Just a few thoughts that have been rattling around in my head about last week's action...
It seems to me that the discounted earnings model method is no way to value a company like Sepracor. You make an assumption about what drugs will make it to market, another assumption about what sales will be, another about the type of sales arrangement made (whether outlicensed or direct), margins, R&D, SG&A, and then you get your profit figure. If Sepracor were a typical biotech with one or two horses, this might make some sense, but since we have dozens of prospects, many of which are dark horses (eg R-fluoxetine--the initial publicity was about the S-isomer, as a migrane prophylactic), and some of which will die unexpectedly, (eg Bayer dropping SEPR's ketoprofen), or hit speedbumps (eg the extension of Allegra patents and still unresolved patent interference issues), any guess what Sepracor's revenue mix will be in 2003 is subject to change, to say the least.
And even in considering the fortunes of one particular new drug, the analysts have a bad track record. Hardly anyone saw the $1 billion potential of Amgen's two blockbusters (recalling an old Barron's article from memory). More recently, Biogen's Avonex has proven hard to forecast: last year BGEN was in the low 30s as the market worried Avonex growth had peaked, now the stock has doubled as enthusiasm for Avonex waxes. In both these cases, the time frame was much more short-term, as the drugs were on the market or nearly so, yet estimates proved wildly inaccurate. Granted, in both cases, the indications were new and so the market had to be created, but some of the Sepracor ICEs will also address new markets that cannnot now be so clearly seen (eg R-oxybutynin).
But, hey, it's your job to guess, so you guess, and say, "we see Sepracor earning $1.15 in 2000 increasing to $3.80 in 2001" (a pretty spectacular growth rate, by the way, which if it comes off will drive the stock price nutty). So what's that $1.15 worth in 2000? Or $3.80 in 2001? Now come all the market assumptions. First you have to apply a PE, which now is thought to be reasonable about 30. In '93, of course, Merck was trading at 15x, and 20x is probably its average PE over the past 15 years. There is market PE set by interest rates, industry PE set by growth prospects, company PE set by particular factors (all very roughly of course), and all these things will be different in 2003. But let's say 30, since that's what it is now and at least that gives a fair basis for comparison. Now what's your discount rate to get a present value? This, as someone here pointed out, should be lower for Sepracor than for say Xoma, since there is lower risk (many products, many already on market in different forms).
The result of the mental gymnastics is this:
>>Maris pegs the stock's fair value at $52 to $57, substantially lower than its price today.<<
But also this:
>>Douglas Lind, an analyst with Morgan Stanley in NY, rates Sepracor a "strong buy" and sets a $73 one-year price target. He calls his assumptions quite conservative and says he could make an argument for $120 by tweaking growth rates.<<
So what's SEPR worth? Say, somewhere between $52 - $120. Thanks, guys.
Granted, the analysts need to say something, and there does need to be some basis for comparison between companies with no earnings. But to issue buy and sell recommendations on the basis of a specific price derived from one of these cranky models is nonsense.
Just my opinion.
Sorry for the novel.
Bruce
|
