The six-year arc of the cost of credit, and what it teaches

The Deloitte CFO Survey’s cost-of-credit net balance hit −73% in Q4 2021, the cheapest reading in the series, then climbed to +86% by Q2 2023, a level not seen since the GFC, before easing to the +30–45% range through 2025 and into 2026. A 159-point swing in six years. That swing is the structuring lesson for every facility a lower-mid-market company signs from here.

The signal is the speed and magnitude of the move, not the level at any single quarter. The cost-of-credit net balance, the gap between the share of CFOs who find credit expensive and those who find it cheap, swung roughly 160 points from floor to peak in less than eighteen months between late 2021 and mid-2023. No-one who borrowed in 2021 planned for that. None of the companies that absorbed it without structural pain had predicted the rate cycle. They had built optionality into the terms at signing.

What did the arc actually look like?

The descent into cheap was gradual. The Bank Rate was cut to an emergency 0.10% in March 2020, and the cost-of-credit perception fell in step through 2020 and 2021, reaching its most negative in Q4 2021 at a net balance of −73%: a large majority of UK finance chiefs rated new credit as cheap rather than costly. A borrower in the market that winter had the full weight of the era behind their term sheet.

The ascent was anything but gradual. By Q3 2022 the net balance had crossed from cheap to expensive, a move of 120 points in nine months, and by Q2 2023 it sat at +86%, a level last seen in the GFC era. The chart below captures that rising-and-easing phase in full. Borrowers who had locked facilities on fixed or semi-fixed terms in 2021 or early 2022 with no refinancing rights watched that move from the sidelines; they could not act. Those who had negotiated flexibility (soft prepayment terms, contractual refinancing windows, or margin ratchets tied to leverage) had at least the option to respond.

What was the hard rate doing underneath?

The perception chart tracks what CFOs felt; the Bank Rate chart shows what was actually happening to the price anchor. The two are closely related but not identical. Perception led the rate on the way up; the CFO survey was registering expensive credit before the Bank had finished tightening. It has lagged on the way down. As of early 2026, Bank Rate sits at 3.75%, down from the 5.25% peak, but the cost-of-credit net balance remains positive at +43%: still expensive by any reading of the 2020–21 window, even if substantially below the 2023 peak.

What should a borrower do with this?

The cycle does not repeat on a predictable schedule. A borrower who tries to time the rate, waiting for the optimal level before signing, is making a forecast that the last six years have shown is very hard to get right. The direction reversed twice within the period; participants who were certain of each continuation were wrong at both turns.

The structuring lesson sits in the contractual terms of the facility rather than the timing of it. A facility signed in 2021 at the bottom of the cycle with free prepayment and soft refinancing rights captured the window perfectly, not because the borrower called the rate but because the terms preserved the option to act. Conversely, a facility with hard break costs, long lock-up periods or no ratchet mechanism left the borrower holding the rate for as long as the lender had negotiated into the document.

Today’s rate is above the zero-floor era and below the 2023 peak: a mid-cycle position that is neither obviously the right moment to lock long nor the right moment to wait. That ambiguity is structurally permanent: it will look the same at most points in the cycle. Rather than trying to resolve it, structure around it. Negotiate prepayment flexibility. Ask what the break cost is, and get it written clearly. Build in a margin ratchet if leverage is expected to improve. Keep the term short enough that a refinancing window opens before conditions become pressing. These are contract terms, not rate forecasts, and they are available in every negotiation.

The forward read

For a lower-mid-market company refinancing into the 2026 maturity wall, the starting position is better than the sentiment suggests. Bank Rate has come off its peak; the supply of private credit is near record levels; and lenders are competing for good credits. The cost-of-credit net balance at +43% reflects a panel of large-cap CFOs whose comparison point is 2021. Against the longer arc, mid-2026 looks like a normal market re-established after an abnormal era of near-zero rates, not an expensive one.

The task is to enter the market with terms that preserve optionality rather than to guess its timing. The same borrower who builds prepayment rights and a refinancing window into a facility signed in mid-2026 will be structurally better placed whether rates fall further, hold here, or move in the other direction again. The arc of the last six years is the argument for doing that work in the document, not the forecast.