1. Lecture 4: Firm Income Tax Discounted Cash flow, Section 2.1 © 2004, Lutz Kruschwitz and Andreas Löffler

2. No. 2 2.1 The unlevered firm Companies are indebted, i.e. levered. Why should we consider unlevered firms, i.e. firms without debt? Valuation requires knowledge of cash flows  from business plans, annual balance sheets, etc. taxes  from tax law cost of capital  from «similar companies» What is a «similar company»?

3. No. 3 2.1 «Similar» companies Companies are similar with respect to: business risk financial risk (=different leverage ratio). We eliminate the financial risk by determining the cost of capital of an unlevered firm («unlevering») and then of a levered firm («relevering»).

4. No. 4 2.1 Notation levered firm index l unlevered firm index u free cash flows (after taxes) value cost of capital

5. No. 5 2.1.1 Value of unlevered firm We have, analogously to Chapter 1 (even with the same proof!) Theorem 2.1 (value of unlevered firm): With deterministic cost of capital

6. No. 6 2.1.2 An important assumption Aim: We want to show that for the unlevered firm, costs of capital (expected returns) are discount rates as well! Means: An assumption is necessary, which was already used by Feltham/Christensen, Feltham/Ohlson, also everyday business in statistics: «autoregressive cash flows». Remark: This assumption will concern the stochastic structure of the unlevered cash flows.

7. No. 7 2.1.2 Independence vs uncorrelation «Autoregressive cash flows», also called AR(1): In finance the usual assumption is: noise terms are pairwise independent. But here we only assume that the noise terms are pairwise uncorrelated – which is less restrictive: are independent if for all functions f and guncorrelated if

8. No. 8 2.1.2 A formulation using conditional expectations Furthermore, our growth rate g t can be time-dependent – that is why we speak of «weak» autoregression. Assumption 2.1 (weak autoregression): There are growth rates (real numbers!) gt such that for the cash flows of the unlevered firm Is this the same as definition above? Yes, by application of the rules!!

9. No. 9 2.1.2 Noise and weak autoregression We will show now 1.Noise has no expectation 2.Noise terms are uncorrelated if where noise defined by

10. No. 10 2.1.2 Proof (1) Noise terms have no expectation:

11. No. 11 2.1.2 Proof (2) Noise terms are uncorrelated:

12. No. 12 Proof (2) continued

13. No. 13 2.1.2 Implications What follows from weak autoregressive cash flows? Two important theorems: 1. There is a deterministic dividend-price ratio. 2. The cost of capital of the unlevered firm may be used as a discount rate.

14. No. 14 2.1.2 First conclusion: Gordon–Shapiro Theorem 2.2 (Williams, Gordon-Shapiro, Feltham/Ohlson): If costs of capital are deterministic and cash flows are weak autoregressive, then holds for a deterministic dividend-price ratio. (Our first multiple!)

15. No. 15 First notice that 2.1.2 Proof of Theorem 2.2

16. No. 16 2.1.2 Proof of Theorem 2.2 (continued) From Theorem 1.1

17. No. 17 2.1.2 Second conclusion: discount rates Aim: to show that cost of capital are discount rates! First let us precisely define discount rates. Notice that discount rates depend on the cash flow we want to value, on the point in time where we determine this value (index t), and on the actual time period (index r) we are discounting. We use the notationfor discounting from r+1 to r.

18. No. 18 2.1.2 Definition of discount rates Definition 2.2 (discount rates): Real numbers are called discount rates of the cash flow if they satisfy Interpretation of lhs: value, follows from fundamental theorem. Interpretation of rhs: the way we use discount rates.

19. No. 19 2.1.2 Costs of capital and discount rates Now, finally, our second implication from weak autocorrelated cash flows. Theorem 2.3 (equivalence of valuation concepts): If costs of capital are deterministic and cash flows are weak autoregressive, then or: costs of capital are discount rates!

20. No. 20 2.1.2 Meaning of Theorem 2.3 Notice that sums are equal Theorem 2.3 tells us that summands are equal as well, This result is not trivial at all!

21. No. 21 2.1.2 Proof of Theorem 2.3 We skip the proof! The shining of the proof...

22. No. 22 2.1.3 The finite example We assume that k E,u = 20%. The expectations are The value of the firm is given by

23. No. 23 2.1.3 Determining Although not clear yet why, we want to determine the market value at t = 1: Hence

24. No. 24 2.1.3 Determining Q Let r f = 10%. Another additional result is the determination of the risk-neutral probability Q. Due to Theorem 2.3 («costs of capital are discount factors») we have Assume that state occurred at time t = 2. Then this equation translates to

25. No. 25 2.1.3 Determining Q (continued) Also, the conditional Q-probabilities add to one This is a 2×2–system that can be solved!

26. No. 26 2.1.3 Q in the finite example

27. No. 27 2.1.3 The infinite case Let k E,u = 20%, then the value of the unlevered firm is Let r f = 10%. Then, as above Q can be determined: regardless of t and w. Remark: The factors u and d cannot be chosen arbitrarily in the infinite case if the cost of capital k E,u is given, because must hold in order to ensure positive Q-probabilities.

28. No. 28 Summary We will consider unlevered and levered firms, Cash flows of the unlevered firm are weak autoregressive, i.e. noise terms are uncorrelated. The costs of capital of unlevered firms are discount rates. The multiple «dividend-price ratio» is deterministic.