Bulgaria's Parliament approved a new Law on Consumer Credit at first reading on 23 September 2026, transposing Directive (EU) 2023/2225. Most of what's been written about it since focuses on two things: the new registration regime for lenders, and the exact number attached to the new rate ceiling. Both matter. Neither is the point.
The point, stated plainly in the directive's own reasoning and repeated through the Bulgarian draft's explanatory memorandum, is narrower: a person taking out a loan is not expected to be able to assess it the way a lender can. They didn't choose to become a participant in a credit market the way a lender chose to become a participant in a lending business. They have no training in reading an APR calculation, no visibility into how a scoring model priced their risk, and no practical way to know whether a contract term is standard or predatory before they've signed it. The entire architecture of the law follows from that one asymmetry: since the person taking the loan usually can't protect themselves, the law has to do it before the fact, not just give them somewhere to complain after.
That's a different thing from what most of the coverage has been arguing about, and it's worth tracing precisely — because the directive's own logic has three distinct parts, and the gap between that logic and the version of the law that just passed first reading sits in exactly one of the three.
The directive's logic has three parts, and they aren't equally intact in Bulgaria's draft
First: it closes the old loophole that let the smallest loans escape protection entirely. The previous EU rules (Directive 2008/48/EC) excluded any credit agreement below €200 from consumer-credit protection altogether — no creditworthiness check, no standardized disclosure, nothing, purely because the amount was small. Directive (EU) 2023/2225 deletes that floor. Its own reasoning is explicit that this was deliberate: small-value, short-term, and "buy now, pay later"-style credit were exactly the products slipping through, and closing that gap is one of the recast's central purposes. Bulgaria's draft reflects this correctly — its scope provision carries no lower threshold at all, only an upper one (€100,000).
Second: having removed the floor, protection is meant to apply regardless of size. The creditworthiness duty, the advertising restrictions, the forbearance obligation — none of them are written with a size threshold in the directive. A €50 loan and a €50,000 loan face the same assessment duty, the same advertising bans, the same "prevent over-indebtedness" standard. The logic here matters: a borrower's vulnerability doesn't scale down with the loan amount. If anything it runs the other way — someone borrowing €50 typically has fewer options and less capacity to absorb a bad outcome than someone borrowing €50,000 against a mortgage-grade credit file. Bulgaria's draft keeps this part intact too: the assessment and advertising rules apply uniformly, regardless of amount.
Third: it draws one clear ceiling on cost, not a sliding one. The idea is a single, size-independent answer to "how expensive can credit legally be" — a bright line, not a scale that loosens for smaller amounts. This is the part that breaks in the version Parliament just approved. Rather than one ceiling, the draft creates two: the general rule caps the annual percentage rate at five times the statutory default interest rate, but loans up to roughly three minimum monthly salaries — which the opposition estimated during debate covers around 85% of all fast loans actually issued in Bulgaria — fall instead under a separate rule: total cost capped at 20% of principal for a one-month loan, or 30% for a three-month loan. That's not a blanket exemption like the old €200 floor was — no loan escapes protection entirely. But it recreates the same underlying assumption the floor represented: that small loans are somehow lower-stakes and can be governed by a looser rule. A concrete example raised in the debate: a €50 loan could end up costing roughly six times more in interest and fees than under the current rules, precisely because it falls inside this separate, weaker tier rather than under the general cap.
How much weaker, in an actual number
It's worth working out exactly how large that gap is, because "weaker" understates it.
Bulgaria's statutory default interest rate for the second half of 2026 — the reference rate the general 5× cap is built on — is the ECB's main refinancing rate plus 8 percentage points, currently 10.40% annually. Five times that is 52% — a real, if generous, ceiling.
The 20%-of-principal cap for a one-month loan needs to be converted to the same basis to be compared fairly, and that conversion has to be done the way GPR - in Bulgarian ГПР (the annual percentage rate of charge) is actually defined — as an effective, compounding annual rate, the same logic as an XIRR calculation, not a flat number multiplied by twelve. Bulgaria's own draft law makes this explicit: the ГПР is calculated "по формула, съгласно приложение № 3" (Art. 22, p. 2) — the standard EU annuity-equation formula, which solves for the single rate that equates a drawdown today with a repayment later. For a loan of principal D repaid once, at time t, as amount C: D = C ⁄ (1+X)ᵗ.
