Here’s the written submission I made to the BIS Select Committee in advance of yesterday’s evidence session (recording now available).
It should be read in conjunction with the Rothschild review transcription, as that document explains what is going on with the current loan sale proposals.
Here is the pdf: BIS select committee submission
False Economy have published the Rothschild review with a commentary from me. What’s contained therein underpinned my evidence to the BIS Select Committee yesterday.
The transcription of the Rothschild feasibility study into the sale of loans was a team effort, so I only saw this passage this morning.
Conventionally capitalism requires that a real return on capital should exist i.e. in other words that RPI should be below interest rates. However, right now:
-
Capitalism is suspended,
-
Base Rate is an issue of national policy, and
-
RPI and Base Rates are currently Inverted
In such circumstances, Rothschild is recommending that the government or, indeed, universities should pick up the risks that investors cannot tolerate. Risk is ‘socialised’, private returns protected.
The BIS Select Committee will be holding a meeting on ‘Student Loans’ on Tuesday 17 December.
I will be appearing alongside Toni Pearce (NUS president) and Bahram Bekhradnia (Director of HEPI).
It starts at 9.30am. You can watch it live online at Parliament TV or even attend as it is open to the public.
Update 17 December
A full recording of this morning’s session is now available.
Q: You are selling the student loan book. That just generates a one-off fee. Do your calculations take into account the long-term effects of losing this income?
No, they don’t says Osborne. The Treasury figures just look at the immediate costs. The OBR looks at the long-term consequences.
The implication that this is a long-term matter is incorrect; even for his ‘scorecard period’ the graduate repayments sold to purchasers amount to £1.7billion. If he is still claiming that the change in cash outflow as a result of additional grants and loan outlay will be covered by net sale proceeds over 2013/14 to 2018/19, he may just be right. (‘Osborne says there are two costs: grants, and borrowing money to lend to students.This is a cash flow issue.’)
But once 2019/20 is included, he’s wrong.
Moreover, he cannot continue to maintain the Autumn Statement’s claim that this policy will also reduce Public Sector Net Debt over the period.
Here’s a quote I gave to the Guardian yesterday:
How will the planned expansion of undergraduate places will be funded after 2015/16? We don’t have the details yet: setting out policies using gross, rather than net, proceeds is incompetent. If you sell the loans, you no longer receive the associated income stream. That should be obvious.Beyond 2019/20, there are no more sale proceeds, but income will continue to be £1billion lower than previously estimated. What happens then? This is hit-and-hope policy making.Higher Education deserves better – a clear, sustainable financing solution without gimmickry.
I could have added: not only is income lower, but you’ve also got those additional places to fund.
The BIS Select Committee will be holding a meeting on ‘Student Loans’ on Tuesday 17 December.
I will be appearing alongside Toni Pearce (NUS president) and Bahram Bekhradnia (Director of HEPI).
It starts at 9.30am. You can watch it live online at Parliament TV or even attend as it is open to the public.
Update 17 December
A full recording of this morning’s session is now available.
On Monday morning, I put my questions about the Autumn Statement to the Treasury. I’d spotted an omission of £1.7billion: the downwards revision to estimated graduate repayments as a result of selling a portion of the loan book.
Yesterday evening, I finally received a reply:
“Consistent with the OBR’s approach and for reasons of commercial sensitivity, table 2.5 in the Autumn Statement did not specify a figure for lower repayments, however these figures are fully consolidated into the calculation of Public Sector Net Debt. The Government has always been clear that selling the student loan book would obviously reduce the income received from repayments. Even when the OBR’s estimate of lower repayments are included, the sale of loans more than finances the cost of new, additional loans over the scorecard period.”
It took them two days to concoct that.
I particularly like the ‘obviously’ and that the change to graduate repayments is somehow ‘commercially sensitive’, but the ‘gross proceeds’ from the sale programme are not. Let’s hope no one else reads the OBR’s Economic and Fiscal Report!
The IFS called the proposal to fund more loans through the sale of loans ‘economic nonsense’.
I think it’s worse than that – back of the envelope calculations missed basic facts about financial assets. Watch this space.
Update
I can’t resist pointing out the ‘kettle logic’ in the quote: ‘these figures are fully consolidated into the calculation of Public Sector Net Debt. … Even when the OBR’s estimate of lower repayments are included …
Table 2.5 of the Autumn Statement needs urgent revision.
Mr Willetts: Estimates for the impact of RAB charge changes from 2016-17 are highly dependent on the future growth of earnings, and forecasts of spending for years beyond FY 2015-16 have not yet been made.
