Mini-Perms and Refinancing Risk: The Step-Ups Cost 3.9%, the Wall Is 64.8%

September 17, 2026 · Dezzmond Team

Project Finance Project Finance Mechanics Refinancing

A soft mini-perm punishes a sponsor for not refinancing. The margin steps up, a cash sweep switches on, distributions stop. Those provisions get negotiated hard, and they should — they look punitive on the page.

On the worked facility below, they cost equity $4.5m, or 3.9% of the present value of its distribution stream. Meaningful, but considerably less than sponsors fear, because the penalties bite in year eight and equity discounts year eight heavily.

The number nobody negotiates is the other one. At the mini-perm date the facility still has $97.1m outstanding — 64.8% of the original amount — and it has to be refinanced in whatever market exists seven years from financial close.

The step-ups are an incentive. The wall is the risk. This post covers both, what makes a refinancing achievable, and how to stress a balance that nobody has committed to take out.

ℹ️ Note: All figures are labelled assumptions on a $150m facility at 6.5% amortising over fifteen years. Market terms vary; the relative magnitudes are the point.

What Is a Mini-Perm?

A facility whose economic life is deliberately shorter than the asset's, structured so that the debt is refinanced part way through. Lenders provide construction and early operating funding, then hand the asset to a different market — a longer bank facility, an institutional term loan, or a project bond — once it has an operating record.

“Hard” and “soft” are useful analytical labels, but filed financings use a range of legal maturities, amortization profiles, margin ratchets, sweeps and distribution conditions. The critical distinction is whether the refinancing date is the legal maturity or an economic incentive date before a later maturity.

Hard or Soft?

	Hard-style structure	Soft-style structure
Refinancing date	coincides with legal maturity / balloon	precedes legal maturity
Failure to refinance	maturity payment becomes due; non-payment can become an Event of Default	financing remains outstanding, but pricing / sweep / distribution economics may worsen
Main risk	liquidity and default risk at the wall	economic pressure plus later refinancing risk
Terms to model	balloon, maturity, extension rights, takeout fees	margin ratchets, sweeps, restricted payments, target balances

In a hard-style mini-perm, the legal maturity itself creates the refinancing requirement. Bancolombia's public filing describes a hard mini-perm as a project-finance structure with legal maturity around five or seven years and a bullet or significant balloon. A specific asset can be shorter or longer.

In a soft-style structure, the legal maturity is later and the documents create economic pressure before that date. The exact tools can include margin increases, target-balance tests, cash sweeps or restricted-payment conditions; do not assume every “soft mini-perm” uses the same bundle.

The trade-off is well understood. The hard structure gives lenders certainty that refinancing will happen at prevailing market rates and lets them price and amortise upfront fees over a short period. Its main disadvantage is the introduction of a new, possibly unnecessary, default risk for every party — funders, borrower, and in a PPP context the government.

Why Do Mini-Perms Exist?

Three reasons, and only the first is about the project.

Construction risk is priced differently from operating risk. A bank willing to take construction risk is a different institution, with different pricing, from one willing to hold a twenty-year operating asset. A mini-perm lets each do what it is good at, which genuinely lowers the blended cost.

Tenor and capital allocation. A mini-perm lets lenders limit the period for which they commit capital while leaving the project to refinance once construction risk has fallen and operating evidence exists. The exact lender motivation is institution-specific.

Pricing and balance management. Shorter legal tenor, margin ratchets and sweep mechanics can change lender returns and balance-sheet usage. Those economics are deal-specific and should not be treated as the borrower's refinancing case.

The sponsor is accepting a future market dependency in exchange for the initial financing terms. An operating record can improve the asset's financeability, but future rates, coverage requirements, remaining contracted life and lender appetite can move in either direction.

What Do the Step-Ups Actually Cost?

Less than they look, and the reason is discounting.

Assumptions, labelled as such:

Facility                                       $150,000,000
Rate                                                   6.5%
Amortisation                                       15 years
CFADS                                           $30,000,000
Asset life                                         20 years
Equity discount rate                                    12%

Soft mini-perm penalties, from year 8:
   Margin step-up                            6.5% → 8.0%
   Cash sweep                                        100%
   Distributions                                  blocked

Equity NPV, refinanced as planned              = $115,430,150
Equity NPV, penalties bite                     = $110,932,152
                                                 ------------
Cost of not refinancing                        =   $4,497,998   (3.9%)

Under the penalty case the 100% sweep repays the facility in year 11 instead of year 15 — four years early, which is worth something back to equity and partly offsets the higher margin and the lost distributions.

