“I was once asked to assist with the valuation of a plantation operator. The expected annual cash flow was R50m p.a. before tax (assume a 30% tax rate for this purpose) ceasing after 15 years. A pre-tax fair rate of return was considered to be 30% p.a. before tax or 21% p.a. after tax. ”

Fullstop: On His High Horse: The Tax Element

I was once asked to assist with the valuation of a plantation operator. The expected annual cash flow was R50m p.a. before tax (assume a 30% tax rate for this purpose) ceasing after 15 years. A pre-tax fair rate of return was considered to be 30% p.a. before tax or 21% p.a. after tax.

There was confusion as to whether the valuation should be arrived at by discounting the:

1. pre-tax cash flow at a pre-tax rate, i.e. R163m or
2. post-tax cash flow at a pre-tax rate, i.e. R114m or
3. pre-tax cash flow at a post-tax rate, i.e. R224m or
4. post-tax cash flow at a post-tax rate, i.e. R157m or
5. pre-tax cash flow at a post-tax rate and deducting tax, i.e. R224m – R67m = R157m or
6. post-tax cash flow at a post-tax rate and grossing it up, i.e. R157m divided by 0,7 = R224m

The principle in valuations is that one should discount after tax cash flow at an after tax rate. Unfortunately, half of the auditing firms in RSA and most Merchant Banks do not subscribe to this principle. The correct answer to the problem above is to recognise the biological asset in the statement of financial position at R224m and to recognise a liability for deferred tax of R67m. The discounted value of the cash flows is, therefore, R157m, the net amount.

Despite the confusion, Mr. Market fully endorses the correct principle. Here are two examples to illustrate. In the interests of simplicity, I have ignored the fact that the cash flow is receivable bi-annually and have ignored accrued dividends at the date of the valuation. Both have little impact on the valuations.

Example 1: Standard Bank has in issue R1 irredeemable fixed 6,5% preference shares. The formula for valuing such an annuity is A/r where A is the annuity (6,5 cents) and r is the discount rate. The discount rate is traditionally arrived at by adding to a government bond rate (presently 9% p.a. for long dated bonds) a premium for risk. Consensus is that a premium of 1,5% should be added to the risk free rate to arrive at the discount rate. The school that disagrees with the above principle would take 9% plus 1,5% = 10,5% and arrive at a value of 62 cents for the preference share. The school that accepts the basic principle above would take 9% less 40% tax rate (marginal rate for individuals) and add 1,5% = 7,4% p.a. and arrive at an answer of 94 cents per share. The market price was 95 cents at the date of writing.

Example 2: Standard Bank has in issue R100 irredeemable variable rate preference shares paying a dividend of 70% of prime, i.e. R100 x 70% of 9% = R6,30 p.a. Using the government bond rate above and a risk premium of 1%, the school that disagrees with the above principle would arrive at a valuation of R6,30/(9% + 1%) = R63 per share whereas the school that agrees with the above principle would arrive at a valuation of R6,30/(60% x 9% + 1%) = R98 per share. The market price was R100 at the date of writing.

When I ask those of the other school to tell me why they ignore tax I get a variety of answers, some of which are any of the following:

1. We have never taken tax into account in the past (the rut is so deep that they cannot see out of it)
2. The capital asset pricing model ignores tax (this model was developed for Utopia, not for the real world)
3. We were told at varsity to ignore tax (tax was a separate department at varsity)
4. We adjust the risk premium by the fact that we ignore tax on the risk free rate (try this on the above examples – it does not work)

During a presentation at a large bank I got into a heated discussion with a bright young banker. He flatly refused to consider deducting tax from the interest rate when valuing a dividend-paying share. So I asked him the following question: “You have a wealthy grandmother who is considering an investment in a bond vs a preference share. They both have the same risk profiles and redemption periods. The bond is paying 9% and the preference share 7,5%. Which of the two would you recommend to her?”. Without hesitation he said: “Clearly the preference share.” When I pointed out that the bond was paying a higher return than the preference share he said: “When comparing the two one must take tax off the bond interest rate.” Enough said.

The views expressed in this article are those of the author and not SAICA. asa

Author: Charles Hattingh CA(SA), Chartered Financial Analyst, is the Managing Member of P C Finance Research cc.