Flickr / geezaweezer.
Albert Einstein once admonished that if you can't explain a concept in a way a six-year old would understand it, you probably don't understand it yourself. The idea is that if you have to use jargon or highly complex explanations, you're probably covering something up -- in this case a lack of expertise.
How does this relate to company valuation? It turns out that how a 10-K filing is written is just as informative as what's inside. You'd be wise to pay attention when you're falling asleep: one researcher found a significant correlation between the readability of a 10-K and the stability of earnings in future years.
Using linguistics to measure earnings
What exactly makes something readable? One way is to use the Fog index, which looks at a document and takes the number of words with three or more syllables and the number of words per sentence and plugs them into a formula. The resulting score is roughly equivalent to the number of years of education you need to understand something on a first reading.
The idea is that the more arcane the language and the longer each sentence, the harder it is to process the information.
Making lower earnings look better with complex language
Using the Fog index, Professor Feng Li at the Ross School of Business at the University of Michigan looked at the complexity of 10-K filings to see if there was a relationship between readability and earnings. He found that companies with lower earnings tended to use more complex language in their reports.
While this effect was small in terms of dollars, Li also found that the persistence of earnings was significantly tied to complexity. In other words, profitable firms that had less readable 10-Ks also had less stable profits -- in a way that made a significant economic difference.
Don't all 10-Ks make readers want to cry?
Well, yes. While the Fog index of the average Wall Street Journal editorial is about 15.2, the average score for 10-Ks was 19.2 -- significantly harder to read.
But, as it happens, they don't need to be quite so complicated. Li finds that reports vary widely in complexity, even within the same document. He notes that "the MD&A section of the report is much easier to read than the document as a whole," and there is an enormous amount of variation among reports from one year to another. From the group of hardest-to-read reports in one year, only 44% remain in that category in the following year.
What on earth is going on?
Perhaps they're hiding something. "I find that, indeed, the positive earnings of firms with 'foggier' or longer annual reports are less persistent," Li said in his paper. In other words, more complexity in a profitable report could mean bad news is on the horizon. The same holds true for the length of the report: Longer reports are associated with less persistent positivity.
Whether this implies deceit or just a correlation between complexity and performance is not answered by Li's research, but it's certainly food for thought.
So pay attention to how those 10-Ks are written, and if you're falling asleep or completely zoning out, it might not just be a need for coffee. It could be that there's more to this company's 10-K than meets the eye.
