If A "Full-On" Centocor Win Is Worth 4% to 7% to JNJ's EPS. . .

I've set out to be considerably more precise -- about what such an outcome would mean to Merck, then.

I think the Wells Fargo analysts have it about right -- a 100 percent, straight-ahead win for J&J/Centocor in the arbitration would be worth a four- to seven-percent boost in J&J's earnings per share in 2011. More on that -- in a tic -- below.

Meanwhile, though -- Pete Loftus, writing for The Wall Street Journal, this afternoon, quotes J.P. Morgan analyst Chris Schott to suggest that "Merck's partnership with J&J is worth about $2.50 per share for Merck. . . ." I am decidedly dubious about that -- it seems too low. By a lot, given the Wells Fargo analysis.

Before we proceed to hash through the numbers, however -- let me offer a suggestion as to motive (yep, JP Morgan co-led the $4 billion stock and converts offerings old Schering-Plough made in mid-2008 -- ones that remain underwater, if only slighly, today). Yes, over at JP Morgan, there would be motive to calculate a low downside figure, to New Merck. Moreover (in digesting these coming links), regular readers would recall that going all the way back to 2008, JP Morgan has been very Merck-friendly (an updated January 2010 PDF file, that). So that JP Morgan-generated $2.50 is suspect, as a "sissy Mary's analysis" -- at least in my book.

So -- what is the "real" number (in my opinion)?

Well, if Wells Fargo has a win for J&J generating between four and seven percent at the EPS line -- then, on the $4.40 of reported 2009 J&J EPS, the 4 percent would add $0.17, and the 7 percent would add $0.31 per share. At year end 2009, J&J had 2.789 billion fully-diluted common share equivalents outstanding, so the increase in net earnings, after taxes (to be carried to the J&J EPS line), would be between $474 million and $865 million (2.789 billion shares, times either $0.17 or $0.31, per share).

Alright. It is safe to assume that -- because J&J doesn't have a non-US arthritis sales force, trained and ready to go -- it would have to build the salesforce. But those expenses, are plainly already in the earnings figures that Wells Fargo estimates. They are net earnings, not gross sales.

Said another way, the $474 million to $865 million of earnings are net of sales costs (and all other costs). Those figures would also be net of very similar existing costs at New Merck. Now, it may be true that (if it were to lose the two products) Merck would be able to save some money -- if it could fire a purely dedicated Euro-zone, and/or Pan-Asian, arthritis sales-force. On the contrary, it is highly likely that this salesforce is "shared" at Merck. That is, these people probably also sell other biotech drugs, from legacy Schering-Plough, primarily (and to a lesser extent, from legacy Merck). So, Merck may shave off some variable expenses (but not a ton), should it lose the J&J/Centocor arbitration.

Thus, assuming that the Wells Fargo analysts did not presume that J&J could create a much lower-cost salesforce, outside the US, than the one Merck presently deploys (and why would they?), then it is fair to assume that Merck would lose between roughly $450 million and $850 million of earnings (not sales), if the arbitrators ruled against Merck's reverse merger ploy. That is an annual figure, and it excludes any additional damages Merck would pay to J&J, for the lost profits due J&J for sales from Novemeber 3, 2009 to early 2011 -- or about an additional $700 million.

Okay, then -- on Merck's net earnings line, we may assume that between $1.15 billion and $1.55 billion of earnings will vanish in 2011 (the $700 million in damages, paid from net earnings -- added to the between $450 and $850 million in 2011 lost earnings). Dividing by 3.132 billion Merck common share equivalents outstanding at Q2 2010, that would be between $0.37 per share and $0.49 per share of lost 2011 GAAP EPS.

In all of the full-year 2010, Merck is still projecting GAAP EPS of only $1.15 to $1.50 per share. This loss, then, would take between one third and one quarter of all of that away. That would be an absolute worst case scenario, though. On non-GAAP basis, it's considerably better.

Of course, Merck projects more like $3.27 to $3.41 of NON-GAAP EPS for 2010. That makes the loss number seem more swallow-able. However, recall that to "fill the hole", at Merck's 28 percent average gross margins (TTM), it will need 1/0.28 dollars of NEW sales -- or $3.57 of new sales, for each dollar of the missing earnings, just to "stay even".

If the missing earnings are $1.15 billion to $1.55 billion, then, Merck will need new sales of between $4.11 billion, and $5.53 billion, in 2011 alone. And that will likely be on something like $40 billion of 2011 revenue. Where on Earth could Merck possibly scare up an incremental $5 billion in sales, or an additional 12.5 percent? I have no idea.

If Merck (as the venerable house of Goldman, Sachs has opined) is fully-valued at $37, including a complete WIN in arbitration, I get a 12.5 percent NYSE linear per share price decline (on a complete Merck loss to Centocor, in the first year, alone) or a $4.63 price decline -- or, nearly double what JP Morgan has projected (at $2.50 per share).

Motto? Be careful out there. The $37 of Merck NYSE per share value (tonight) could easily be $32.27 in early 2011 -- just due to the arbitration's outcome.