, Britain
In September 2021 an energy supplier in Cardiff had ninety thousand customers, a bank balance that looked healthy, and a wholesale gas price on the screen that had done something the model did not have a column for. The finance director had been asked in a board meeting that spring how far forward the book was hedged. The answer was eight months, and at the time that sounded prudent.
By the autumn it meant the business knew exactly when it would run out of cheap gas, and roughly what it would cost to buy the next lot.
Between August 2021 and the following spring, twenty-nine suppliers failed and their customers were moved to other companies under the supplier of last resort process. Almost all of them had a healthy-looking bank balance a month before they went.
The cap says what you may charge and the market says what you must pay
Ofgem's default tariff cap, introduced under the Domestic Gas and Electricity (Tariff Cap) Act 2018 and in force from the start of 2019, limits what a supplier may charge a customer on a standard variable tariff. It is not a cap on profit or on cost. It is a ceiling on the unit rates and standing charge, calculated to allow an efficient supplier a reasonable margin, and it is set for a period at a time.
The mechanism has a lag built into it, and in 2021 the lag was the whole story. The cap level for a period is calculated from wholesale prices observed in a window that closes before the period starts. So a cap taking effect in the autumn reflects a market that has already moved on, and if the market moved sharply upward inside that window, every supplier is selling below cost until the next reset, by law.
A supplier cannot respond in the usual way. It cannot put the price up. It cannot decline to supply a customer it already has. Energy supply is a licensed activity with obligations attached, and one of them is that you keep supplying.
A hedge is a promise about volume as well as price
Buying forward is how a supplier makes a fixed tariff possible. If you sell a customer a fixed price for twelve months, you should be buying the gas and electricity to serve them at a known cost over the same period, so that the margin is locked rather than gambled.
In practice that is harder than it sounds, because a hedge is a bet on how much you will need as well as what it will cost. Two things move the volume.
The first is churn. Suppliers built their hedges on assumptions about how many customers would leave during the year. In a market where switching was cheap and constant, those assumptions were reasonable. In late 2021 switching effectively stopped, because there was nothing cheaper to switch to, so customers who were expected to leave stayed, on tariffs that were now well below cost, for the whole of the winter.
The second is the weather. A cold winter means more consumption per customer than the hedge covers, and the shortfall is bought on the day, at the day's price.
The reverse risk is real too and it caught firms in other years. A supplier that hedged heavily and then lost customers is left holding energy it cannot use, in a falling market, at a price it agreed to. Hedging is not a way of avoiding risk. It is a way of choosing which risk you are taking.
The money in the account belongs to the customers
Domestic customers on direct debit pay a level amount every month rather than paying for what they used. In summer they use very little and pay the same, so a credit balance builds up. In winter they use a great deal and pay the same, so the credit balance is drawn down.
This means a supplier's cash position peaks at the end of summer, which is precisely when it looks strongest and is about to be tested. In September the bank balance contains months of customer money that has been collected and not yet earned.
There is a second pot with the same quality. Suppliers collect environmental and social obligation costs through bills continuously and settle some of them, including the Renewables Obligation, once a year. Between collection and settlement the money sits in the supplier's account looking exactly like working capital.
None of it is the company's. All of it can be spent, and spending it is the single most reliable way to finish a supplier, because it funds growth and losses today with an obligation that arrives in full on a date somebody else has already set.
A failing supplier looks solvent until the week it is not
This is why the sequence in 2021 surprised people who were watching the wrong numbers. A supplier with a full bank account, a growing customer base and a market-leading tariff can be deeply insolvent, because the cash is other people's, the growth is loss-making and the tariff is the reason.
Growth makes it worse rather than better. Winning a customer costs money in acquisition and takes months to earn anything back, and a customer won onto a tariff priced below the cost of supply is a liability signed up voluntarily. In a rising market, every new customer deepens the hole while making the top line look better.
Ofgem had already tightened the test applied to new suppliers before the crisis, and it introduced checks on the financial resilience of existing ones, including the power to restrict a supplier from taking on new customers. After 2021 the requirements were tightened further.
When a supplier did fail, Ofgem appointed another company as supplier of last resort. Domestic customers kept their supply and their credit balances were honoured by the new supplier. The cost of doing that, along with the network charges and obligation payments the failed supplier had collected and not passed on, was recovered through an industry levy, which is to say through everybody else's bills. Where a company was too large for that process to work, a special administration regime existed and was used. Each failure therefore made the survivors slightly weaker, which is a quality fast crises have and slow ones do not.
Where this shows up in the game
The Cap and the Curve puts you in charge of an energy supplier in Cardiff in September 2021, with ninety thousand customers and eight months of hedge, and runs for a year. It is marked Impossible, and it starts with the bank having already given notice on your overdraft, which removes the facility most suppliers were quietly relying on.
The hedge covers eight months of demand and cannot be extended at the old price. The price cap is fixed for each six-month period and you cannot charge above it. Customer credit balances appear in your cash as they do in real life, and the game will let you spend them, which is the decision the scenario exists to offer you. A quarterly resilience check can stop you taking on new customers, and growth remains available throughout, and remains fatal.
You are scored on whether you are still supplying at the end of the twelve months, and the epilogue reports how much of your book was hedged when the price moved, which is the honest measure of what you did. The page for The Cap and the Curve sets out the starting figures and the rules. The seed is fixed at 600020, so everybody attempting it meets the same market and the same winter, and the scores can be compared. It runs under ironman rules: one save, and what happens happens.