Project Hero and the student loans sell-off – an update and report transcript
By Andrew McGettigan
Earlier this year, False Economy revealed that government advisers had called for either higher interest rates on existing student debt or taxpayer subsidies for investors’ profits in order to help privatise the student loans book.
Drawn up by investment bank Rothschild in 2011 in a secret report called Project Hero, the recommendations were to enable a sale of existing student loan debt starting in 2013 – a timetable that has since slipped.
Both the current government and its predecessor have long wanted to privatise the student loan book – which would mean former students owed their student debt to private investors – but had struggled to come up with a proposal that would be both politically acceptable and sufficiently enticing to potential buyers.
The government has since said it will not go ahead with the inflammatory call to allow repayment terms on existing student debt to be retrospectively raised – but the option of a ‘synthetic hedge’, under which taxpayers would guarantee profits for investors, remains very much on the table.
Today we are publishing a transcript of the Project Hero report – the copy of the report provided by the Department for Business, Innovation and Skills (which commissioned Project Hero) was heavily redacted and difficult to read, so our transcript combines the unredacted material with the swathes of text that were blacked out.
You can download it here (PDF 1.8MB)
Words in square brackets are those we are uncertain of; where we could not read the word at all it is marked with a ‘…’ and charts are not republished. The rest of the text is exactly as Rothschild wrote it. We have transcribed some of the footnotes, although due to the redactions it is not always clear to what text the footnotes are referring.
Things have moved on a bit since then, with the government flogging their remaining holding of the old mortgage-style loans issued before 1998. The sale of the post-1998, pre-Browne loans is currently pencilled to start in 2015, though much uncertainty remains on the detail.
However, certain aspects of Rothschild’s proposals are worth re-examining in closer detail – the rejected models, and the ‘tranching’ proposal that opens the door to subprime student loans.
The rejected models
One of the most interesting aspects of the Rothschild review is the detail provided on the rejected models, not least because of the centrality of the suggestion that universities could be encouraged to underwrite the risk of poor graduate repayments through debt issues or equity stakes in special purpose vehicles.
The aims of the sale are:
- Reduce Public Sector Net Debt
- De-risk the balance sheet
…set against the test of providing ‘value for money’.
This is a relatively straightforward idea: offer incentives for borrowers to make extra repayments on top of the mandatory ones. The idea outlined is a 15% discount offered on an outstanding balance of £10 000. A borrower could pay £8 500 to close the account.
This is a regressive measure offering incentives to those with disposable sums. It would pay down PSND more quickly than the current scheme, but not as quickly as an outright sale. Its value for money test is therefore explicitly tougher: ‘equal to or greater than retention value’.
Extra repayments are welcome but an effective incentive that would mitigate the regressive tendencies and deliver value for money was deemed to be too challenging.
The two other models involve setting up a new company to take over the ‘management’ and, potentially, issuance of student loans.
They are differentiated by the corporate form used: a company limited by guarantee, even a trust, or a company limited by share, potentially publicly listed on the stock exchange. Rothschild’s preference was for the latter.
Company limited by guarantee or trust model
This company would be funded by bonds – ‘debt issued in the capital market’ – or by universities or other private sector bodies: to de-risk the balance sheet the government must move the loans on to the balance sheet of a separate, private entity. Funding through government grants is not to be preferred as it cannot offer ‘value for money’ if annual government expenditure is passed to the ‘trust’.
There are two variants discussed: the ‘cash distribution’ model and the ‘trust’ or ‘Network Rail model’.
“The beneficiaries of the Trust (‘shareholders’) would receive distributions of all excess cash the sold loans pay out over time.”
This is particularly unclear as a company limited by guarantee has no shareholders and cannot distribute dividends. Instead, it is likely that the ‘cash distribution’ model would be funded through bond issues with annual payments linked to loan repayment performance (a variable ‘coupon’) – investors receive payments if repayment performance exceeds the prices paid to purchase loans.
The trust, or ‘Network Rail model’, would instead reinvestment proceeds into financing higher education.
The plan was rejected as it was not clear how to structure the plans in its early days in such a way as to remove government liabilities (‘getting the loans off the balance sheet’). Without initial government support the company would not be seen as robust enough to issue the size of bonds needed to take the loans off the government’s hands.
The Utility model resembles the above suggestions except that the company would issue shares and be publicly listed on the stock exchange. The government would hold a minority stake but otherwise the other shares would be issued in an initial public offering.
It would be subject to ‘economic regulation’ in the sense that the price it paid the government for each year’s set of loans would be determined independently. (The aim would be to have the company eventually take over the ‘origination’ of loans itself).
It would also distribute ‘profits’ to shareholders, making profits insofar as the outlay on new loans was exceeded by repayments.
In the early days, it might take some time to reach this ‘steady-state’ and then move into profit. The OBR estimates that the government will achieve ‘steady state’ on the new loans sometime in the 2030s.
Rothschild believed that this would offer a repeatable model for sale ‘including post-Browne loans’. Crucially, this was ‘the preferred option during most of the first half of 2011’ and thus explains the optimism of the White Paper, which appeared on 28 June 2011.
1.41 We want to find a solution that will manage all current and future ICR loans on an ongoing basis (unlike the one-off sales of the late 1990s). …
1.43 Rothschild is currently finalising the feasibility study and further detail on whether and how Government will proceed with the monetisation of ICR loans will be set out later in the year. [my emphasis]
Rothschild saw the share and dividend structure as a way to ‘incentivise the Utility to work to enhance the [careers] of its borrowers’. The better graduate earnings, the more return to investors.
However, there were clear problems: returns would not materialise for several years and it was unclear how much equity would need to be raised, especially to purchase the £12bn+ of loans to be issued annually from 2014/15. There is also the regulatory risk of operating in a democracy that might choose to move to a different HE funding scheme!