This is a genuinely useful result because it runs against intuition. A clause bundle that reads as severe — a 150 basis point step-up, a full cash sweep and a distribution ban — is worth under four percent of equity value, because it operates from year eight onward and a 12% discount rate reduces year eight to 40 cents in the dollar.

The practical consequence is that a sponsor spending its negotiating capital on softening the ratchet is spending it on a four percent problem. There is a larger one in the same document.

One qualification, because the result is sensitive to a single input. The 3.9% figure assumes the penalties begin in year eight. Move the mini-perm date earlier and the cost rises sharply, because less discounting applies: a five-year mini-perm with the same penalty package is worth substantially more than a seven-year one, and a three-year structure — which appears in some construction-plus-ramp financings — is worth more again.

The general relationship is the same one the cash sweep post established from the other direction. The value of any penalty or acceleration provision is dominated by when it operates, not by how severe it is. A savage clause that starts in year ten is cheaper than a mild one that starts in year three, and a sponsor comparing two term sheets should look at the dates before the percentages.

That also means the hard mini-perm and the soft one converge in cost as the date moves earlier. At seven years the soft structure's penalties are a modest, discounted cost and the hard structure's default risk is the material difference between them. At three years, the soft penalties are expensive enough that the choice is much closer than the default-risk framing suggests.

The Wall

The balance that has to be refinanced.

Outstanding at end of year 7                    = $97,133,341
As a percentage of the original facility        =        64.8%

Nearly two thirds of the original debt, falling due into a credit market seven years away. That is the actual exposure, and unlike the step-ups it is not discounted away — a failure to refinance it is a liquidity event in the year it happens, not a spread applied to later cash flows.

Stress it properly. What the market will lend against the same $30m of CFADS depends on the coverage ratio it demands, the rate, and the tenor it offers:

Refinancing market	Capacity	vs the $97.1m wall
Benign: 1.30×, 7.5%, 10 years	$158,401,868	comfortable
Tighter: 1.45×, 8.5%, 8 years	$116,672,751	adequate
Stressed: 1.60×, 9.5%, 7 years	$92,805,229	$4.3m short

Even in the stressed case the shortfall is $4.3m — real, requiring an equity injection or a partial repayment, but not existential. That is the honest conclusion on this particular set of assumptions, and it is the sort of conclusion a stress test should produce rather than a binary.

Two things move the answer sharply, and they are the ones to test.

CFADS at the refinancing date, not at close. Everything in market structure material applies: a project whose basis has widened, whose curtailment has risen or whose capacity accreditation has been cut will refinance against a smaller number than the one it was originally sized on. On these assumptions a 15% fall in CFADS moves the stressed case from a $4.3m shortfall to roughly $18m.

Contracted status at the refinancing date. A project refinancing with twelve years of PPA remaining faces a different market from one refinancing with three. That is the merchant tail analysis applied to a refinancing rather than an original financing, and it is why mini-perm dates and PPA expiry dates should never be allowed to converge.

What Makes a Refinancing Achievable?

Five things, in roughly the order lenders assess them.

An operating record. Actual generation, availability, basis and operating-cost data can reduce uncertainty relative to the original development-stage model, but it does not guarantee better refinancing terms.

Remaining contracted life. The single most important structural variable. Refinancing a project with a long contracted tail is ordinary; refinancing one about to go merchant is a different transaction.

Amortisation achieved. The wall is smaller the more the facility has amortised, which argues for a faster schedule during the mini-perm period even at a cost to distributions — the reverse of the argument the sweep post made about ongoing sweeps.

Market conditions. Uncontrollable, unforecastable, and the reason this is a risk rather than a plan.

Sponsor standing. A refinancing is a new credit decision, and a sponsor with a track record of completed refinancings and no distressed assets will find one where a stressed developer will not.

What Does the Refinancing Itself Cost?

Enough to matter, and it is routinely omitted from the model that assumes the refinancing happens.

