Asymmetric information, moral hazard, the systemic-risk externality, bubbles and market rigging; the excessive-risk diagram and the capital requirement that corrects it; the FPC, PRA and FCA, and the costs of regulating.
From 2007 until it banned them in 2021, the Financial Conduct Authority allowed “discretionary commission arrangements” in car finance: the dealer chose the interest rate, and the higher it was, the more commission the lender paid the dealer. Customers were not told. After the Supreme Court’s ruling in August 2025, the FCA confirmed on 30 March 2026 an industry-wide redress scheme covering about 12.1 million agreements signed between April 2007 and November 2024. It expects compensation of about £7.5bn, averaging £830 per agreement, and a total cost to lenders of £9.1bn once administration is counted. Lloyds, which owns the largest motor lender, had set aside £2bn.
It is the largest consumer redress since payment protection insurance, which cost banks £38bn between 2011 and 2020.
Why do financial markets fail so often, so expensively, and what does regulation do about it?
Source: Financial Conduct Authority, PS26/3 Motor finance redress scheme, 30 March 2026; FCA analyst briefing, 30 March 2026.
Asymmetric information. Sellers of financial products know more than buyers (motor finance, PPI) and borrowers know more than lenders (adverse selection, which is why lending to the unknown is rationed). Moral hazard. A bank whose depositors are insured and which expects a rescue takes risks it would not take if it bore the cost: too big to fail. Externalities. One bank’s failure damages others through contagion, freezes credit and cuts output: the private cost of risk is below its social cost, so too much is taken. Speculation and bubbles. Herd behaviour drives asset prices away from fundamentals (dotcom, US housing, crypto), and the bust destroys wealth and bank capital. Market rigging. LIBOR and foreign-exchange benchmarks were manipulated by traders at the largest banks.
The UK response since 2013 has three parts. The Financial Policy Committee at the Bank of England watches the system as a whole (macroprudential): it sets the countercyclical capital buffer and limits on high loan-to-income mortgages. The Prudential Regulation Authority, also at the Bank, supervises individual banks and insurers (microprudential): capital and liquidity rules, stress tests, ring-fencing. The Financial Conduct Authority polices how firms treat customers and markets: product rules, the Consumer Duty, fines, bans and redress schemes.
Regulation has costs: compliance, higher lending costs, activity pushed to unregulated shadow banks, regulatory capture by the industry it oversees, and moral hazard of its own if customers assume anything regulated is safe. The 2025 “Leeds reforms” loosened some rules to promote growth; the argument about where the balance lies never ends.
About 12 million car finance agreements signed between 2007 and 2024 are being compensated. The bill is £9.1bn.
Dealers arranging the loans were paid more commission the higher the interest rate they set, and did not tell the customer. The customer saw a monthly payment and a friendly salesman.
What kind of market failure is that?
Asymmetric information is the characteristic failure of financial markets: the product is complicated, the seller is paid to sell it, and the buyer finds out years later. PPI, endowment mortgages, motor finance: the same failure three times. It is why finance has its own regulator.
The market for risky lending. Demand shows the return on each extra pound of high-risk loans; the bank’s private cost (MPC) is the expected loss it bears itself. Left alone it lends £8bn at a return of 12%. Deposit insurance and the expectation of rescue keep MPC low: if the loans fail, depositors and taxpayers pay. That is moral hazard, and it is built into the position of the MPC curve.
Check: Which statement describes moral hazard in banking?
The protection that stops runs also removes the discipline that would otherwise restrain risk-taking.
Add the external cost: contagion, frozen credit, lost output, bailouts, here 4 per £. The social cost curve MSC sits above MPC. The optimum is £6bn, where MSB = MSC; the market lends £8bn. The extra £2bn costs society more than it earns: the welfare loss triangle, £4bn. 2008 was this triangle at national scale: a 6% fall in GDP and £65bn of public capital.
Check: The social cost of risky lending exceeds the bank’s private cost because:
The gap between MSC and MPC is the external cost of systemic risk: contagion, lost output, public money. In 2008 the UK economy shrank 6%.
The correction. A capital requirement forces the bank to fund risk with its own money, so losses fall on shareholders first: MPC rises towards MSC and lending falls to £6bn. The FPC can raise the countercyclical buffer in a boom and release it in a bust. Liquidity rules, stress tests and ring-fencing work the same way: they make the bank pay in advance for the risk it imposes on others.
