UPDATE: Amigo gave some more information about the Scheme on 25 January, see my article Amigo’s Scheme – timetable for approval & how it may work. But there weren’t any of the useful numbers I had been hoping for, so I have not made any changes to the rest of this article.
This article looks at some possible numbers to see how this Scheme might pan out if the Financial Conduct Authority (FCA) allows it to proceed.
I am interested in:
- how Amigo can assess claims in line with Financial Ombudsman decisions
- what refunds customers might get
- whether there is enough money in the Scheme to make it workable, and
- whether it is equitable for all Amigo’s stakeholders
I have had to make a lot of estimates and guesses to be able to do any calculations. Some of the numbers and estimates are in Footnote 1 at the bottom of this article.
How much money is going into the Scheme pot?
The Company anticipates the immediate cash payment under the Scheme towards the Redress Claims Pool will be £15 million, with the option for this to increase should the anticipated Balance Adjustments be less than expected, up to an extra £20 million of cash. In addition, the Company will make a cash contribution to the Scheme based on 5% of profit for the next three financial years ending 31 March 2024. In light of the cash contribution to the Scheme and the Balance Adjustments to the loan book, the Company believes the requirement for the current unutilised complaints provisions remains in line with expectation.
The previous provision for complaints was £159m in the half-year accounts. Some of that will already have been used for complaint settlements and expenses. A significant amount will be needed for the cost of setting up and running the Scheme. Together somewhere between £30 and £50m? Let’s say there may be £120m available from the Complaints Provision for the Scheme.
I am not clear why Amigo is apparently distinguishing the money in the provision from the new money available to the redress claims pool. If it wants to put all customers into the same “class” to vote on the Scheme, I think there has to be a single pot available to that class. If the balance adjustments come to a lot less, then why doesn’t all that extra go to the pool for cash refunds? If the balance adjustments come to more, then surely they have to take priority and there may be no money left for any other refunds.
How will claims be assessed – details are needed
In payday loan administrations, the administrators have faced this problem. Their approach is to develop an automated process where the decisions would be broadly in line with what FOS would decide and have a similar uphold rate.
Amigo needs to do the same here. It has not given any details as to how it would do this. FOS has been upholding over 80% of Amigo cases.
This matters because the FCA needs to be sure that this Scheme isn’t an attempt by Amigo to uphold fewer complaints than FOS would. No authorised firms should be able to think they can undermine FOS decision-making by setting up a Scheme.
Amigo’s own records don’t have verified income and expenses, which is what FOS works from. But it seems to me that Amigo could make a reasonable attempt to uphold cases in line with FOS by accepting any complaints with one or more of the following characteristics:
- customer appears to be in a debt management plan or has multiple payment arrangements;
- recent significant CCJ/defaults/payday loans;
- essential items missed from expenses or have amounts which are unfeasibly low;
- customer was assessed at an implausibly high disposable income;
- has applied for a top-up but finances had got worse;
- customer was relying on partner to pay essential bills but this was never checked;
- any payments by the guarantor;
- no verification of self-employed income.
What Amigo can’t do is reject a complaint because it was the first loan, because all payments were made on time or because it considers the income and expenses on the loan application were incorrect. FOS is routinely upholding guarantor loan complaints in all these situations.
Amigo also needs to explain how some cases will be handled:
- will the Scheme make a deduction for “unpaid interest” when not all loans are upheld? FOS does not do this.
- what will happen if a borrower’s complaint is upheld and the guarantor has made payments?
- how will complaints about loans sold to a debt collector be handled?
- will CCJs and charging orders be removed?
- what about complaints where a guarantor complains that the loan is unaffordable for the borrower – will these be assessed on what is known about the borrower?
- what about complaints where the customer, typically the guarantor, says they were coerced, pressured or conned into agreeing to the loan. Will these be considered by the Scheme? In many but not all situations here, the guarantor is also complaining that the loan was unaffordable for them.
Customers with a current balance
The right of set-off
The Scheme announcement said:
there would be an automatic right of set-off for the redress due to claimants against their current outstanding loan balance.
A simple example shows how this treats a borrower very differently depending on how much has been repaid. Take a £4000 loan for three years, monthly payments £195. The interest would be £3020, which is the basic redress if the loan is decided to be unaffordable:
- after a year, the borrower has paid £2340 and owes £4680. The redress is used to reduce the amount owed to £1660. So the borrower “has 100%” of the £3020 redress through a balance write-down.
