles, or GAAP, are used by all public companies, but GAAP allows for a lot of freedom in classifying assets and earnings. What this means is that, when looking at a statistical screen, we may not always be comparing apples to apples. Take a simple, and relatively common, mechanical screening criteria - price to book ratio. How a company calculates book value can differ greatly. Take, for example, Harley-Davidson (HOG). The main competitive advantage for this company is their brand name - but this asset is not recorded anywhere on the balance sheet, so HOG's price-to-book looks higher than a company that does record branding as an intangible asset (like, say, Coca-Cola (KO)). Peter Lynch, in his book One Up On Wall Street, talks about how Pebble Beach's gravel pit real estate asset was worth several times what it was recorded for in the balance sheet! Without due diligence, these kinds of things are impossible to know by using just a mechanical investing approach.

Joel Greenblatt, in The Little Book that Beats the Market, spends a lot of time talking about this point. All too often, an investor decides to base his strategy on a mechanical technique, just to find his portfolio trailing the market after a year (or less). These results lead to the emotional conclusion "this doesn't work!", despite years of data showing that it does. The investor drops the strategy, which then goes on to crush the market for several years. The Magic Formula, and many other mechanical strategies, can be volatile. There will be periods where they trail overall market returns. The mechanical investor must have a strong will and conviction that the strategy will outperform the market over the long term.

Probably an offshoot of the last point. For many investors, blindly buying stock in a company without knowing anything about it seems risky and foolish. There is not much else to say on this one. It is a major hurdle for many people, and for some people it will not bother them at all.

Clearly the biggest problem with mechanical investing is the "driving using the rear-view mirror" aspect. What this means is that you are investing based on past results, instead of taking into consideration future prospects. After all, it is future results that will determine the success or failure of your investment, not past performance. Sure, Crocs (CROX) were all the rage the past few years, earning a ton of money, but is it likely this fad will earn those same dollars going forward? Maybe that low P/E ratio is warranted. Many large banks looked downright cheap on a price-to-book basis at the beginning of the year. However, as huge write-offs were announced, the "B" part of that ratio dropped in a hurry, and now those "cheap" prices don't look so attractive. Doing some fundamental research could have at least raised red flags on these potential future problems, something a mechanical screen would not have found.