Solved for a 20%-of-principal repayment 30 days out (C/D = 1.20, t = 30/365): X ≈ 819% a year. Solved the same way for the three-month tier (C/D = 1.30, t = 90/365): X ≈ 190% a year. Those are the correct, compounding-consistent ГПР-equivalents of the draft's own 20% and 30% figures — not the roughly 240%/120% a flat annualization would suggest, and dramatically further from the general cap than a flat comparison implies. The one-month tier's true ГПР-equivalent runs at roughly sixteen times the general 52% ceiling that applies to every loan above the threshold.
For a sharper domestic comparison, look at what Bulgaria's own courts already treat as the outer limit of acceptable interest — not a penalty for late payment, but the ordinary, regular contractual interest rate charged for the use of the money itself. A settled line of Supreme Court of Cassation decisions (No. 906/2004, No. 378/2006, reaffirmed in No. 901/2015) holds that an agreed contractual interest rate exceeding three times the statutory rate on an unsecured claim, or two times on a secured claim, is void as contrary to good morals — applied to the everyday price of borrowing, the same thing the new law's cost cap is meant to govern. At the current 10.40% statutory rate, that's a ceiling of 31.2% a year for unsecured lending and 20.8% a year for secured lending.
The new law's 20%-for-one-month tier, expressed on the same compounding basis the ГПР formula requires, runs at roughly 26 times the ceiling Bulgarian courts already apply to an ordinary unsecured interest rate, and roughly 39 times the ceiling for a secured one. This is not a penalty being compared to a non-penalty. It is the same thing — the regular, contracted price of borrowing — measured the same way, and the gap between what courts already treat as void and what Parliament is about to write into statute is not a matter of degree.
To be clear about what a fair international comparison looks like on the nominal figure alone, before the compounding conversion: 20% for a one-month loan isn't unusual by international standards on its face. The UK's price cap on payday loans — in force since January 2015 and the most widely cited benchmark globally — sets 0.8% per day, which the UK regulator itself calculated works out to about 24% in total fees for a 30-day loan repaid on time. Bulgaria's 20% figure is the same order of magnitude, expressed the same nominal way. The difference is what surrounds that number. The UK's daily rate is one part of a three-part cap: it also limits default fees to £15, and — separately — caps the total, lifetime cost of the loan at 100% of the original principal, so a borrower can never be made to repay more than double what they borrowed, no matter how many times the loan is extended or rolled over. Bulgaria's 20%/30%-of-principal figures aren't one component of a layered cap like that. For the loans that fall under them, they're the entire ceiling on their own — and, converted to the GPR (ГПР) basis the law itself uses everywhere else, they're an order of magnitude larger than the nominal figure suggests.
Does any other EU country actually do this? No.
Before treating the tiering itself as a defensible policy choice, it's worth checking whether any other EU member state gives small or short-term loans a weaker cost ceiling than its general rule. None of the comparable cases do — and where a similar-looking mechanism exists, it runs the opposite direction.
- Finland applies one flat cost cap — 20% per year — to all consumer credit, with no size-based tiering whatsoever. That's not an oversight: an earlier version of Finnish law did cap only loans under €2,000, at a much higher 50%/year, precisely the kind of two-tier structure Bulgaria's draft now creates. Lenders structured their lending around that boundary, and Finland abandoned the tiering in 2019 specifically because of it, moving to one uniform rate for every consumer loan regardless of size.
- Poland does have a short-term-specific formula for non-interest costs (its "MPKK" cap) — but it runs the other way. For loans of 30 days or less, the ceiling is 5% of principal; for longer-term credit, the formula's fixed component alone is 10%, plus a further time-based charge. Poland's short-term tier is stricter than its general one, not looser.
- Spain, transposing the same Directive (EU) 2023/2225 that Bulgaria is currently transposing, created its own distinct tier in 2026 for high-cost, short-term credit — and built in the exact safeguard Bulgaria's draft lacks: "the maximum cost may not exceed that of a twelve-month loan for the same amount under the general regime", alongside a mandatory minimum three-instalment repayment term designed specifically to stop a short duration from inflating the effective annualized cost.