- after two years, the borrower has paid £4680 and owes £2340. The redress clears the balance, so the borrowed has 100% of £2340 and then only gets the standard pence in the pound payout for £680. The pence in the pound payout may be very low, see below.
- after three years, the loan is just paid off, the borrower has paid £7020, owes nothing. The borrower gets only the standard pence in the pound payout for £3020.
Of the 150,000 people with a current loan, roughly one-third of the ones who only had one loan and who hadn’t taken a Covid-19 payment break would still owe a balance and two-thirds wouldn’t after redress. That is a big simplification with a lot of assumptions. See Footnote 2 if you are interested in these.
Clarification: someone asked in the comments below whether many more customers would not have paid off a lot as the Scheme will take a year or more to give the balance write-off? I think the answer is No, because in Administrations and the Money Shop Scheme, current debtors who have a complaint upheld get a full rebate of the money they have paid in since the Scheme started. So dragging out the Scheme will not reduce the cost of balance reductions.
These differences are important for borrowers. They are also key to the cost of the Scheme for Amigo. Balance write-offs have to be covered in full.
This right of set-off would also apply in administration, so the customers who appear to get a generous write off are not actually benefiting more from the Scheme than they would if Amigo goes under.
How many are likely to make a claim?
This is a big unknown.
One group won’t be claiming – those who have already had a complaint upheld. But they are probably a small amount of the loan book as their balances will already have been reduced.
Of the rest, past experience with payday lenders isn’t much help because Amigo’s loan book and customer details are very different:
- Amigo loans are large. Quite a few payday loan customers only took one or two small loans and repaid them without problems so they would never claim.
- Over a third of Amigo’s current borrowers have probably had some financial difficulty this year as they applied for a Covid-19 payment break. The fact their loans are now being repaid doesn’t mean their financial problems are solved, just that they are desperate to protect their guarantor.
- If the borrower misses the email or doesn’t think it applies to them, the guarantor will also be getting an email so may encourage the borrower to claim.
- Unlike payday lenders, Amigo isn’t carrying a large number of old, defaulted loans with many inaccurate customer contact details.
I find it hard to imagine fewer than half the current customers would claim and it could well be three-quarters. Perhaps 100,000 claims?
What would this cost Amigo in balance write-offs?
Assume an uphold rate of 80% (see above) so 80,000 claims upheld out of the guessed number of 100,000.
Average loan book value of £3400. It would be better to have the average redress value for the upheld claims but that’s complicated. Let’s assume the average redress for these customers is the same as the average loan book value – this simplification is in Amigo’s favour.
My theoretical calculations above suggested one-third of people with a single loan would get their balance reduced by the whole amount of the redress on that loan.
I shall assume that of the remaining upheld complaints, a half get a write off of two-thirds of the amount and a half get a write off of one-third of the amount.
On those assumptions I make the total write off about £180m. That is more than Amigo is proposing to put into the Scheme.
I don’t know what Amigo is contemplating doing in this situation… if it covers the write-offs in full, would it still be solvent? If it is proposing to restrict the balance adjustments, then these customers would get less than they would in administration which cannot be acceptable.
Amigo seems to be thinking the total balance write off would be less then the provision – so £120m. The only way I can get the total write off down to that is by using low complaint numbers or an uphold rate that is much lower than the Ombudsman’s.
Customers whose loans were settled
- 500,000 previous borrowers;
- 50,000 previous guarantors who had made a payment;
- average redress £5000;
- assume claim uphold rate of 80%.
Again the big question is what the complaint rate will be.
The complaint rate will be lower than for current loans because Amigo is far more likely to have out of date contact details and some emails sent to the correct address will get put into spam.
But it won’t be that low. No-one forgets having had an Amigo loan. The guarantor may suggest that the borrower should claim. Few people will think of an Amigo loan as something small and unimportant. The FCA’s recent survey on relending found a high level of borrower regret for top-up loans.
Let’s say 20% of borrowers complain and ignore the guarantors. Let’s ignore the people with a small balance reduction who are owed more redress.
That gives 100,000 complaints, 80,000 of which are upheld at an average redress of £5,000. A total redress bill of £400m.
If the pot to be divided is only £15m (it could be less, even zero. Any more seems unlikely.) plus £10m more from the 5% of profit in the next three years provision (that feels very optimistic given previous profit amounts) that is £25m.
That would give a pay out of 6p in the £ over a period of 4 years.
These feel very conservative estimates for complaint numbers, they could easily be higher. So I think 6p is maximum, unless Amigo has a lower uphold rate. In reality it could well be 2-3p in the £.
Treating customers equitably
In its announcement, Amigo said:
A key objective is that all customers are treated equitably.
