Don,
It's not desirable to compare one unique company with another? Of course it is! It's not possible? Rubbish, done all the time!
Show me a person who is not investing in all companies by choice and I'll show you a person who is comparing unique companies with alternatives. Somehow. It's a matter of basis. One possible basis is on the degree to which the business model of the company is enriching to shareholders. Ask Mq! There are others. Ask Pezz. How about you? How do you go about choosing where to put your money? If you make a choice (or entrust others to make such a choice for you), are you (or they) not doing something both possible and desirable?
Is it "accuracy" you think is the undesirable and impossible element? Any method of choice where accuracy doesn't matter at all resembles a game of chance. So much so it might as well be called "gambling". Again, ask Pezz. Do you suggest we use the most inaccurate methods possible on which to base investment comparisons? Feel free.
Whether you agree or not, it is both possible and desirable to compare one unique company with another, to some arbitrary degree of accuracy greater than complete randomness. Perfect accuracy is unnecessary. We only need to be accurate enough.
But you wrote: It seems to me that most of your erroneous conclusions stem from an unwarranted premise that it is either possible or desirable to accurately compare one unique company with another.
Yes, my conclusions do stem from the premise that it is desirable and possible to compare one unique company with another, to a degree of accuracy that is meaningful, though imperfect. Contrary to your opinion, such a premise is indeed warranted. The logical errors here are not mine.
Your mistake is confusing perfect accuracy with diminished inaccuracy. With your examples of showers and ski-lift tickets, you are quite rightly pointing out the irrelevance of factoring in irrelevant costs. Of course, this does not speak against the relevance of factoring in relevant costs of the same nature!
What about showers as a business cost? Forget about 'em if you like. Doesn't mean the cost is nil, just that such an improvement in signal is lost in the noise.
But what if the unaccounted cost isn't insignificant? What if, for example, to attract employees, the shareholders of your New York company are sheltering employees in their own homes and feeding them meals and driving them to and from work and paying a third of their living expenses to boot? That would also be a competitive advantage for the company, an "intangible" benefit of a similar nature as well kept shower stalls. As a point of principle it belongs in the same place on the books too. Whether we choose to ignore it or not? Hmmm... If due to some favorable quirk of accounting rules, this practice became common place for a third of companies, (or most of the NASD companies versus most of those on the NYSE), then when anybody crowed about the relative profitability of this third of companies (or those on the NASD) versus the other two thirds (or those on the NYSE), then yes, when contemplating becoming a shareholder I might even be inclined to bring up the difference.
You suggest we should not? That's your choice. I'm OK that you disagree, but don't try to chock that difference up to logic errors on my part!
Back to the main thrust of the discussion.
Where do we agree?
You suggest we shouldn't account for insignificant costs such as free ski-lift tickets but admit as we might want to account for free automobiles. We agree. The basis for an intelligent determination whether to bother including or excluding a cost in the accounting is not the qualitative existence of a cost, but the quantitative degree to which the cost is relevant. Measured by relative magnitude to the business.
And if we do count for the cost, you suggested we should use opportunity cost as the basis, e.g. for free seats on an airline, or hand-made cars. Here we agree as well.
Now, it's not a long trip down logic lane from where we agree to arrive at the conclusion against which you have been arguing.
If something is significant relative to the other costs of the business, then we should factor it in based on opportunity cost, if not we shouldn't. When stock compensation becomes significant, we should factor it in based on opportunity cost. And the opportunity cost for a stock grant, for example, is exactly what the company could get if it sold it on the open market. Right?
When these things are infrequent and immaterial to the business except to the third significant figure to the right of the decimal point, I'd be all for losing such a faint signal amongst the noise. But when they are material to an objective discussion and rank right up there in magnitude with other parts of the business, and in the first significant figure, then we include them. If we're being intellectually honest and discussing something from a point of principle rather than a point of position.
Right?
Now, if an individual takes an objective look at the degree to which stock option compensation for many companies stands today in contrast to cash compensation, it's hard to argue it away as "insignificant".
Which leaves the only obstacle to agreement being the effect of some mysterious double counting that might end up going on if we adjusted our accounting policies to include a significant but unaccounted for cost. To which you have referred rather continuously, except by way of example. Does this double counting really exist?
Or, even if it does, turnaround is fair play. Twisting the logical hose around in your general direction, is it even relevant?
Let's use your 10 M$, 10 shareholder company, one of whom is an employee (for whatever reason). With our $100,000 per year employee, you showed two equivalent alternatives to contemplate: one where the employee is paid a salary, the other where the employee is paid in shares. And we both agree that these are equivalent positions.
If according to Don Lloyd rules we use the tool of P&L to compare these alternatives, two perfectly equivalent alternatives will appear dissimilar. If our comparison tool uses EPS as the measure of goodness, then one alternative (employee paid in stock, EPS=0) will appear far superior to the other (employee paid in cash, EPS <0).
However, both would be equivalent if the stock grant to the employee was factored into the P&L as a "ficticious" expense! Both would show EPS <0 and precisely by the same amount.
Furthermore, if one compares on the basis of the cash flow statement, both alternatives would be equivalent if the stock grant to the employee was factored in as a "ficticious" expense. Work it out for yourself if you disagree but you should see cash flow of -1 M$/year if you do it right. So no harm is done there.
If one compares on the basis of shareholder equity (balance sheet), both alternatives would be equivalent and show equally declining balance of 1 M$ per year. Again, no harm.
The accounting doesn't change the results, or double count anything. Such a change to accounting just makes equivalents more obviously equivalent in the places where the vast majority of people are trained to do their shopping for things which are inequivalent!!!
For whatever that's worth.
And whether Don Lloyd agrees with the practice of using P&L and EPS and PE ratios and such as a basis of comparison between companies is... well, irrelevant. The fact is that the bottom line of the P&L statement IS the most common, most referenced and most followed comparison tool out there. Which, by the way, is by design, rather than by accident.
Maybe indeed there is some other obscure measure which gets itself broken in the process of improving the accuracy of the most used tools. But I doubt it. Any tool that relies on "E" (which is the only thing changed here) is already broken by the degree to which E doesn't reflect the actual costs!!! And any tool that doesn't rely on E won't be changed by alterations to E!!
Try it out on your example. Your "tool that already works" should give equivalent results when applied to the equivalent alternatives you present, and should be broken in some obviously doubled counting when stock grants are reflected as an expense.
That is, if the double counting you assert exists is really there. Which I doubt. In fact, I bet if you try to show something the counting of which would be doubled, I'll show you a mistake. Possibly one which exists under current accounting too!
Suggesting we don't improve something used by some people 'cause some other people might make mistakes is a good recipe to stop progress in its tracks.
Suggesting that we don't improve something 'cause we can't obtain perfect accuracy is even worse. In our very real and very imperfect world, all we need to move the yardsticks forward is reduced imperfection. Perfect isn't the standard of good enough.
John |