Teacher tip: Students know the externality diagram from pollution; showing them the same diagram for banking makes the capital requirement a Pigouvian tax by another name. Then ask what the equivalent of “measuring the externality” is: nobody knows the true MSC, which is the regulator’s problem.
Check: A capital requirement makes banks fund risky loans with more of their own equity. On the diagram this:
Shareholders now lose first and lose more, so the bank prices risk as society does. Capital requirements internalise the externality the way a tax does.
Conduct. Asymmetric information is corrected by rules on disclosure, bans on products and practices (discretionary commission in 2021), the Consumer Duty requiring good outcomes, fines, and redress schemes: £9.1bn for motor finance, £38bn for PPI. Market rigging brought criminal convictions and fines of billions over LIBOR. Information failures cannot be priced away like an externality; they need rules about what must be said.
Check: Which failure is the Financial Conduct Authority, rather than the PRA, mainly responsible for correcting?
The FCA regulates conduct: how products are sold and markets behave. The PRA regulates safety and soundness; the FPC the system; the MPC monetary policy.
The limits. Regulators cannot see MSC exactly, so they may set capital too low (pre-2008) or too high (the growth complaint behind the 2025 Leeds reforms). Rules raise the cost of credit and push lending into shadow banks and private credit, outside the perimeter. Regulatory capture is a risk when the regulated hire their former regulators. And no rule prevents the next bubble forming somewhere new.
Check: The strongest argument against ever-tighter financial regulation is:
Regulation is itself subject to failure: cost, evasion, capture, and the moral hazard of an official stamp of safety.
Q* is marked at 10, above the market quantity of 8. The social cost curve lies above private cost, so the optimum must be below the market outcome: £6bn, where D meets MSC. A negative externality means too much, never too little.
Evaluate the view that regulation of the financial sector does more harm than good.
Financial markets fail in ways that justify regulation. Sellers know more than buyers, so products are mis-sold: motor finance dealers were paid more commission for higher interest rates without telling customers, a failure now costing lenders £9.1bn in redress after PPI cost £38bn. Insured depositors and the expectation of rescue create moral hazard. And a bank’s failure imposes costs on others, through contagion, frozen credit and bailouts, so the private cost of risk lies below the social cost. On the diagram the bank lends £8bn where its private cost meets demand; the social optimum is £6bn, and the excess generates a welfare loss. In 2008 that triangle was a 6% fall in GDP and £65bn of public capital for RBS and Lloyds.
Regulation corrects these failures. Capital and liquidity requirements, stress tests and ring-fencing, set by the PRA and the Financial Policy Committee, raise the bank’s private cost of risk towards the social cost, as a tax on an externality would, and reduce risk-taking to the optimum. Conduct rules, the Consumer Duty and redress schemes from the FCA address asymmetric information by requiring disclosure and banning practices such as discretionary commission.
**However,** regulation has real costs and failures of its own. Equity is dearer than deposits, so higher capital raises the cost of credit; activity migrates to shadow banks and private credit outside the perimeter, moving risk rather than removing it. Compliance costs fall hardest on small firms and new entrants. Regulators cannot observe the true social cost, so they set rules too loose before 2008 and, banks argue, too tight after it, which is why the 2025 reforms loosened them in the name of growth. Conduct regulation tends to arrive late: the commissions ran for fourteen years before the ban, and compensation comes two decades after the first loans. Regulatory capture and the false comfort of an official stamp add moral hazard of their own.
On balance the view is wrong for prudential regulation and only partly right for conduct regulation. The capital and liquidity rules imposed after 2008 bought a banking system that passed through the pandemic and the 2022 gilt crisis without a bailout, a large return on their cost. Conduct rules have a weaker record, catching each scandal after it has happened. Regulation does most good where the externality is largest, in the biggest banks, and least where it substitutes the regulator’s judgement for the customer’s; the right question is not whether to regulate but how much, and where.
Motor finance figures: Financial Conduct Authority, PS26/3 Motor finance redress scheme, 30 March 2026 (about 12.1 million agreements from 6 April 2007 to 1 November 2024; estimated redress £7.5bn at 75% take-up, average £830; total cost to firms £9.1bn). The risk-taking market on the diagram is invented. Diagrams drawn with Rough.js and perfect-freehand, both MIT licensed.