You Cannot Price an Interest Rate Without a Loan Term

A colleague told me a story some years ago about a project he was onboarded to carrying significant debt, upwards of $10M, due to a full reset and a loose trajectory. He put together a proposal and thought it was enough to bring to the VC panel.

His boss smiled and asked if they had any questions. The first was about an NDA, with some laughs. The next was a comment that the information was appreciated, and if it became a concern they would reach out. The negotiation was around $300M give or take. His number was a give or take. And if it became a concern, they would drop by $10M and charge back $14M in migration advisory and implementation services.

They would make a little money off the debt and everyone is happy.

That story shifted my beliefs, as has happened so many other times, and I went quiet on the topic for a while. Technical debt in those terms is just trailing earnings or asset value. The debt behaves accordingly: linear, absorbable, split with the buyer, no drama. It is real, and it is not wrong for that kind of exit.

Two decades on technical debt, first out of interest and challenge, then out of necessity, and finally giving the lessons back. Most of it spent watching the same tension play out: the moment realization meets self-preservation, at every level of the process.

I’ve ended on this. Ownership does not care about the philosophical framework. They already have a pricing model for it, in three parts: balance-sheet debt, which costs what the fix costs; risk-of-event debt, which costs what the blast radius costs; and the one that matters most, velocity and optionality debt, which costs whatever it costs against the plan.

The model is already theirs. Ours is the work of filling it with something rigorous enough to hold, and compressed enough to survive the room.

At the executive altitude

The executive concerned with navigating technical debt discussions with the board is representing velocity and optionality debt, whether or not they use those words.

This debt never appears as a liability. It appears as a growth rate the org can’t hit, because it can’t ship, can’t integrate the next acquisition, can’t respond to a competitor as fast as the growth story assumes. It isn’t on a balance sheet. It’s in whether the forward projection is credible, and forward projections are what multiples are priced against.

Which is where the panel’s arithmetic stops working. When a buyer is paying for a growth trajectory rather than a static earnings number, they aren’t pricing off current profit minus known liabilities. They’re pricing off a multiple, and the multiple is a bet on how fast the company can keep doing what it’s been doing. Market advantage enters the math not as a line item, but as the thing that sets the growth rate the multiple is bet on.

The CFO and the engineer

The engineer is the input, the CFO the output. And like your keyboard to your monitor, unending levels of abstraction between the two.

The CFO can easily accept, conceptually, that technical debt is an accrued liability. That framing appears favorable to the engineer. Principal, interest, default. Quantifiable, then. And the obvious question follows: how do we quantify it?

We ask the engineers. They can quantify it, and they actually do. Cyclomatic complexity, deprecated and transient dependencies, coverage gaps, on and on. We’ve identified the principal.

Engineering management compresses that into velocity, capacity, and burndown. That’s the interest, the carrying cost. What it costs every sprint to keep operating with the debt in place instead of having it paid down. Those metrics are measuring it even when nobody names it that way. A team moving slower than it structurally could is paying interest on principal nobody has retired.

But what happens next is that this becomes “the number.” That’s the folly, and seeing why requires understanding each stakeholder’s motivations and expected reactions. The engineers have only quantified the principal. The fix cost. Retire the debt today, independent of anything the business plans to do. It’s the only leg that doesn’t require anyone outside engineering to have an opinion. So we stay in engineering.

Which is where the first lesson lands:

You cannot price an interest rate without a loan term.

Debt carried for six months before the system retires pays effectively nothing. There’s no time for the slowdown to compound. The same debt in a system becoming more central every year carries a rate that’s effectively rising.

The asymmetry

Line the three lenses up and something falls out of them.

Balance-sheet debt prices today with no roadmap at all. It’s an inventory. Known, fixable, linear. Subtract it or escrow it, done.

Risk-of-event debt also prices today with no roadmap. It’s asymmetric, but it’s self-contained: probability and blast radius, independent of business direction. It doesn’t cost the fix price, it costs the probability-weighted price of a breach, an outage, a key person leaving with undocumented knowledge. That can tank a deal or blow up a multiple far past what the fix would have cost, which is why it’s priced like insurance rather than like a bill.

Velocity and optionality debt is the one that behaves differently. You cannot price it without knowing what the business is trying to do, because “this slows us down” is meaningless without a target speed.

So the roadmap isn’t a prerequisite for technical debt conversations in general. It’s specifically the prerequisite for the lens that’s worth the most money.

That’s why quantification frameworks that measure debt in isolation fail. They produce numbers that can’t distinguish between debt in systems that matter and debt in systems that don’t. Every allocation model reserving twenty percent for debt work assumes all debt is equally worth paying down, which is only true if the organization has no plans that would help prioritize.

None of the three legs is new. Balance-sheet debt is how due diligence already categorizes. Risk-of-event debt is what a risk register already does. What’s worth anything is that they’re assembled against one organization’s actual situation. Novelty isn’t the bar. Correctness and internal coherence is.

The finding

Organizations that struggle to quantify technical debt usually have a prior problem. The business functions don’t have coherent multi-year plans. Customer service doesn’t know its roadmap. Finance doesn’t know how its systems will evolve. Sales doesn’t know what platforms it needs.

That reads like someone else’s problem, two levels up, outside what you can fix. It isn’t, quite.

If velocity debt can’t be priced without a roadmap, the ask was never budget for remediation. The ask is what each function intends to do with the system over the next three years. Which systems become more important. Which stay stable. Which get retired.

Ask it and one of two things happens. You get the answers, and you can finally price the lens that’s worth the most money. Or you don’t, and the absence is the finding.

At that point, the engineering metric ceases to be a technical complaint and becomes a diagnostic tool for corporate strategy. The inability to price technical debt is a lagging indicator of an organization operating without a horizon. For a leadership team pricing its future on a growth multiple, the arithmetic stops working without it.