What was most crucial though is managing the decades-long transition to steady-state. Rothschild makes clear, the private sector do not have that kind of risk appetite:
“However the [risk] allocation of the Utility structure involved the private sector taking short term risks (and rewards) on the performance of the portfolio, with Government retaining the longer term risks (and rewards).”
Meaning that the government would continue to cover the longer term risks even with a ‘subsidy and rebate mechanism’. Basically it needed engineering to distribute the annual income pool in such a way as to keep the private sector interested. With the government tied in as backstop, the Office for National Statistics determined that insufficient risk would be transferred to the private sector. The main aim of a sale was stymied.
‘Utility’ was therefore discarded and the government has fallen back on the idea of a ‘one-off’ sale of its existing stock of income contingent repayment loans: precisely what the White Paper said it would not be pursuing.
Rothschild are clear: they would like the case to be reviewed and a way found for ‘universities (in the private sector) [to] take the long term economic risk’. Were this to be packaged up with exemptions from the maximum tuition fee, with some of that income deferred through the ‘Utility’ dividends, then some wealthier universities may well consider such an option (despite the transformations it would effect on academic relations).
Since the government seems to believe that it can move towards a sale of the ‘coalition loans’, this may be a model at the forefront of HE policy in the 2020s.
Tranching – two kinds
Rothschild conclude their report with a new suggestion in relation to all loans: that the government look further into the possibilities offered by ‘tranching’ the loan book in order to attract a better price. In effect, suggesting that the product to be sold is broken down into smaller sizes than whole cohorts. This has significance for assessing whether the much riskier ‘coalition loans’ can be sold (as David Willetts appears to be suggesting as a future funding solution). As each of these cohorts will enter the repayment system with £12billion of debt in total, any feasible sale requires more sophisticated financial engineering.
There are two approaches to tranching outlined by Rothschild: the first relates to a structuring of repayments and risk according to different ‘seniorities’; the second to structuring the product by the ‘underlying characteristics’ of the individual loans themselves. Let’s take them in order.
Tranching by seniority
On the current programme of sales, one purchaser will bid for a whole ‘cohort’ (by year) of loan accounts. With tranching, the income stream associated with that same cohort can be parcelled out into different shares of the income stream. Rothschild discuss two kinds of product: ‘senior’ and ‘equity’/’junior’.
To simplify: Those purchasing senior tranches are paid first from the annual cashflows and may pay a purchase price that reflects a guaranteed minimum annual income. Their investment would be more secure than those who take an equity, or ‘junior’ tranche: these buyers get paid second and therefore run the greater risk of fluctuating annual cashflows reducing their return. They effectively purchase a collateral stake in the cohort – they may also get better returns for their investment but it is riskier.
In a typical Collateralised Debt Obligation, the sponsoring bank would likely take the equity stakes. Here Rothschild is suggesting that the government (or a proxy) play the role of the bank.
With this structure Rothshild was confident that pricing and demand could be improved.
“By tranching the capital structure, we are confident that significant demand exists for the ‘safer’ debt tranches (‘A’ rated or above) at attractive yields.”
The problem would lie with the equity stakes. Would the government have achieved the aims of the sale if it was still exposed to non-repayment risks? Rothschild again:
“However there is significant risk in not finding sufficient buyers for the riskier (‘equity’) tranches at a price that delivers acceptable overall value. If Government were to retain significant amounts of these riskier tranches, limited risk transfer will have taken place and therefore the classification treatment is unlikely to be favourable. If this option was of interest to Government, we would recommend a further round of market testing with potential acquirers of these riskier tranches and further discussions with ONS to confirm the classification treatment.”
The ONS determines whether a sale has moved liabilities and risk off the balance sheet effectively. Rothschild identified the ‘key question’:
“… whether ONS will require a full sale of the junior tranche or, for example, a sale of 51% of that tranche where pricing might be improved in order to achieve overall value for money?”
That is, how much equity risk can the government hold on its books? Would selling over 50% of that product be sufficient to achieve the classification change?
One suggestion is that universities might be encouraged to purchase such equity stakes in graduate debt (it would need the involvement of large funds whether endowment, pension, insurance or hedge funds). As they are classified in the private sector, the government would then have ‘de-risked’ the public sector balance sheet.
Tranching by ‘Underlying Characteristic’
“Tranching each cohort of loans by the characteristics / quality of underlying loans may be deliverable from a classification perspective. However the modelling approach developed (the Hero model) was not designed to value individual loans on this basis. Furthermore, valuing and monetising the loans in this way would require complete and accurate data on individual borrowers”
Here the tranching would not be based on the timing and amount of annual cashflows, but by selling different products built out of particular classes of graduate borrowers.
The jargon hides the obvious idea that Oxbridge or Russell Group graduates, or medicine or STEM graduates, may be more attractive to investors (given the variance in projections for male and female graduates, indirect separation by gender should not be discounted). Note that the ‘complete and accurate data on individual borrowers’ tallies in some ways with the ‘variable human capital’ project being pursued by Neil Shephard and Anna Vignoles. While both the Office for National
Statistics and the National Audit Office have recently produced analysis of graduate earnings and repayment performance which underscores the potential investor value to be had from this form of tranching.
Again improved pricing could be achieved and there may not be an equity stake as such, but this approach may not ‘de-risk’ the government’s balance sheet as the ‘rump’ of non-repayment risk might be left behind – no purchasers may be forthcoming regarding loans with certain ‘characteristics’. Again, universities could also be encouraged to enter the frame: but here they might be buying their own graduates’ debt, rather than taking the equity stakes in a pooled investment.
(Abusive or off-topic comments will be deleted)