This past week, I got a very nice e-mail from a reader of my book who lives in Hong Kong. I could tell he was knowledgeable about the balance sheet. He wondered why the balance sheet does not play a more important role in investing and why so much emphasis is placed on future expectations, growth and earnings. If you’ve read my book and followed my blog posts, you know that this reader is in the same camp of thought as I am.

The aim with stock investments is, of course, to buy at low prices and sell at high(er) prices. We have to get the first part right (buying low) before the second part (selling high) becomes possible. My position is that without the balance sheet, it’s not possible to define a low price. Earnings and growth expectations must be considered when buying, but should not trump the balance sheet’s primary role in finding a low price.

The kind reader posed an excellent question about how to best consider earnings at the time of buying a stock. My practice described in the CSW book is to add the past 12 months earnings per share (eps) figure to my initial floor price in order to derive the adjusted floor price, when there are no analyst projections for the “next” 12 months eps. Oftentimes, there are no analysts covering small company stocks so with those stocks we are often without a forecast of the next year’s eps.

Recall that the initial floor price is found by assessing the balance sheet only. So, by adding 12 months eps, consideration is being given the earnings at the time of assessing a low price. Note that if you have not read my book, some of this terminology (i.e. initial floor price) will probably seem strange.

The reader wondered if, when having to use past earnings to adjust the initial floor price,  it would be better to use a multi-year average annual eps number (extracted from past years) rather than using the past 12 months eps, when adjusting the initial floor price. While I gave a more exhaustive answer during our electronic conversation, I’ll give a simplified version here.

Rarely would the eps number be significant relative to the size of the initial floor price number. So, the precision of a past eps number selected is less critical to the final floor price (adjusted floor price). That said, we still don’t want to ignore earnings altogether but rather subordinate the role earnings play to the role of the balance sheet when defining a low price. Using the past 12 months of eps is the most recent number available in the absence of a forecast of the next (forward-looking) 12 months eps. A more recent number is likely more relevant than older eps numbers.

However, an annual eps number extracted as an average of multiple past years might well be more appropriate if the past 12 months eps are anomalous (highly unusual) due to inclusion of non-recurring (one-time) earnings, like a gain on the disposition of a segment of company operations, for instance. That is, an average number from several years could mitigate somewhat for any unusual events (unlikely to be repeated in the future) included in the past 12 months of earnings.

We don’t actually need to go back several years to mitigate for an anomalous effect in the past 12 months eps number, though. Rather, we can go to the Securities and Exchange filings (see my prior blog posts on corporate financial filings) and determine if any non-recurring amount is included in the past 12 months eps number and simply exclude this one-time part from the eps number we blend with the initial floor price.

I appreciate the astute reader from Hong Kong sharing his thoughts with me and posing his excellent question.