General Comments / Chat, page-12361

Whilst things are a little quiet...

I still find it puzzling that Triangle applied the 50% haircut but didn't explain the reasoning for it.

Particularly when you consider that discount had an impact that can be measured in billions of dollars.

It leaves readers to speculate on the reasons for applying such a significant discount; some here even implying that the reason for the haircut was simply because 'the numbers were too big'.

Of course, the numbers are very big, but that's not the reason for the 50% discount.

The practice of pharmaceutical commercial forecasting is complex; actuals rarely match forecasts, and when there is a mismatch it's even rarer still to find scenarios where actuals are greater than the forecast.

It's usually the other way around.

Simply put, the forecasted adoption rates stated by physicians are typically over-optimistic.

So what to do about it?

The same group that produced the Triangle report for bisantrene also published a report 'Distilling Insights from Primary Research' where they comment:

'Everyone who has experience doing physician market research will be quick to point out that one must apply a standard research “haircut” to account for bias. Take 50% off of physician-reported responses, they say.Or potentially chop based on a top two box score. Either way, it’s acknowledged common practice, everyone does it.'

That paper was published 6 years before the Triangle report for bisantrene. (It goes on to discuss how there might be more nuanced approaches in certain settings, but evidently Triangle did apply a 50% haircut for the bisantrene analysis.)

Far from being an outlier to make the numbers seem less outlandish for bisantrene, pharmaceutical commercial forecasters have been applying discounts to physician-stated adoption rates for a long time before the Triangle report for bisantrene.

The 50% haircut was not applied after Triangle had derived an expected adoption rate; it was applied to the physician-stated adoption rate in order to derive the Triangle-forecasted adoption rate.

My criticism is not about the application of the discount (in fact, I'm glad they did), but I'm surprised they didn't explain the rationale (maybe they don't consider a few $B between friends to be of any significance...).

Incidentally, Triangle published another paper on this topic 'Pharmaceutical commercial forecasting: Case Studies on the Importance of Informed Assumptions', which only compounds my puzzlement at them somehow failing to recognise the importance of informing us of their own assumptions in the bisantrene report.

Regardless, much more important for me as an investor is that I don't want unrealistic/unfounded optimism. I want realism. Unicorns and rainbows are not for me.

So I ask myself a number of questions:

#1 Do I think the 50% haircut is a reasonable multiplier?.
=> Maybe. Maybe not. This analysis is produced by a group that we consider the experts in this line of work and stake their reputation on it.

#2 If we consider (worst case) that the 50% haircut to be valid, do the numbers still stack up for investment analysis?
=> Are you serious? Check my sentiment in my signature.

I worry much more about risks associated with program execution than a 50% haircut on physician-stated adoption rates.