BBY Growth Pattern

I haven't looked at BBY in several years, so this is off the cuff.

Is your model looking at historical results or is it strictly for future projections? Either way my answer should hopefully help.

If you look at BBY since inception or as far back as there is publicly available data, you would see three clear stages in broad terms. There would be an initial stage of high growth as the store network was built out and they gained market share. The second stage would be a flatline or stablization stage with limited growth and a mature store base in a mature market. Sales may have been up during the housing ATM bubble when people bought new 50" plasma TVs or whatever, but that should still be considered the mature stage on a normalized basis. The third stage is decline, which we are almost certainly in. I think if you're modeling this on a long time horizon, you have to assume that all of the per store metrics will deteriorate -- inventory turns, sales / store, sales / sq ft, margins will contract, etc.

The bottom line is that BBY is a high fixed cost capital intensive business with structurally negative trends. Amazon is eating their lunch. Wal-Mart has cut in. Piracy hurts. And the general digital trend is not their friend (pretty soon you're going to have game downloads for XBox, etc. -- no need to buy a hardcopy at BBY anymore). I am about 95% sure that if you were to look back at all of the major pure play electronics retailers that have existed since, say, 1950 just to pick a round number, that none of them are still alive except BBY (Circuit City died a predictable death in 2008 and went into liquidation -- worth more dead than alive). Best Buy is going to try to restructure their store base and cost structure, but big box is extremely hard to turn once it sours. At the end of the day they have massive fixed costs, including very expensive fixed cost leases, for a concept that is going to permanently lose volume over time. There might be upward spikes based on increased consumer confidence, etc. but the ship is slowly sinking.

To make actual EPS and sales projections, it would make more sense to do it on a per store basis rather than pick a headline number -- model it from the bottom up in a sensitivity model.