John, thanks for your detailed answer. But let me critique your approach in greater detail by honing in on one of the critical assumptions you make that seems to underpin your entire approach: book value is a valid basis for valuation.
You said I know that stockholder equity is an accounting fiction. But it is a useful fiction. It is useful because everything I want to know -- earnings, dividends, and stock prices -- are related to equity/share. Once I forecast equity/share forecasting earnings, dividends and stock prices is much easier.
First, let's take the case of acquisitive companies (LU is a good example). Depending on the accounting method employed (purchase or pooling) you can get widely differing book values in spite of the fact that corporate cash flow is exactly the same. Furthermore, the widespread practice of writing off IPR&D under cash accounting leads to a precipitous drop in book value when compared to amortizing the difference (goodwill) between the cost of the acquisition and the market value of the net assets. Of course, at the end of the amortization period the difference disappears. But that can be 10 years which is an eternity in technology companies.
Second, some companies are very aggressive in revenue recognition, leading to an overstatement of earnings, and hence, an overstatement of book value. Companies like the former HBOC and OXHP come to mind.
Both of these effects lead the unwary to the assumption that book value is growing more rapidly then it would under more conservative accounting treatments.
So, I would suggest that discounting free cash flow is a more rational way of analyzing the value of a company because it discounts cash flows rather than accounting estimates (which is consistent with the spirit of DCF analysis). Let me illustrate briefly:
Suppose you are considering an investment in a process that is expected to to generate $150MM net cash per year, and cost $500 MM. Let us further assume that at the end of 5 years the process will have no salvage value. Using accounting practices we would show earnings of $50 MM per year, and assuming a 15% risk adjusted discount rate we get an NPV of $167.6 MM. But this approach is incorrect because it ignores the timing of the cash flows. Substituting the actual cash flow of $150MM we get an NPV of $2.82MM ($502.82MM - $500MM). This kind of analysis is completely lost when you focus on book value.
I would feel much more comfortable with your approach if you used free cash flow analysis, which, by its very nature, takes the timing of cash flows into account, and cuts through much of the accounting mythology. It still fails in one important respect -- it fails to take into account the off-loaded costs of issuing employee stock options (but that is also a failing of your method).
The major problem I have with LU is that heretofore, operating cash flows and free cash flow has been weak, owing to a combination of high inventory levels and high A/R levels. I have not yet analyzed this quarter's results, so I will withhold comments here.
My final criticism of your use of book value in your method is that it assumes that stock price is somehow determined by book value, and you do this by assuming a terminal value for the stock. By doing so you assume your conclusion. I understand that your calculation is aimed at determining the value of the stock to the individual investor, but how can you apply a DCF approach without estimating the terminal market value of the equity?
Sorry to ramble on,but this is a topic dear to my heart.
TTFN,
CTC |