The cost stack can include arrangement fees on the new facility, prepayment or break costs on the old one, hedge termination, and transaction costs such as legal, technical, insurance and model audit. The amount and sign of each item are document- and market-specific.

In the worked case, use an illustrative refinancing-cost assumption rather than a market rule. The takeout amount needs to cover the old debt plus whatever arrangement fees, legal / technical costs, hedge settlement, make-whole or break costs are actually triggered by the old and new documents.

The hedge termination is the item most likely to surprise. A swap struck at financial close and running to year fifteen, terminated in year seven, settles at its mark-to-market. If rates have risen the sponsor receives a payment; if they have fallen it makes one, and the amount on a $97m notional with eight years remaining can run to several million dollars in either direction.

Two practical instructions follow. Size the new facility against the wall plus transaction costs plus a contingency for the hedge position, rather than against the bare balance. And check the prepayment provisions of the original facility at financial close, not at year seven — a make-whole or a hard non-call period in the original documents can make an early refinancing uneconomic regardless of market conditions.

Is a Project Bond the Answer?

A project bond can be one takeout option, but it is not the assumed destination of every mini-perm.

A bond takeout may require a rating process, sufficient issue size, an investor-acceptable covenant package and operating / contractual evidence appropriate to the credit. The threshold for each depends on the market and transaction.

What it offers in return is tenor and pricing. Institutional investors will hold twenty-year paper that commercial banks will not, and on a rated, contracted, operating asset the pricing can be materially better than the bank market.

Bond covenants, amendment mechanics, call protection and funding timing can make the structure less flexible than bank debt. Model the actual make-whole / call schedule and proceeds timing if a bond takeout is part of the strategy.

The choice is therefore less about price than about what the asset needs next. A project that is complete, contracted and stable benefits from bond tenor and does not need bank flexibility. A project still making capital decisions — an expansion, a repowering, a co-located battery — will find the bond's rigidity expensive.

When Does a Refinancing Release Value?

This is the sponsor's actual motivation, and it is worth separating from the risk.

A successful refinancing can also create a distribution if the new debt raised exceeds the amount needed to repay the existing facility and transaction costs, subject to restricted-payment and leverage conditions. Whether operating history supports more debt depends on the new lender's coverage, tenor and revenue assumptions.

That is a legitimate and substantial source of return. It is also the reason mini-perms are sometimes preferred by sponsors rather than merely tolerated: the structure creates a scheduled opportunity to re-lever at a point where the asset is demonstrably better than it was at close.

Two constraints on it are worth knowing.

In PPP and concession structures, the gain is often shared. Refinancing gain sharing provisions require a defined proportion of the benefit to be paid to the public sector counterparty, on the reasoning that the improved terms arise partly from the concession itself. Where such a clause exists, the calculation methodology is worth reading carefully at financial close rather than at refinancing.

Lenders will resist a full re-lever. A refinancing that returns the gearing to its original level puts the new lender in the same position the old one was in, without the benefit of the amortisation that has occurred. Expect the new facility to be sized below the theoretical maximum, and expect that to be the negotiation.

How Do You Model Refinancing Risk in Excel?

As a balance at a date, tested against a range of market capacities — not as an assumed rollover.

The two outputs

PF_RefinancingWall     = Outstanding balance at mini-perm date
                       = $97,133,341   (64.8% of original)

PF_StepUpCost          = Equity NPV base − Equity NPV with penalties
                       = $4,497,998    (3.9%)

Report both. The second is what gets negotiated; the first is what determines whether the structure works.

The refinancing capacity test

Refi_Capacity = (CFADS_at_date / Target_DSCR) × Annuity(Refi_Rate, Refi_Tenor)

Shortfall     = MAX(0, Refinancing_Wall − Refi_Capacity)

Run it across a grid of target DSCR, rate and tenor rather than at a single point. The output is a surface, and what matters is how much of it produces a shortfall.

The CFADS sensitivity, which dominates

CFADS at refi date       Stressed-case capacity      Shortfall
$30.0m (as modelled)             $92,805,229           $4.3m
$28.5m (−5%)                     $88,164,968           $9.0m
$25.5m (−15%)                    $78,884,445          $18.2m

Every haircut from market structure material lands here twice — once in the original sizing and again, compounded, in the refinancing capacity.

