Why Stock "Value" Systems Have No Value
John Price, PhD
| We regularly get asked how Conscious Investor is different from the countless "value" based software products and books available. The vast majority of "value" approaches are based upon a standard discount cash flow (DCF) model. They all purport to find what is called the intrinsic value or true value of a stock. This is the value the proponents claim that all rational investors should pay for the stock. Many of the authors and investment websites claim to base their intrinsic value approach upon the methods of Warren Buffett. In fact, it is actually very unlikely that he really uses any of these methods in anything vaguely resembling a formal approach. For example, Charlie Munger, the old friend of Buffett and the Vice-Chairman of Berkshire Hathaway, says that he has never seen Buffett do any discount value calculations. This is consistent with the answer he gave when asked about intrinsic value at the annual meeting of Berkshire Hathaway in 1996. "There is no formula to figure it out." He replied. "You have to know the business." Until recently, Buffett had never used a computer for anything, let alone for implementing any value models. Now he only uses it to play bridge on-line. When people write about Buffett I usually agree with most of the general observations made about his approach. The one area where Conscious Investor differs from the majority of Buffett followers (but quite likely not with Buffett himself) is in relation to the "valuation" model used. There are large numbers of DCF models. Stable growth models, two-stage models, three stage models and so on. Each of these models calculate the intrinsic value of the stock by discounting back to present time the stream of "cash" that is generated by the business. All I can say is that it is hard to believe that such simplistic and unreliable material is still taught in universities and promoted by stock analysts. The main problem is that people think that just because they can put some numbers into a formula they have found a useful and realistic model. Here we are talking about all the academics and writers who have limited understanding of the interface between mathematics and the real world. Two Fatal Weaknesses There are two fatal weaknesses of any DCF model.The first is that DCF models are unstable — small changes in the input values can lead to such large changes in the output that almost any number can be obtained. Here is a simple example showing just how unstable intrinsic value calculations are. The model uses the standard two-stage approach. We assume that the company spends ten years in the initial stage during which the cash per share (generally the free cash flow) that it generates grows by the rate given in the second column. After the initial stage comes the steady state period. During this period the cash is assumed to grow at the rate described in the third column. Finally, everything is discounted back to present time using the rate given in the fourth column. With small changes in the input variables, the output can shift from $23 to $58. I can't help getting the image in my mind of the host of an old television show saying, "Would the real intrinsic value please stand up?"
And then you have to do the same estimate for the discount rate. If a model with this level of instability was proposed in a science class, it would be thrown out of the window. The second fatal weakness is that just because some model says it is generating something called intrinsic value does not mean that it is providing anything that really is "intrinsic value". And it certainly does not mean that it is giving something useful for investors. For example, just because a stock is undervalued (by some model or other) does not mean that it won't stay undervalued. This is quite different from saying that if a company has a strong economic performance, then eventually the market will acknowledge this by increased stock prices. Another weakness The instability described above is compounded by the fact that it is impossible to confirm the accuracy of two of the input variables. For example, the entry for the final growth rate requires that you estimate the growth rate of the cash not for another ten years, or even twenty years, but out to infinity! This is despite large studies showing that analyst forecasts for earnings over five years are no better than random.In contrast, in Conscious Investor we don't try to calculate the mythical concept of intrinsic value. We don't talk about whether a stock is undervalued or overvalued, whatever that may mean. Rather we define value in terms of the return you will get on an investment. Instead of intrinsic value, we talk about investment value or investment return. This is calculated using the proprietary tools STRETTM and STRETD®. STRET is a calculation of the annualized percentage profit or rate of return from owning the stock. STRETD is similar except that it assumes that dividends are reinvested. By calculating the actual return you can anticipate on your purchases, you get practical criteria whether it is worthwhile buying stock in a particular company or not. So in a nutshell, if you were to compare the stocks selected by Conscious Investor with other "value" models freely available on investor websites, you are unlikely to find much overlap between the selections. Despite the best intentions of would-be intrinsic value systems, the way that DCF models work either provide you with more or less random stocks or with stocks that you like and you (unconsciously) manipulate the data to make them appear undervalued. The Conscious Investor for more details ... The book The Conscious Investor: Profiting from the Timeless Value Approach by Dr John Price covers over 30 valuation methods including their assumptions, and their strengths and weaknesses. |
http://www.conscious-investor.com/articles/articles/ar04value.asp
No comments:
Post a Comment