However, the impact of 1% pt increase in the RAB charge on loans issued in 2014-15 would be around £100 million. The impact of 2% pts increase in the RAB charge on loans issued in 2014-15 would be around £200 million. The impact of 5% pts increase in the RAB charge on loans issued in 2014-15 would be around £500 million.
Any change to the RAB charge will impact the net expenditure and balance sheet position in the Department for Business, Innovation and Skills annual accounts, but does not impact the current deficit or Public Sector Net Debt.
…
Mr Willetts: In March 2011 we estimated that around 30% of the value of post-2012 loans would not be repaid. We currently estimate that around 35%-40% of the value of these student loans will not be repaid, a change of 5%-10%. This is largely due to an increase in the value of the £21,000 repayment threshold relating to forecast earnings.
That means the expenditure needed to cover each year’s loan outlay has officially climbed by £500m to £1billion. But that’s on an estimated annual outlay of £10bn – that’s already out-of-date.
Will the promise to the Lib Dems be kept? That the repayment threshold will rise in line with earnings from 2017? The relevant regulations had still not been made at the end of 2012.
Last week’s Autumn Statement announced an expansion of higher education by 60 000 places from 2015/16. Little detail has been available on how that expansion will affect the departmental budget for Business, Innovation and Skills after that first year.
The Statement did however claim that loan outlay for those additional places would be funded from the planned sale of a portion of the student loan ‘book’.
§1.203 This expansion is affordable within a reducing level of public sector net borrowing as a result of the reforms to higher education finance the government has enacted. The additional outlay of loans over the forecast period will be more than financed by proceeds from the sale of the pre-reform income-contingent student loan book. Taking the two together, public sector net debt by 2018-19 will be lower as a result.
The details were presented in the table below taken from page 83 of the Statement.
[Thanks to Jonathan Clifton, IPPR for making this image available online: click to see larger version]
There is however a problem with this table. A significant ‘impact’ has been omitted.
Gross proceeds from the loan sale are shown (£2.3billion per year from 2015/16) but the Treasury has forgotten that if you sell part of the loan book you will lose the graduate repayments that are now going to the purchaser.
As noted on Friday, The Office for Budgetary Responsibility wrote in its Economic and Fiscal Outlook:
§4.145 Selling the loan book reduces repayments over the latter years of our medium-term forecast, by just under £1 billion in 2018-19, and beyond, whereas removing the numbers cap increases forecast outlays by around £2 billion by 2018-19.
OBR have also confirmed to me that for 2016-17, they estimate a resulting downwards revision to repayments of £200million, and £500million for 2017-18.
Leading to a total downwards revision over the three-year period to 2018-19 of £1.7billion.
If you plug those figures into Table 2.5 the impact on ‘total policy decisions’ goes into the red with a loss overall of £570million across the full period shown.
In 2018/19, just focusing on loans, £2.3billion of sale revenues minus £1billion of lost repayment income is a long way short of the £1.93bn needed to provide loans to the additional 180 000 students (3 years of an extra 60000 places).
Of course, you can strip out the Green Investment Bank, Start Up Loans, rent to buy and ‘unlocking large housing sites’ from the above table. You could then argue that strictly speaking the claim made about loans and debt was correct. But PSND is not going to be reduced as a result of this suite of proposals.
I put my questions to the Treasury this morning and they were unable to provide comment by this evening. BIS were unwilling to discuss ‘a Treasury document’ but released a fact sheet this morning:
“… a final decision to go ahead with the sale has not yet been taken. Any decision will require a full assessment of the value for money to the taxpayer of selling the loans against the cost to Government of retaining them.”
So … what to make of it all? And what happens after 2020? The ingenious coup de théâtre of Thursday now seems closer to bad bunko gimmickry.
Update – 11 December
This omission was confirmed by the Treasury. Table 2.5 of the Autumn Statement needs fundamental revision.
So maybe it’s also worth setting out the basics of loan sales clearly. Or at least trying to.
The government is hoping to raise between £10 billion and £15 billion in ‘sale revenues’ between 2015 and 2020. It’s ‘central estimate’ which is repeated throughout the Autumn Statement documents is £12 billion, with £2.3bn each year on the basic modelling published.
§2.16 Borrowers’ loan terms will be fixed prior to a sale. The interest rate charged will remain at the Retail Prices Index (RPI) or base rate plus 1%, whichever is lower.
