IBM: How much for this Elephant?

“No amount of sophistication is going to allay the fact that all your knowledge is about the past and all your decisions are about the future.” – Ian Wilson

Valuing any company in the middle of a major transition is tricky. I saw this, when I tried to value Staples, recently. Usually, history is a good guide for the future.

For instance, if I want to guess what the weather will be like tomorrow, I can do pretty well with just two data points: (i) the weather today, and (ii) (seasonal) averages for each day of the year.

If a business is not cyclical (pharma company), then my expectations of tomorrow are shaped largely by an extrapolation of the recent past.

If a business is cyclical (oil company), then knowing where I am in the cycle is useful additional information. If the recent past has been anomalous, then knowledge of the full cycle can help us discount it.

When a company, like IBM, is drastically reshaping itself, the utility of the past is somewhat diminished.

I use two simple ways to value IBM.

Multiple of pre-tax Income

For high quality businesses, Buffett has this famous “10x” thumb-rule, i.e, he is happy to pay up to 10x pre-tax income (PTI).

For 2015, IBM’s PTI was \$15.9B. Dividing by the number of shares (0.965B), I get a per share PTI of 15.9/0.965 = \$16.48.

Therefore, according to this rule it might be okay to pay up to 10 * 16.48 = \$165/share.

DCF

A stated goal, according to management, is “high single-digit EPS growth with return to shareholders”. In the recent past, IBM has essentially paid out all of its FCF in the form of dividends and buybacks.

One can temper these two facts, adjust them downwards, and come up with some conservative PV using a dividend discount model.

The current EPS is \$13.50. I can ramp up the growth rate from 2% to 5% in years 4 and 5, and then ramp it down to 1% in year 10. This is loosely based on the idea that 2016 seems to be an inflection point in IBM’s strategic imperatives. I assume that growth resumes from 2017, and can be sustained for 3 or 4 years – hits a peak of 5%, and then tapers down to a terminal 1% by year 10.

Using a discount rate of 10%, and a terminal growth rate of 1%, I come up with a DCF estimate of \$176.50/share. Subtracting nearly \$5-6/share of net debt, I get an estimate of about \$170/share for the equity.

Miscellaneous

Historically, IBM has traded between 8x and 12x TTM EPS. Given this years projected EPS of \$13.5, this implies a range of \$110-\$160. If earnings increase, then this band will shift.

Regardless of how you slice it, IBM’s fair value based on the recent past, overlaid with conservative assumptions, comes out to be, say, between \$150-\$170.

It is currently trading at slightly over \$160. This probably implies that it is fairly valued.

If earnings per share pick up faster than anticipated, then the fair value will increase. Otherwise, one should expect a high single digit to 10% return on IBM, if purchased at this price.