On Valuation

Last time I posted about biotech valuation (here), the model used was very “bottom-up”; that is, I attempted to put in place estimates of numbers of units sold, unit prices, market penetration, inflation adjustments, development costs, and to use this data to arrive at an estimate of cash flow, which is then discounted by the risks of clinical failure (the rNPV method) to arrive at a valuation. Another approach I have seen used is a lot more rough-and-ready and is far more “top down” - specifically, you start with an estimate (guestimate even) of peak sales and odds of success and use that to directly arrive at a valuation.

For example, I have recently been looking at some stuff Antisoma (link) have been doing in NSCLC, and thought that I would rustle up a valuation using this second method.

The image above illustrates the principles fairly clearly I think. To some extent you have to use educated guesses of the likelihood of clinical trial success. For example, taking ASA404 - the Phase II NSCLC trial data is split 1/3 - 2/3 between squamous and non-squamous NSCLC. Seeing as Novartis (who have licensed this program from ASM) have indicated that they intend to develop ASA404 in squamous cell NSCLC (Avastin is approved for non-squamous only), this means that a lot of that Phase II data is no longer really relevant to the proposed Phase III program (only 11 squamous cell patients in the 1800mg/m2 treatment arm!). I chose a 40% likelihood of success - therapeutics which have unprecedented mechanisms of action fail twice as frequently as those that don’t. In addition squamous NSCLC is generally shows shorter PFS, and is more aggressive/less survivable that non-squamous, which will undoubtedly make it a tougher clinical nut to crack. Add in the fact that ASA404 bombed in ovarian cancer, and that prostate data was looking a bit shaky……

How then do you decide on an estimate of peak sales? In this case I looked up how much revenue Avastin currently generates (approx 1.8bn USD per annum and still growing), and divided that by 5 (80% of NSCLCs are non-squamous, only 20% squamous). That gives you approximately a market of ~$360m, which converted into pounds is approximately 180m GBP. Since this product won’t be on the market (if ever) for a number of years, I inflated the peak sales to 500m GBP per annum, which will also allow for some off-label use. [What can I say, I was feeling generous!]

The multiplier used next not only allows for sales from multiple years to be taken into account, but also gives extra credit to ground-breaking therapies that will probably have that “X-Factor” - everything from elements of “coolness” and clinical excitement to the reliability of that revenue stream.

Pulling all this together you arrive at a probability adjusted value, which is used to derive the present value (discounted over the number of years to the sales peak). Bingo Bango, you have a valuation. Repeating the process for the other clinical programs, add in the cash in the bank, and it all works out to give you a current share price target.

You can download my spreadsheet from here.