What range would reflect fair value for Wind's stock?
=====================================================
Stock price is influenced today more by such things as market and sector trends, momentum, ignorance of new technology, and herd mentality among analysts, then the intrinsic value of a stock. Having a feel for the intrinsic value is important though, especially at times when the intrinsic and market value appear to diverge widely. If the assumptions behind the computation of the intrinsic value are borne out, then the investor who invests by following intrinsic value is likely to win (and sometimes win big).
Theoretically, the price of a share should equal the present value of the future cash flows that this share generates. Cash flows are harder to predict than earnings (hard enough as earnings are), and the theoretical price is often computed as the present value of the future earnings that one share generates. To calculate this, the three most important parameters are:
1. number of years to project
2. earnings growth rate
3. interest rate for discounting
Number of years to project
-------------------------------
Classical formulas often imply perpetuity. I do not agree with this – companies do not last forever, especially in the field of technology. Besides, even if they last a very long time, who can predict earnings, say 20 years away? I often use 15 years. By not using more years, one is adding a reasonable dose of conservatism.
Earnings growth rate
-------------------------
This is the trickiest. We can get some idea from analysts, pundits, company statements and market trends for the next five years but not beyond. I therefore like to look at the first five years separately from the years thereafter.
Interest rate
--------------
One should use a long-term treasury interest rate – say 6%. Some would argue that a higher interest rate should be used to reflect the greater uncertainty with corporate profits than with the Government's ability to repay debt. I would argue that any such uncertainty should be factored into the projected growth rates (i.e. use risk-adjusted earnings projections).
------------------------------
First, let us use the above methodology to blow away immediately the myth of a much used measure – the PEG. Many, including market analysts, believe that the PE of a stock should be very close to the earnings growth rate. There is no basis to this theory (someone please educate me if there is). It has never been clear to me in the first place if the earnings growth rate is the current year's or the long-term. Neither makes any sense. If it were the former, clearly the long-term growth rate should be more relevant to the price of a stock then just one year's growth (although important).
Let us use our model to show how off-base the PEG theory is. The P/Es (forward P/Es) for various long-term growth rates are as follows. Results are shown for two interest rates.
l-t growth - interest=6% - interest= 10%
---------- ----------- -------------
0% - 9.7 - 7.6
5% - 13 - 10
10% - 19 - 14
20% - 39 - 27
30% - 85 - 56
The PE ratio should always be higher than the long-term growth rate (unless the company is losing money), and much higher in the case of high growth companies.
Wind's price using the model
---------------------------
For projecting Wind's future earnings, I have used the following assumptions:
1. 1999 EPS of 0.77 (consensus estimate – likely to be low).
2. Growth rate for first five years: 40% per year. This is the consensus estimate and does not include significant I2O revenue. Nor does it include some of the possibilities we have been talking about here such as internet appliances, DSP, growth in the Auto sector, etc. (all of which are real possibilities and could dramatically push up revenues in the intermediate/long term). Growth rate after 5 years: 20% per year (which is still extremely good as a sustainable long-term growth rate). This is pure speculation. Any thoughts/discussion on this would be appreciated.
3. Price = discounted value of future earnings for 15 years, using an interest rate = 6%, as discussed above.
The above assumptions produce a price of $61, growing to $160 in five years' time (21% per year increase). This assumes continued growth for another 5 years at 20% per year. Now, if you were reasonably certain that the price will be close to $160 after five years, would you not borrow money at 10% to buy more shares? The break-even price for one not to do so ($160 discounted at 10% for five years) is $100. One can argue that this “adjusted”price of a Wind share should be $100. The target price range can then be thought to be in the range $61-$100.
Sensitivity Tests
-----------------
1. Now, what if interest rates all of a sudden go up to 10%? The target price range then becomes $41-$69.
2. If interest rates stay at 6%, but the long-term growth rate after five years is 10% per year, not 20%, the target price range becomes $42-$47.
3. If interest rates stay at 6%, but the long-term growth rate after five years is 30% per year, not 20%, the target price range becomes $93-$218.
Growth Implications of Current Price
------------------------------------
Using our model, let us work backwards and see what kind of growth rate is implied in the current price ($19). There are, of course, an infinite combinations of growth rates, but here are the most interesting ones:
1. 40% for 5 years, -18% therafter (!)
2. 20% for 5 years, 7% thereafter
3. 14% forever.
These show the relative importance of the growth rates during the short/intermediate term.
============================================================
I will leave it to each of you to form your own judgement as to whether Wind's current share price is justified, based on the above model. I would also welcome any thoughts on the methodology or assumptions.
Regards,
Erwin
|
