The Fixed Cost of Being Allowed: Compliance Economics and Who Gets Served | HL Hunt
The Fixed Cost of Being Allowed: Compliance Economics and Who Gets Served
A licence, a compliance officer, a model governance framework, an audit, an examination, and a reporting stack cost roughly the same whether a lender writes a thousand loans or a million. Almost the entire cost of being permitted to lend is fixed, and that single fact does more to shape the consumer credit market than most of the individual rules do. It sets a minimum size below which a firm cannot exist, it decides which products are worth offering, and it pushes the market away from small loans and thin files, which is exactly where a different view of a borrower would be worth most.
In this report
What the cost structure looks like
| Cost | Scales with volume? |
|---|---|
| Licensing and applications | No |
| Legal and regulatory advice | Barely |
| Compliance staff and training | Weakly |
| Policies, procedures, and controls | No |
| Model governance and validation | No, per our validation analysis |
| Audits and examinations | Barely |
| Regulatory reporting and record systems | Mostly no |
| Disputes and adverse action notices | Yes |
| Complaint handling | Yes |
Only the bottom two rows grow with the book. Everything above them is a standing cost of operating at all, and a firm with one customer needs nearly the same version of it as a firm with a million.
Per our pricing analysis, this is the same structural point that report made about origination and servicing, applied to the regulatory layer. Fixed costs are the dominant term in small-balance lending, and compliance is one of the largest fixed costs there is.
Minimum viable scale
The first consequence, and it follows arithmetically.
Suppose the standing annual cost of being a licensed lender in a handful of states, with a compliance function, governance, audits, and reporting, is a figure in the low millions. That number has to be spread across whatever business the firm writes.
| Accounts | Fixed compliance cost per account, at $2m a year |
|---|---|
| 5,000 | $400 |
| 25,000 | $80 |
| 100,000 | $20 |
| 500,000 | $4 |
The figures are stylized, and the shape is the point. At 5,000 accounts the regulatory overhead alone can exceed the entire margin on a small loan. At 500,000 it is a rounding error.
Which produces three effects at once:
- A minimum size below which a firm cannot be viable, regardless of how good its underwriting is.
- A capital requirement before revenue, since the costs arrive before the first loan. Per our licensing analysis, a new entrant is paying for permission long before it has customers.
- An advantage to incumbents that is unrelated to how well they serve anyone.
From cost per firm to cost per loan
The second consequence, and the one that reaches borrowers directly.
Per our pricing analysis, a fixed cost per loan is a far larger share of a small loan than a large one. A $30 fixed cost is 6 percent of a $500 loan and 0.3 percent of a $10,000 loan. Adding regulatory overhead to that fixed component makes small loans harder to offer at any price a borrower would accept, and easier to abandon altogether.
So the arithmetic pushes lenders toward:
- Larger balances, where the fixed component is diluted.
- Standard data, because per our thin-file analysis, assessing someone the standard file does not describe costs more per applicant.
- Simpler products, since each additional product structure carries its own compliance work.
- Fewer states, per below.
Every one of those tendencies points away from the borrowers with the least access. Per our small-dollar analysis, the small loan is the product the market struggles most to supply at a tolerable price, and this is one of the reasons why. Per our rationing analysis, when a product cannot be priced viably, it is usually not priced higher, it is simply not offered.
Why geography multiplies it
Per our licensing analysis, in the United States the same fixed cost repeats across jurisdictions. A lender operating in several states is meeting overlapping and sometimes conflicting requirements, each with its own application, examination cycle, reporting, and rule set.
What that produces:
- Expansion decided by regulatory cost as much as by market opportunity. A state with a small population and a demanding regime may never be worth entering.
- Sequencing by convenience, so firms enter the easiest and largest states first, and residents of smaller or more demanding states see fewer options.
- An advantage to arrangements that avoid the state-by-state path, which is why partnership structures matter so much commercially and attract so much legal attention.
- Duplicated work of a kind that produces no additional protection, where the same control is evidenced repeatedly in different formats.
The last is the part worth separating from the rest. Meeting a standard protects consumers. Demonstrating the same compliance five times in five formats mostly protects nobody, and it is a real share of what small lenders spend.
The effect on variety
The consequence this desk cares about most, because it connects to a finding we have made from several directions.
Per our monoculture analysis, when every lender reads the same data the same way, a decline stops being one firm's opinion and becomes a verdict, because no lender sees the applicant differently. Fewer and larger firms make that convergence worse for reasons that have nothing to do with the modelling:
- Fewer independent decision makers means fewer chances that someone assesses an applicant differently.
- Large firms buy similar vendor components, per our vendor analysis, which pushes methods together.
- Experimenting with a different data source carries fixed cost, in validation, documentation, and fair lending review, so it is affordable mainly to firms with scale.
- Small lenders with local knowledge are exactly the firms the arithmetic removes, per our place analysis.
So the cost structure does not only decide how many lenders exist. It decides how many different opinions exist, and a market with few opinions serves the people who need a second opinion worst.
The cost structure decides how many lenders exist, and therefore how many different opinions about a borrower exist.
The case for the rules
Stated properly, because an analysis of regulatory cost that ignored regulatory benefit would be one-sided and wrong.
- The harms were real. Most consumer credit rules exist because of documented practices that damaged people, and per our consent analysis, disclosure alone did not prevent them.