Put plainly: there is no EU practice of a small or short-term loan facing a weaker cost ceiling than everything else on the market. Every comparable country either refuses to tier by size at all, tiers in the stricter direction, or tiers with an explicit clause preventing the short-term rate from ever exceeding the general one. Bulgaria's first-reading draft is the outlier, and it's an outlier in the direction the directive's own logic argues against.
Aligned with the directive, or not — stated plainly
It's worth being direct about which parts of Bulgaria's draft actually carry out Directive (EU) 2023/2225's purpose and which one doesn't, rather than leaving it as an impression.
Aligned with the directive:
- No size-based floor (Article 2 scope; Recital 15). The directive's own reasoning states that closing the gap for small-value and short-term credit — the products that escaped the old €200 exclusion — was a central purpose of the recast. Bulgaria's draft carries no lower threshold at all. This part is correctly transposed.
- Uniform creditworthiness and advertising duties (Articles 7, 8, 18). These apply the same way regardless of loan size in both the directive and the draft. Also correctly transposed.
Not aligned with the directive's purpose — even though nothing in the directive's text is technically violated:
- The two-tier cost cap contradicts the reason Member States are allowed to set a cap at all. Recital 73 permits Member States to "maintain or introduce" caps on borrowing rates, APR, or total cost of credit for one stated reason: to prevent abuse. A cap structure is only doing what Recital 73 exists for if it prevents abuse where abuse is most likely. Bulgaria's draft does the opposite: it hands the weaker ceiling to the loan tier — small, short-term, necessity-driven — that is by every measure the most abuse-prone segment of the market. The directive doesn't specify what level a Member State's cap must be set at, so nothing here is a breach of the text. But a cap that provides less protection exactly where the directive's own justification for having a cap is strongest fails that justification on its own terms.
- It reintroduces, in substance, the same size-based discount on protection that Recital 15 was written to eliminate. Not the identical mechanism — no loan is fully excluded the way sub-€200 credit once was — but the same underlying premise: that a small loan is lower-stakes and can be governed by a looser rule. Recital 15 rejects that premise for scope. The draft reinstates it for cost.
- It answers risk with a higher price, when the directive's own answer to risk is not to lend at all. This is the sharpest point of divergence. Article 18 does not say a lender may extend risky credit provided the cost reflects the risk — it says credit may only be provided once the assessment concludes the consumer can meet the repayment terms. The directive's mechanism for handling a borrower who looks likely to struggle is exclusion from the loan, not a risk premium on it. A cost structure that lets exactly the segment most associated with thin credit files and necessity-driven borrowing carry a far higher price is applying ordinary risk-based pricing — the standard commercial logic of "charge more for the riskier customer" — inside a framework whose entire stated purpose was to replace that logic with a simpler rule: if the risk is real, the loan shouldn't be made, at any price.
That's the precise shape of the gap: two-thirds of the directive's own logic, faithfully carried into Bulgarian law; the third piece present in name (a cap exists, it is not unlimited, it technically satisfies "a ceiling was set") but structured in a way that undercuts the specific purpose the directive gives for having one.
Why this is the part worth watching
None of this means the law is badly conceived. Removing the old size-based floor and applying the creditworthiness and advertising rules uniformly are genuine, structural improvements — and they track the directive's stated purpose faithfully. But a law whose entire justification is "ordinary borrowers can't be expected to protect themselves, so the law has to" is only as strong as its weakest cost ceiling, because the cost ceiling is the one protection that doesn't depend on anyone reading, understanding, or negotiating anything. It's supposed to be the backstop that holds regardless of what else goes wrong. Splitting it into two tiers — and giving the weaker tier to the smallest, most necessity-driven loans — puts the backstop exactly where it's least protective and calls it protection anyway.
The law still has further readings ahead of it. Whether that split gets corrected, or at minimum brought closer to what Bulgaria's own courts already recognize as the line between an acceptable and an unconscionable rate, is a better measure of whether this legislation does what it says it's for than any of the registration or disclosure detail that's dominated the coverage of its first reading.