So it looks like a ‘synthetic hedge’ will be used to persuade enough buyers to take a piece of the action. I won’t go into that here, but it amounts to offering purchasers a deal whereby they are compensated if interest rates remain below RPI.
There are three ‘values’ needed to understand loan sales:
- ‘face value’ – the aggregate total of outstanding loan balances. At the end of March 2013, loan balances made to English graduates stood at £43billion;
- ‘fair value’ – this is what is recorded in the departmental accounts. It is the value of the asset in terms of anticipated future cash flows (in net present value terms). The fair value of those same outstanding loan accounts was assessed as £31billion at the end of March;
- ‘price’ – the upfront cash purchasers would pay the government. This will be lower than ‘fair value’ perhaps as much as 5-6p lower per pound of face value. That is, if the government thinks £1 of face value is worth 65p in fair value terms, then it might get 60p in price.
So, the first point is that much of the discussion of the sale of the ‘old style mortgage loans’ announced last Monday missed the difference between face value and fair value. Those loans had a face value of £900m but were worth considerably less given their particulars.
The second point is that any sale will lose ‘value’, if you compare the price to the fair value. But a seller who wants cash upfront today may be prepared to forego an income stream of greater present value if it is spread over 30 years. (You might wonder why the government should be in that position given that it doesn’t appear to face any lending constraints at present).
The question then is: if you want to raise £10-15bn, what proportion of the loan book do you have to sell? The government probably has a fair idea about price, but is not so sure about how much market appetite exists.
This sale will be structured by auctioning whole cohorts (determined by year of graduation commencing with the oldest, the 2001-03 vintages which are expected to fetch a better price) and is envisaged to focus on pension funds and insurance companies. Purchasers have to be large enough to purchase tranches of £1billion at a time and want this kind of asset. (ICR loans are unusual with odd repayment patterns meaning that many potential buyers may be reluctant to take them on – this is one reason the price will be lower than the government’s fair value). In 2011, Rothschild told the government that total ‘appetite’ would be in the region of £10billion. That is, it’s not clear that such a sale is repeatable. Hence the concerns about quite what was announced yesterday for sector expansion.
The government may try to sell as much as possible so we may be looking at half of all existing loan accounts. The sales will be spread across the five years, 2015 to 2020. The decision to sell rests with the Secretary of State, Vince Cable; loans can be sold without consultation or the consent of borrowers.
The IFS called yesterday’s proposal to fund more loans with the loan sale, ‘economically nonsense’. One thing to consider: why would you sell this asset class at the bottom of the market? that is, when the economy is only just beginning to recover from recession. If you believe in ‘the growth to come’, wouldn’t you be better holding on to an income stream tied to graduate earnings?
Now’s there’s a good question. Will we be in the perverse situation of the alternative providers labouring under greater restrictions than established, HEFCE-funded universities and colleges?
The OBR writes:
§4.144 The number of new students undertaking courses with alternative providers is expected to peak in 2013-14, but to remain above the March [2013] forecast throughout.
And it seems clear now that HNC and HND courses, subdegree qualifications, will continue to be restricted after a harum scarum explosion in 2012/13. Responding to a parliamentary question from Liam Byrne this week, Willetts revealed that numbers on those courses leapt from 2 590 in 2011/12 to 15, 320 in 2012/13, while student support outlay (loans and grants) shot up from £21m to £131m. BIS stated very recently that they want to concentrate on degree-level provision at alternative providers.
The Autumn Statement reaffirms that in 2014/15 numbers controls will be introduced for the first time at private providers who want access to the student support. The institutional caps will be based on 2012/13 recruitment levels but the OBR comment above suggests that the limits will be rebased so that allocation turns out to be lower .
§1.204 From 2015-16, [the government] will allow student numbers at alternative providers to be freed in a similar manner as for HEFCE-funded provision. The higher education sector has an internationally excellent reputation for quality. The government will continue to closely monitor quality of provision across the sector and reserves the right to reimpose number controls on institutions that expand their student numbers at the expense of quality.
By ‘similar manner’ there, I read the idea that we will see transitional arrangements not unlike the current high grades policy. It would be bizarre to introduce caps for one year, 2014/15, and then immediately lift them.
But it’s good to see quality brought back into the frame. I understand that the current consultation on ‘designation’ status will be withdrawn and replaced with something that looks more closely at institutions. Obviously, several things remain unclear in practical terms (one aside: what happens to HE in FE in the new terrain?), but it does look like the government has learnt some lessons. Still, the regulation needed to effect this change means that private providers will have to sacrifice more independence to preserve access to student support.
Critical Education 