It seems to me that there are three parts to this.
First, are the complaints being upheld in a fair manner? Customers need to be able to make a reasonable guess as to whether their complaint would be upheld in the same way it would at the Ombudsman. Given Amigo’s very poor complaint handling this year, many borrowers will have no faith in Amigo saying what it is proposing is fair and equitable without a lot of details before they are asked to vote.
Secondly, is the division of the money between the customers fair? I think customers would understand that people with a balance will do better. But I am not sure they will accept that people who have repaid the loans should only get a few pennies in the pound where people with balances get a full write off, even though that is what happens in insolvency. I repeat what I said before – these groups of customers are in such different positions that they should be treated as two separate classes in a Scheme.
Thirdly, are customers being treated fairly compared to other stakeholders?
Looking at this Reuters article, two weeks ago the Amigo bonds were on a yield of over 27% – other bad credit lenders such as Provident and NSF had bonds trading at 9-11%. That suggests that the bondholders do not expect they will get all their money back if Amigo goes into administration. So why are they not being asked to share in the pain now, with a debt for equity swap ,for at least part of their holdings?
Shareholders would of course lose the entire value of their holdings in administration. Instead of this, the Scheme will give them the benefit of 95% of the profits over the next three years. That feels fundamentally unfair to me. I would suggest that customers should get 50% of the profits each year until they have received at least 50p in the £, and during this time there should be no dividends paid.
When I looked at the Money Shop’s proposed Scheme last year, the important numbers turned out to be how many people would complain and what the uphold rate would be. The same applies here.
I needed a lot of other numbers to make estimates of what refunds customers could get under the Amigo Scheme. I had to use estimates and assumptions. Amigo probably knows many of the actual numbers – I hope they have shared them with the FCA.
On my calculations, it looks as though Amigo may not be putting enough money to allow people with a current loan to have full set-off. And people with settled loans would be getting 6p or less in the £.
Meanwhile the bondholders are not being asked to take any losses at all and shareholders get away with only a tiny impact on future profits.
I can’t see how this can be an acceptable proposal to the FCA.
1 – some numbers and their sources
- Amigo has 150k customers with a current loan, and 500k past borrowers (from this BBC article.)
- customer numbers were rising until lockdown: 222k in 2019/20; 201k in 2018/19; 169k in 2017/18. (from 2019/20 Report & Accounts.)
- average redress amount – estimate £5000. (Average redress won’t swing wildly so 10% would be typical for sensitivity analysis. The 2019/20 Report & Accounts used +/- £500, implying £5000 average.)
- 28% of loans were relending in 2019/20, down from 39% in 2018/19. (from 2019/20 Report & Accounts.)
- statutory profit/loss: -27m in 2019/20; 88m in 2018/19; 50m in 2017/18. (from 2019/20 Report & Accounts.)
- 513m estimate of gross loan book at end 2020 (from half-year accounts – gross loan book 558m at end September 2020 and loan repayments had been 91m in six months).
- average balance owed to Amigo is estimated at £3400 (gross loan book divided by the number of current customers).
- 57,000 Amigo customers took a Covid-19 payment break, ie more than one in three. (from half-year accounts).
- Amigo had 275 CCJs in Northern Ireland in 2018. That suggests there may be tens of thousands of CCJs across the whole book. (Registry Trust statistics)
- Amigo often says that less than 10% of guarantors are making a payment. I am unclear if this is less than 10% of guarantors ever make a payment? Or a snapshot at a point in time.
2 – assumptions in working out balance write off
- the open loans are evenly spread over the 4 years before the first lockdown. Weighting the loans to be more recent as Amigo’s business has been increasing means more people are likely to owe a balance after redress.
- all the loans are over one year old – ignores the low number of loans post lockdown;
- taking a Covid-19 payment break means people are more likely to still owe a balance after redress as they have paid off less;
- having had previous loans means people are likely to get more redress so they are less likely to still owe a balance after redress. )
3 – it’s complicated to work out average redress for loans with balances
Across all complaints, the average redress value was estimated at £5000. (UPDATE – Amigo has confirmed that £5,000 is a reasonable figure for the average redress). But that may not be right for people with an open loan who have made a complaint:
- complaints about older loans that have been settled several years will tend to get a higher amount so the average redress for people with an open loan will be slightly lower;
- only people with a higher than average balance could get a balance write-off of £4000 – needs the distribution to be modelled.
- the average balance is reduced artificially by including customers who have already won their complaint;
- it seems likely that people with a larger debt are more likely to complain.