Biomarin’s market cap lost $7.5bn on yesterday’s US rejection of its haemophilia A gene therapy project valoctocogene roxaparvovec (valrox). The number is huge, coming close to what some of the most bullish analysts say the entire asset is worth, and being higher than a calculation of valrox’s entire NPV based on consensus sales forecasts.
An obvious reading of this is that the market has entirely written off valrox. True, the buyside’s own assumptions were likely even more overblown, but it serves as a cautionary tale; enthusiasts of competing gene therapies from Roche and Pfizer/Sangamo should think carefully before celebrating their win.
The problem of valrox's waning efficacy has been rumbling for some time (Biomarin hopes long-term valrox data can stem the bleeding, May 22, 2019). Gene therapy players are making a case for premium pricing – $2-3m per valrox dose has been mooted – logic that would be destroyed if such a therapy were not, after all, a “once and done” procedure.
As such Biomarin’s claim that the US FDA had moved the goalposts in demanding long-term data before approving valrox, aka Roctavian, a central point of yesterday’s complete response letter, rings hollow. The question for competitors is whether they too must now brace themselves for similar regulatory demands.
Investors would apparently rather not think about such worries. Sangamo, whose haemophilia A gene therapy SB-525 is partnered with Pfizer, closed up 3% yesterday. Biomarin’s other key rival here is Roche’s Spark-derived RG6357, while Uniqure’s AMT-180 was recently discontinued for competitive reasons.
True, none of the competitors has suffered a waning of efficacy similar to that of valrox, but then none has reported as comprehensive a dataset as far out as Biomarin. Since it is not clear whether a patient can even be re-dosed with a gene therapy, never mind its cost, smart investors should think carefully about the risk.
House of cards
It is also worth investigating what expectations had been priced into a potential gene therapy like valrox; it seems that many analysts’ models were little more than a house of cards.
Take valrox’s consensus sales estimates. Before yesterday’s setback EvaluatePharma’s consensus of six banks that had provided recent forecasts saw valrox revenues peaking at $1.7bn in 2031; plugging these numbers into a cashflow model and discounting at 9% cost of capital yields an NPV of $3.1bn – a number wildly out of kilter with yesterday’s $7.5bn market cap decline.
How high would assumptions have to have been to justify such a loss of value? One of the most bullish banks to have put out a peak sales estimate, Leerink, had at one point suggested that valrox would sell $4.5bn in 2030, but later reined this in to $3.5bn.
As shown in the model below, generated by Evaluate Vantage to replicate such a scenario, even a peak sales estimate of $3.5bn yields an NPV of only $5.4bn.
Yesterday Leerink cut its valuation for Biomarin from $25bn to $20.5bn, saying that without valrox the group was worth $16bn. The implication is that valrox is worth somewhere between $4.5bn and $9bn.
It is hard to get revenue forecasts to back this up. Yesterday investors said either that valrox was a zero, or that their expectations of it had been even higher than those of some of the most bullish sellside analysts.
|The valrox bull case before yesterday's CRL|
|Gross profit ($m)||22||259||704||1,226||1,946||2,802|
|Free cash flow ($m)||(97)||(36)||328||614||1,084||1,675|
|NPV at 9% WACC ($m)||5,392|
|Note: sales decline steadily beyond 2030, and end in 2040. Source: Evaluate Vantage assumptions.|