Skip to content

And this is just glorious

Phil Stokoe says:
November 14 2022 at 5:21 pm
Hi Richard,

Thanks to Andrew for actually quoting the System of National Accounts 2008 – but the UK actually follows the European version of the SNA – the European System of Accounts 2010 (predating Brexit of course), for compiling National Accounts (GDP etc) and the Public Sector Finances, and then this guidance is supplemented by the guidance in the accompanying Manual on Government Deficit and Debt 2019 Edition (MGDD).

Rightly or wrongly, the guidance on recording interest on index linked securities, when the index is a general price index like RPI or CPI is clearly set out in paragraph 4.46 (c) (1) which states:

“(1) the amounts of the coupon payments and/or the principal outstanding are linked to a general price index. The change in the value of the principal outstanding between the beginning and the end of a particular accounting period due to the movement in the relevant index is treated as interest accruing in that period, in addition to any interest due for payment in that period;”

There is a longer explanation in MGDD section paragraphs 38-45 which includes a kind of helpful mathematical example as well (though it unhelpfully muddies the waters by adding movements in the market price of the bonds as well)

I think, following this guidance, the ONS is right to record the uplift in principal due to the performance of CPI in a given year as an interest expense in that period.

I’ve weighed in on this before of course, and I understand that Richard disagrees with this, but ONS are genuinely just following the recording rules here, and other countries with index linked bonds in the EU also follow these rules – and then outside the EU, where countries issue index linked bonds, they would be guided by the IMF to follow essentially identical guidance in the Government Finance Statistics Manual 2014 paragraph 6.77.

From a personal point of view, having worked in this area for 13 years now, I don’t have the same issue that you have with this. Unless their are periods of deflation, the costs of redeeming an index linked bond will rise each year, and this rise gets added to the outstanding debt each year, and we record that rise as interest in that period.

Your proposed approach, for me, breaks the accrual principal – we record transactions when the economic event takes place, and the uplift to the amount owed in principal because of, say 8% inflation in 2022 occurs in 2022 and should be recorded in 2022. I don;t see the logic of recording interest in future periods due to movements in the index in 2022. To provide an extreme example – if the index increases by 8% in this year, then stays perfectly flat for the next 10 years, under the approach in the manuals we’d record the 8% interest now. In your approach that interest would spread over the remaining life of the bond, even when the index is unchanged and there is no change in the amount to be repaid at redemption. That feels wrong to me.

So again, I’d just like to note that while you and others disagree with this treatment, and have different views on the correct time of recording (you agree in principle that the eventual increase in principal to be repaid on redemption is interest on the amount originally advanced) it would be helpful if you could recognize that the problem here is not the ONS, or for that matter Treasury and the OBR, but the international organizations and their standard setting committees that have directed that index linked bonds be recorded this way.

Richard Murphy says:
November 14 2022 at 7:46 pm

I am aware of your expertise, and respect it, but you ate relying on three things here

One is the rules say this is right. And they are, a# you note, U.K. rules, and they could be wrong. They can certainly be changed. That is the business of government.

Second, you assume the UN’s flexibility no longer exists, but I question that, and the EU rules definitely need not apply, so this is a UK choice.

Third, as we both know, there are many measures of national debt, so an IMF compliant version can be offered (as a Maastricht one still is) but not be used fur decision making purposes. I think you would agree.

Fourth, accruals requires substance not form to be reflected in accounting, so the charge must be accrued over the life of the bond. Recognising the cost in full when no payment is due is the exact opposite of accruals in this case.

So, I regret I do not agree this time. This is deeply misleading accounting and there is no need for the ONS to use it

The bloke who tries to write international accounting standards insists that following international accounting standards is wrong?

11 thoughts on “And this is just glorious”

  1. He clearly doesn’t understand a word of the preceding paragraphs, but has to defend his position at all costs. Wanker.

  2. I feel sorry for his twin brother – who Moqifen says is actually quite pleasant. He probably agreed to sit Murphy’s accounting exams for him Under duress and now he’s got to observe his brother making a complete fool of himself for posterity to see.

  3. It’s been a while since I was in the business, back when God was an articled clerk, but this:

    Quote: Fourth, accruals requires substance not form to be reflected in accounting, so the charge must be accrued over the life of the bond. Recognising the cost in full when no payment is due is the exact opposite of accruals in this case./quote

    Surely costs must be accounted for as soon as they arise? So, an increase in the value of the bond because of indexation must happen as soon as the indexation is applied, not left until it is redeemed.

    One of us has got this wrong…!

  4. He seems to be arguing there is a divine principle, in addition to the true and fair override, that allows departure from the rules-based standardisation of accounting when Murphy, in his omniscience, determines it should be so.

    An excellent basis for comparability indeed!

    We could call it the Murphy override, or the Half-Baked Potato override?

  5. Nautical Nick

    I think I’ve got it:
    “Recognising the cost in full when no payment is due is the exact opposite of accruals in this case”

    The *whole point* of an accrual is to account for a payment due but not made. That’s literally what accruing a payment means.

    His beef is not with the accrual per se, it’s whether the payment is *due at all*. If the payment isn’t “due”, it doesn’t need to be accrued.

    But he’s wrong either way. 🙂 This has got to be accounted for as occurring in this year. Otherwise you end up having to show the cumulative slices of every previous year (which you will then have to accrue anyway if it’s being added to principal)

  6. Dennis, CPA to the Gods

    The Pedant-General nails it.

    As Phil Stokoe says, “…we record transactions when the economic event takes place…”

    End of discussion.

  7. What’s Murphy saying here? Stokoe seems to imply that Murphy would accrue the current measurement of inflation as if it did not change over the remaining life to maturity.

    Which I think gives the following scenario; the bond is issued with a twenty year life, fixed coupon, but the maturity value is par plus the indexation. In year ten, inflation spikes up 5x, from 2% pa to 10% pa. Murphy would immediately take on to the books that 10%, as 10% every year, over the next ten years?

    Is that right?

  8. Ducky McDuckface:
    No, I think the Spud would increase the coupon over the remaining ten years, so that the accrued obligation at year twenty matched the actual obligation, after all inflation adjustments. I don’t know how Gilt ILBs work, but for US TIPS there is an increase in the actual cash interest paid (flat nominal rate, but the nominal principal increases with underlying inflation – I can’t see how Gilts wouldn’t have an equivalent interest adjustment), so presumably Spud’s stated interest cost would be on top of that. It seems like amortizing a premium or discount – but you would be stacking up a new premium (or discount) every year – seems like an awful lot of work to provide less useful information.

  9. Technically an accrual, in accounting terms, is to recognise costs in the period they relate to…regardless of when cash changes hands.

    You can also have negative accruals -prepayments – where we don’t recognise payments in the books that are made in advance of the period they relate to. Trivial example – an annual subscription is paid in January but we prepay it to adjust the books so that a twelfth of the cost falls every month. That way your month to month profitability measures don’t get distorted.

    Accruals and the point of payment really don’t have to be related at all.

  10. Surely someone should ask him to provide a worked example, taking an issue from 2010, showing the accounting entries he would make, compared with how the markets show it

Leave a Reply

Your email address will not be published. Required fields are marked *