- Protections that work are usually the substantive ones. Per the same analysis, the rules that function are those that do not depend on the customer reading anything, and those are precisely the rules that impose design and process costs on firms.
- Consistency requires documentation. Per our fair lending analysis, testing decisions for disparate impact is only possible because firms are required to record what they did.
- The cost of weak supervision falls on the same people as most other costs in this market, which is a serious argument against treating compliance cost as pure waste.
- Some fixed costs buy genuine capability. A model governance function that catches a broken attribute, per our attribute analysis, prevents harm at scale.
None of that is in tension with the argument. The claim is not that the rules are too strict. It is that the way a requirement is structured determines how much fixed cost it creates, and that two requirements with identical protective value can impose very different amounts of overhead depending on how they are written and examined.
Lowering cost without lowering standards
The constructive section. Each item reduces the cost of meeting a standard rather than reducing the standard.
- Requirements scaled to size or risk. A firm with 5,000 accounts and a firm with 5 million can meet the same substantive obligation with proportionate process, and many regimes already recognize this in places.
- Standardized artifacts. Common formats for reporting and examination evidence let one piece of work serve several purposes rather than being rebuilt for each.
- Reciprocal or coordinated review across jurisdictions, so the same control is not evidenced from scratch repeatedly.
- Clear published expectations. A large share of small-firm compliance spend goes on working out what is required, which is a cost imposed by ambiguity rather than by the standard itself.
- Shared infrastructure. Per our reporting analysis, common rails for furnishing and disputes already demonstrate that shared plumbing lowers per-firm cost without lowering obligations.
- Supervised routes for testing new approaches, which lower the fixed cost of trying a different data source and therefore protect variety.
Item four is the cheapest and least discussed. Ambiguity is a tax that falls hardest on firms without a large legal function, and resolving it costs a regulator far less than it costs the fifty firms currently guessing.
The strongest objections
"This is a lobbying argument dressed as analysis." A fair suspicion, since regulated firms routinely complain about compliance cost, and HL Hunt is one of them. The test is whether the argument leads anywhere self-serving, and this one does not: it asks for proportionate structure and shared infrastructure, not for weaker protections, and it concedes that the substantive rules are the ones that work.
"Small firms are not obviously better for consumers." Correct, and worth stating. Per our place analysis, local discretion was frequently discriminatory, and scale brings real benefits including lower costs and better controls. The argument is about variety of assessment rather than virtue of smallness, and variety could in principle come from large firms using genuinely different data.
"You have not measured any of it." True. The cost figures here are stylized and the effects are argued from cost structure rather than estimated, and the honest position is that firm-level compliance cost data is not generally published in a form that would settle the question. The tests below are how it could be.
Testable implications
- Compliance cost per account should fall steeply with firm size, which is the foundational claim and is measurable from firm disclosures where they exist.
- The number of small licensed lenders should decline as requirements accumulate, and should fall fastest in the most demanding jurisdictions.
- Minimum loan sizes offered should rise with regulatory overhead, since small loans are the first to become unviable.
- Entry into states should track regulatory cost as much as population, observable from licensing records.
- Adoption of new data sources should concentrate in large firms, because the fixed validation cost is the barrier.
- Proportionate regimes should show more small-firm entry without worse consumer outcomes, which is the test that matters for policy.
The sixth is the decisive one, and it is answerable wherever a jurisdiction has introduced tiered requirements. If a scaled regime produces more small lenders and no measurable increase in consumer harm, then the fixed-cost structure was excluding firms without protecting anyone. If harm rises, the fixed cost was buying something, and that is worth knowing with equal precision.
The conclusion we would hold: the cost of permission is fixed, so it falls hardest on small firms and on small loans, and it quietly shapes the market toward larger balances, standard data, and fewer independent views. The answer is not fewer protections. It is requirements whose cost tracks the size and risk of the firm, so that meeting a standard does not require being large.
Frequently asked questions
Most of what it requires exists once per firm rather than once per loan: licences, policies, governance, audits, examinations, and reporting systems.
Fixed cost per firm becomes fixed cost per loan, which is a larger share of a small loan, pushing lenders toward bigger balances and cheaper-to-assess applicants.
No. The harms were real and the substantive rules are the ones that work. The useful question is whether a requirement's cost scales with firm size and activity risk.
Scaled requirements, standardized artifacts, coordinated review across jurisdictions, clearer published expectations, and shared infrastructure.
Key takeaways
- Almost every compliance cost is a standing cost of operating, not a cost per loan.
- Spread over 5,000 accounts or 500,000, the same overhead is $400 or $4 per account.
- A fixed cost per loan is a larger share of a small loan, so small-balance credit is what disappears first.
- In a state-by-state system the same fixed cost repeats, and expansion follows regulatory cost as much as opportunity.
- Fewer firms means fewer independent opinions, which hurts the borrowers who most need a second one.
- Ambiguity is a tax on firms without large legal functions, and clarifying it costs regulators less than it costs the firms guessing.
This report presents an analytical framework and the authors' interpretation; it is not legal or policy advice, and HL Hunt is itself a regulated participant in the markets described. Cost figures are stylized illustrations rather than estimates of any firm's actual compliance expenditure, and requirements vary substantially by activity, licence type, and jurisdiction. The implications identified as testable are hypotheses.