Securitisation structure & loan sale value for money (2 of 5)

Kit Malthouse: Just on the sale of the book, why would anyone want to buy it? If you think about it from a commercial point of view, you have this large asset, which is subject to the vagaries of the economy; people’s ability to repay is down to their employment prospects, and so you are going to see economic variability on it. They already know that there is going to be a fairly big chunk that will not be repaid.

Joseph Johnson: The facts speak for themselves. There was a very large pool of investors who achieved a competitive price that delivered value for money for the Government for this asset.

Kit Malthouse: Okay. Basically the reason they wanted to buy it is because the price was competitive, in your words, or cheap.

Joseph Johnson: We achieved fair value.

Kit Malthouse: It was cheap.

Joseph Johnson: We believe we achieved at least fair value, otherwise we would not have sold the asset. It passed all the relevant tests that the Treasury requires departments to take when selling assets.

Kit Malthouse: Were you not slightly alarmed that people were so keen to buy it, given the pricing of it?

Joseph Johnson: You seem to be making two different arguments. Your first argument was: why would anybody be interested in it? Now you are arguing there was such interest because we were under-pricing it.

Kit Malthouse: No, I said why would anybody be interested, and you mentioned competitive pricing, which I translated as cheap—the price would be such that it stimulated a lot of people wanting to buy it.

Joseph Johnson: We believe that we achieved at least fair value for the asset and that it cleared the tests that the Treasury requires departments to have for the sale of assets.

The above exchange is taken from the transcript of Jo Johnson’s appearance before the Treasury Select Committee on 19 December.

Did the government sell it cheap or did it achieve value for money? It’s not an either/or. As noted in the first post in this series, the government raised £1.7bn from a securitisation of income contingent repayment loans with a face value of £3.7bn.

For the purposes of assessing asset sales, the government uses what it calls the ‘Green Book’ framework to ‘test’ value for money. The government believes that no policy is advanced by continuing to hold student loans, so the decision before them is whether it makes financial sense to sell the assets. It needs to assess how to value certain cash today against less certain repayments spread over a decade and more.

Matt Toombs, Director Student Finance and Analysis at the Department for Education, appeared alongside Johnson at December’s Treasury Select Committee hearing. He explained the approach:

The assessment of value for money involved looking at the alternative uses the Government could make of the money that was held within those assets if it was invested elsewhere. That is why they looked at the Green Book value-for-money framework in assessing whether they could achieve value in selling the loans.

… The Green Book looks at the assessment of the overall value for money for what you receive relative to the face value. It looks at the economic opportunity cost applied to the whole asset. It is a comparison of the kind of price you receive against the value to the Government of holding on to that asset.

What you won’t hear when any official explains the workings of the “Green Book” test is that it involves using what’s called the “social time preference” discount rate, which has been unchanged for a decade. That discount rate is set at RPI plus 3.5%. For the sale of financial assets, one percentage point is deducted to represent the removal of “catastrophe risk” and then a risk factor pertinent to the asset is added. In theory, the discount rate for the loan sale would have as its minimum, RPI plus 2.5%, though in practice it was probably closer to RPI plus 3%.

This is much higher than the financial reporting discount rate used to value the loans in the government’s accounts. That is only RPI plus 0.7%. A higher discount rate means that less value is placed on future cash. (Note that the financial reporting rate was recently reduced from RPI plus 2.2% to better reflect the government’s cost of borrowing. No such change has been made to the Green Book rate, a cornerstone of austerity theology. The alternative use envisioned for the cash received is likely to be reducing the government’s gilt issuance this year).

The same cash stream – project future repayments – is valued in two different ways. At RPI of 3.9%, £1 pound next year would be worth 96p today when using the financial reporting rate, but only 94p to even the minimum Green Book rate. The test is therefore skewed so that a loss-making sale would pass the VfM test.  For the loans just sold, the discount rate “spread” has a greater effect, as these loans are only written off when the borrower turns 65. DfE expects these loans to last until the late 2030’s.

Even so, passing this test took some effort. When the government restarted the sale process at the end of October, the accompanying press release read:

For the first time, Government is structuring a sale of student loans as a securitisation comprising four tranches of notes of varying seniority and maturity from senior, investment grade rated tranches to an unrated tranche; as such, it is expected to appeal to a wide range of different investor groups including pension funds, insurers and asset managers.

… This securitisation structure will enable the Government to maximise value from the sale for the UK taxpayer by creating separate tranches of securities attractive to a wide-range of potential investors, including, for example, pension funds and annuity providers looking to match index-linked liabilities, through to insurers and asset managers.

Of the four tranches, the two most “senior” achieved investment ratings of A from Fitch and Standard & Poor’s. A third tranche, with far fewer notes, achieved B from both. While the unrated tranche represented over half of the principal raised. This represents an unusually high share in a securitisation for the risky, junior tranche, but illustrates the unusual nature of income contingent loans and the associated risks of non-repayment.

• Class A1 are the most senior notes: holders are paid first from available funds. They were priced at 99% of par and represented 23% of the offering. These are “pass-through” notes with a coupon of 1% above one-year sterling LIBOR. The pass-through structure means that purchasers receive cash as and when it is available until the principal is repaid.  It might help to think of this tranche as the right to first use of a cashflow using a bucket that can be filled once.  They are expected to be short-lived, but do not have a fixed maturity date as that depends on the short-term cashflows.
• Class A2 notes were priced at 93% of its denomination value and accounted for 20% of the deal. These securities have a defined schedule of payments and a fixed coupon of 2.5% of the face value (the Treasury is currently issuing 10-year gilts at around 1.3%). They face some risk that payments will not materialise if there are insufficient repayments made by borrowers. These notes provide a stable long-term return.
• Class B looks to me like the means to manage the risks associated with the interest rate terms for student loans. The notes represent under 4% of the deal with coupons based on Limited Price Index, which is RPI but with a minimum (0%) and a maximum of 5%. Borrowers get 1.45% on top of LPI. These are also pass-through notes that get paid when funds are available.
• The unrated tranche, Class X, padded out the deal and was priced at the very low mark of 8.5% of face value. These notes only get paid with whatever is left after the other noteholders have been paid each year. They represent a big buffer for the more senior tranches against the risk of loan non-repayment. Holders might get very little back, though if repayments are higher than predicted they get those surpluses. They have a coupon with a nominal value of 0.5%, but may not see any repayments for a while, if at all.

You can see how this complex structure addressed the issues with a sale. Most of the cash raised came from the sale of senior tranches enabling the VfM test to be passed, while a cheap sale of the risky junior tranches enabled the government to shift its risk. It didn’t want to be left holding any risk or the sale might not be classified as such (more on this in Part 3 when we turn to the classification decisions facing the ONS).

All tranches were oversubscribed. But Class A2 – which looks very much like a government bond but with more return for investors – and the unrated Class X saw twice as much demand as notes available. Make of that what you will.