The Discount Rate Is a Lie: Risk Isn’t a Constant in Biotech

There is a quiet confidence with which financial analysts deploy the discount rate—that tidy little variable in a discounted cash flow (DCF) model that purports to capture the riskiness of future earnings. In biotech, however, this confidence is almost always misplaced. To apply a neat, fixed percentage to an asset class whose chief defining feature is total unpredictability isn’t just bad modelling. It is performance art.

In a normal business—say, a widget manufacturer or a mid-market SaaS company—the idea of applying a time-based discount to future revenues makes some sense. There is a going concern. There is historical performance. There are customers, perhaps even loyal ones. And yet, in biotech, one regularly encounters models that assume the same principles can be used to evaluate a preclinical oncology asset that has never been within 200 yards of a human body.

The default approach goes something like this: take your forecasted revenues, apply a generic probability of technical success (POPTS), slap on a 10–16% discount rate depending on the mood of the partner, and voila—a net present value appears. The process has all the intellectual rigour of reading chicken entrails but is considered gospel in pitch decks, board meetings, and M&A presentations.

But the fundamental problem is this: risk in biotech is not continuous. It does not gently decay with time like the price of a bond. It is binary, asymmetric, and path-dependent. It is also not purely financial. It is regulatory, scientific, and often reputational. And it changes by the week.

Let’s consider a Phase II trial in glioblastoma. The discount rate might be set at 15% to reflect the riskiness of oncology. But the true inflection point isn’t embedded in the rate. It’s in whether the EMA views pseudoprogression as a real outcome or an imaging artefact. That decision can turn a billion-euro asset into a write-off overnight. No discount rate can capture this.

Then there’s the matter of cost of capital. Classic finance theory would have us believe this reflects the weighted average of equity and debt, adjusted for beta. But in biotech, there is often no debt. Beta is fictional. And the cost of capital is essentially: whatever the last investor round implied. Which, during certain frothy years, was somewhere between optimistic and clinically delusional.

All of this leads to a more unsettling conclusion: biotech valuation isn’t about precision. It’s about storytelling. The discount rate exists not to quantify risk, but to lend a veneer of quantitative respectability to what is fundamentally a narrative exercise.

This is not an argument for abandoning financial models. It is an argument for humility in modelling. One could take cues from the insurance world, where actuaries model tail risks with probabilistic distributions that acknowledge the fat tails and structural uncertainties of rare events. Or from venture capital, where some firms have abandoned DCFs altogether in favour of scenario analysis, sensitivity maps, and upside-capture frameworks.

What biotech needs is not another spreadsheet with three decimals. It needs the intellectual honesty to say: “We don’t know, but here are five plausible futures, and here’s what we’ll do in each one.”

In the end, the discount rate is not just a number. It’s a signal. And the signal it often sends in biotech is this: we haven’t really thought this through, but the model balances.

One might argue that it’s harmless—a tool for standardisation, a bridge for communication between investors and operators. But as with any convenient fiction, the problem arises when we start to believe it. When funders, regulators, or boards make strategic decisions based on valuations that depend more on spreadsheet formatting than biological reality, we are not just deluding ourselves. We are misallocating capital.

So what to do? Stop pretending biotech can be valued like an airport or a supermarket chain. Embrace the fact that it is, at best, a series of options strung together by milestones and subject to externalities no model can capture. And above all, stop revering the discount rate like it descended from Sinai.

Because in biotech, risk is not a line on a curve. It is a trapdoor. And you won’t find that in cell B42.