Micron Technology (MU) is a maker of DRAM and flash memory that, numbers-wise, appears to offer a compelling deal. As I dug deeper into its annual reports it began to remind me more and more of another company I’ve already examined: KLIC. It also got me thinking about an article by James Montier about the power of simple quantitative models.

Both P/E and P/FCF numbers are good: 4.2x and 3.19x respectively. P/B is below market and industry averages at roughly 1. Return on invested capital and cash return on invested capital are 16.5% and 26%. The balance sheet is healthy, with a rising current ratio and cash in excess of all long- and short-term debt. The numbers are happy.

The industry outlook and the footnotes are not. DRAM prices are expected to fall in 2011 and MU’s biggest joint venture (which produces almost half its DRAM) will terminate mid-year unless an agreement is reached with the other partners. The company also has big capex plans totalling between $2.4B and $2.9B for the coming year. That exceeds the company’s typical $2B depreciation and might burn through some of that excess cash. Qualitiative factors: much less happy.

The disparity here got me thinking about the Montier article. The questions I ask myself is: am I and other investors overreacting if we pass on Micron due to those impending concerns? Montier mentions overconfidence as a reason that people reject simple quantitative models, but maybe lack of confidence in the model itself is an equally important reason. Perhaps sometimes the numbers suggest an investment that appears too qualitatively unappealing in spite of its low valuation and solid fundamentals. Either way, it’s an issue of too much meddling.

Taking another page from Montier, I played around with a reverse-DCF calculation to gauge what investors are expecting from MU. Figuring a cost of equity/required return of 15% to account for the company’s volatile earnings and constant growth to make the calculations less miserable:

k = .15

FCF1 = FCF x (1 + g) = $2.48B * (1 + g)

Market Price = $7.9B

g = ?

P = FCF1 / (k – g) = [FCF x (1+ g)] / (k – g)

P x (k – g) = FCF x g + FCF

(P x k) – (P x g) = FCF x g + FCF

(P x k) – FCF = (FCF x g) + (P x g) = (FCF + P) x g

g = [(P x k) – FCF] / (FCF + P) = (-1.295) / (10.38)

g = -12.48%

For k of 12%, g = -14.8%

Now, the numbers are a little skewed by the fact that next year’s earnings are expected to be significantly lower, but the general trend is clear: investors expect nothing good from MU. Even if you factor in the reduction in earnings assumed by the forward P/E of 7.7 and work forward from there, implicit long term growth still works out to be -5.2%. That’s still pretty pessimistic.

I figure Micron will be a bit of an experiment – I’ll watch from here to see where it goes. It might even be interesting to make a broader test of the idea to see if people really are overconfident in rejecting the results of quantitative models…a sort of Magic un-Diligence. But, that’s a project for another day.

Current Price: $7.90

Latest 10-K: http://www.secinfo.com/dm25q.r51.htm