Thomas:
Anything based on current numbers alone, no matter how massaged, is just too thin a slice of the pie. I think we are seeing some good examples of this currently where both price and earnings are clearly distortions of anything we could expect in the future.
Absolutely.
Anything based on analyst estimates seems to me little more than flipping a coin.
You are more charitable than I.
The closest suspects seem to me to be the FCF family, but there again I find myself wondering where I would get the numbers to plug in that were meaningful? If my growth figures are just some massage of analyst numbers and instinct, why would I think this gave me anything more than the rudest guess? I suppose if my high, medium, and low estimates *all* told me that
something was over or undervalued one might give it some credence, but how likely is that?
Any projections are inaccurate. Though it bears mention, once again, that failing to make an explicit projection is still making an implicit projection. i.e., to buy without the explicit projection is to implicitly project that results will be at minimum what your "not done" high estimates would have been. But I digress.
For companies with an operating history (e.g, silverbacks) you can certainly use historical information to frame your own projections. You can easily see how much FCF was generated per dollar of sales, and the overall rate at which sales have grown. You can use these numbers to build projections. Perhaps use a continuation as your medium or high projections for business performance going forward.
It is certainly difficult to take this approach with a company without an established operating performance history. In fact, I would say that one should not attempt to do so, as it is too easy to get ridiculous results. However you can "borrow" from the history of other companies to get a ballpark. i.e., if a company was to grow as wonderfully as did one of the silverbacks did in its early years, what would that company be worth? For that matter, you don't need to worry about specific projections. You could work out what growth rate would be needed to justify the price, and then decide if it were realistic. Alternatively, you could work it the other way. i.e, CSCO was rarely priced above 20 x revenues in its early years. So you ask yourself "what are the odds this young company will have a better long term business performance than CSCO - am I certain enough to pay 40 times revenues for it?"
The point is to not get too hung up on accuracy. You aren't going to be accurate. But maybe you can be slightly more accurate than when you implicitly project without the explicit quantitative work.
- Pirah |