The date that must not be allowed to converge

Mini-perm refinancing date                     year 7
PPA expiry                                     year 15
Contracted life remaining at refinancing       8 years    ← check this

ℹ️ Note: A model that assumes the mini-perm simply rolls into a new facility at the same terms has not modelled refinancing risk; it has assumed it away. The minimum honest treatment is the wall, one stressed capacity case, and the resulting equity injection.

To build the refinancing wall, the capacity grid and the step-up cost, prompt Dezzmond with your facility terms and mini-perm date.

What Do Sponsors and Lenders Actually Check?

What is the balance at the mini-perm date, in dollars and as a percentage of the original?
What do the step-ups actually cost in equity NPV, discounted properly?
How much contracted life remains at the refinancing date?
What refinancing capacity does a stressed market provide, and what is the shortfall?
Is the shortfall fundable from sponsor equity, and is anyone committed to it?
Is it a hard or soft mini-perm? One is a default risk, the other a cost.
Does the amortisation profile during the mini-perm reduce the wall enough to be worth the distributions it costs?

Frequently Asked Questions

What is a mini-perm?

A project-finance structure whose legal or economic refinancing point occurs before the end of the asset's economic life. Filed examples range from legal maturities around five to seven years to structures combining amortization and cash-sweep mechanisms with a remaining balance expected to be refinanced.

What is the difference between hard and soft?

In a hard-style structure the refinancing wall is the legal maturity. In a soft-style structure the legal maturity is later and the borrower faces economic pressure before then. The labels are shorthand; the actual maturity, sweep, pricing and distribution provisions control.

How much do the step-ups cost?

On the worked case, $4.5m or 3.9% of equity NPV — less than sponsors expect, because the penalties operate from year eight and are heavily discounted, and because the resulting sweep repays the facility four years early.

What is the refinancing wall?

The balance outstanding at the mini-perm date — $97.1m, or 64.8% of the original facility, on the worked example. Unlike the step-ups it is not discounted away; it is a liquidity requirement in the year it falls due.

What does a refinancing cost?

Whatever the new financing fees and old-debt takeout provisions require: arrangement fees, legal / technical costs, hedge settlement, break costs or make-whole amounts. Model them from the actual documents or explicit assumptions rather than a generic percentage.

Can a refinancing release cash to equity?

Yes. A project with an operating record can often support more debt than its original sizing allowed, and the difference is a refinancing gain distributable to equity. In PPP structures a share of that gain is frequently payable to the public counterparty under refinancing gain sharing provisions.

What most affects whether a refinancing succeeds?

Remaining contracted life and CFADS at the refinancing date. A project refinancing with a long PPA tail and stable cash flow is routine; one refinancing into a merchant period with deteriorated cash flow is a different transaction entirely.

Closing: Negotiate the Wall, Not the Ratchet

The soft mini-perm is designed to look frightening. A margin that ratchets, a sweep that takes everything, distributions that stop entirely — it reads like a penalty clause and it attracts negotiation accordingly.

It is worth 3.9% of equity value. Discounting does most of the work: penalties that begin in year eight, in a structure valued at a 12% equity rate, simply cannot be very expensive, and the accelerated repayment they cause returns part of what they take.

The refinancing wall is worth considerably more attention and gets considerably less. Nearly two thirds of the original facility falls due on a date fixed at financial close, into a market that does not exist yet, against cash flow that will have been reshaped by every mechanism the preceding two contracts and legal materialescribed. In a benign market it refinances comfortably; in a stressed one it is short, and the shortfall is an equity call at the least convenient moment.

The useful discipline is to size the mini-perm around the wall rather than around the penalties. That means paying attention to the amortisation profile during the mini-perm period — the one place where accepting a faster schedule at the cost of distributions genuinely helps — and to the gap between the refinancing date and PPA expiry, which is the single variable that most determines whether the refinancing is ordinary or difficult.

the next post covers the accounts that sit between a shortfall and a default: reserve accounts, how they are sized, and what they actually protect against.

Sources: Bancolombia SEC filing — hard mini-perm described with five-to-seven-year legal maturity and balloon · Atlantica SEC filing — mini-perm with 15-year sculpt but 2027 legal maturity and margin step-ups · OPC Energy SEC filing — project loans combining amortization and cash sweep as mini-perm financing · Dezzmond analysis; worked-example refinancing terms and costs are illustrative.