Consolidation and Lehmans
Posted by Christie Malry on March 17, 2010 at 11:19 pm
Ritchie has a theory about Lehmans and how it all happened:
I’ve been doing a little grubbing around Lehman and its repo accounting.
Report is widespread that the “repo 105” deal that caused the alleged misstatement of Lehman’s balance sheet took place in the UK.
Now it so happens that my information is that Lehman UK reported using UK GAAP i.e. under UK accounting standards.
But, under UK accounting standards a repo is treated as an on balance sheet transaction.After all, it is a sale of an asset matched by an agreement to repurchase it. The transaction has not changed the real structure of the balance sheet as far as UK GAAP is concerned, so the transaction is effectively ignored for reporting purposes. That’s because the risk and reward of ownership of the underlying assets has not changed – they remained with Lehman.
Oddly, as far as I can tell IFRS delivers the same result as far as I can see.
So how come Lehman reported with the benefit of the repos being reflected in the consolidated accounts?
The answer is that US GAAP does not reflect the risk and reward model. It simply asks if there is a sale and then applies a hurdle test. As Prof Mike Page of Portsmouth University wrote on the blog this morning:
Further research on US blogs seems to suggest that SFAS 140 was crucial to scheme. I have, as yet, not been able to track down a full text of it but other comment suggests there may be a ‘bright line’ that allows a Repo to be treated as a sale if it is for more that 105% of current market value. The exact conditions aren’t set out in online summaries. Given the size of the law suits, what is in the UK (and doubtless some offshore) subsidiaries financial statements may be vital to what US investors might have known about the transactions. I think this is going to be one of the most fascinating accounting cases for a long time. As Richard pointed out, the role of fair-value is likely to be important as the notion of ‘market value’ must have been pretty elastic for a lot of the stuff that was apparently put through these Repo transactions.
That’s the hurdle.
And it seems that the Lehman “repo 105” might have been named in honour of that hurdle – over which it jumped.
Some – such as the FT – have suggested that opinions on whether the hurdle had been jumped could be obtained in the UK, but not in the US. Maybe that’s true. But maybe also it was not essential.
The reason is that the resulting misstatement from use of repo 105 will, if I’m right, not be found in the accounts of any of the underlying regulated entities within the Lehman group – all of which were stated in accordance with required GAAP. Rather the misstatement was only to be found in those rather vague entries in the books of account of a group of companies called the consolidation journals.
This takes some getting one’s head round for the lay reader. Basically you have to understand that the glossy published accounts of a group of companies – most especially a multinational corporation – are a work of fiction. As my friend Prem Sikka has often said, most should qualify for the Booker Prize. That’s because of three things:
- There is no entity anywhere that undertakes the transactions recorded in these accounts. The transactions are actually undertaken by a range of other entities – maybe thousands of other entities, which are then added together.
- Except they’re not added together. All the intra-group balances are excluded and all the intra-group trades are excluded. This would not be worrying until you realise that between 60% and 70% of world trade is intra-group – and none of it appears in the consolidated accounts of multinational corporations – which makes clear how misleading they are.
- There is also the possibility that profits and losses can be recorded in the consolidation alone – and balance sheets changed in the process – with the resulting transactions never appearing in the books of any actual company and therefore beyond the reach of taxation and regulation. These are the secretive ‘consolidation journals’. In the wrong hands these provide massive opportunity for abuse.
The real question about Lehman then is not just where did the abuse happen that helped bring it down – but did it happen in the ether of the consolidation alone – in the pure make believe world unrelated to reality that is a set of books and records to which no company lays claim in its own books and records?
I think that might be true.
In which case three things follow:
- We need country-by-country reporting which requires the accounts of an multinational corporation to be grounded in a place – not floating in an ether above it;
- We need, as I have suggested, to review the whole nature of limited liability reporting and what it means in groups;
- we need to review accounting even more urgently.
And, finally, someone needs to work out how to prosecute fraud never recorded in the books of a company but nonetheless reported upon in a set of accounts. There’s a challenge, if this is true.
I'm afraid his thesis is flawed. This isn't the way big group consolidations and big group audits work.
Lehmans UK didn't submit in UK GAAP
A big consolidation is based around the pack. In the old days, the pack was a proforma that all reporting units filled in and sent back to head office. Each pack was certified by the local director. Head office would take all the packs, add them all up, make the necessary consolidation entries and that would produce the group results. These days it's probably a computer file or submitted via an accounting IT system, but the principle is the same.
In all my years of auditing, I have never seen a pack submitted to head office in anything other than company GAAP. This is obvious. Head office don't want to be having to work out what adjustments they need to make to produce their consolidated results. They make their subsidiaries do all the work. It's just about possible that a local unit might produce two packs: a local GAAP set and a set of adjustments to company GAAP. What gets submitted, and what gets audited by the local auditors will be in company GAAP.
Because of this, I think it's extremely unlikely that Lehmans UK submitted a pack to New York that was in UK GAAP.
Murphy's analysis of US GAAP is wrong
If your constitution is strong enough, you can read SFAS No. 140 in full here. The analysis in the examiner's report is very good. The tests to determine whether sale treatment is appropriate for a repo are basically the right tests. However there's a twist in the end, which I'll try to explain here.
Repo transactions are sort of like a pawnbroker (sort of... bear with me, I'll explain). Let's say you want a loan and the pawnbroker is your only option. You don't actually want to get rid of your Rolex; you're just using it as a way of proving to the pawnbroker that you're serious about your commitment to repay the loan. If all goes well, you repay the loan and interest and get your Rolex back.
A traditional repo is the same. A company wants cash, and it gives some assets to another party to prove it's serious about repaying. If all goes well, it repays the loan and interest and gets its assets back. The borrowing company has to put up assets that are worth more than the amount it's borrowing. This difference is called the haircut. A typical haircut is fairly small for big, reputable companies - perhaps 2%. For each $100 you want to borrow, you must put up $102 of assets.
[This is why the pawnbroker example isn't perfect - a pawnbroker will typically require a much bigger 'haircut' from its borrowers than a typical repo counterparty will.]
In accounting, there are two possible treatments. Either we follow the legal form and treat it as a sale and buyback. Or we look to the substance of the transaction and treat it as if the borrowing company never got rid of the assets at all; it just borrowed some cash and paid it back. Most repos are of this latter type.
There are some tests in SFAS No. 140 to determine whether the sale treatment is appropriate. And the tests are pretty sensible. From para. 9 (emphasis added):
The transferor has surrendered control over transferred assets if and only if all of the following conditions are met:
a. The transferred assets have been isolated from the transferor—put presumptively beyond the reach of the transferor and its creditors, even in bankruptcy or other receivership.
b. Each transferee has the right to pledge or exchange the assets (or beneficial interests) it received, and no condition both constrains the transferee (or holder) from taking advantage of its right to pledge or exchange and provides more than a trivial benefit to the transferor.
c. The transferor does not maintain effective control over the transferred assets through either
(1) an agreement that both entitles and obligates the transferor to repurchase or redeem them before their maturity or
(2) the ability to unilaterally cause the holder to return specific assets, other than through a cleanup call.
The examiner's report also focuses on the crucial paragraph 218 (emphasis added):
218. The Board also decided that the transferor's right to repurchase is not assured unless it is protected by obtaining collateral sufficient to fund substantially all of the cost of purchasing identical replacement securities during the term of the contract so that it has received the means to replace the assets even if the transferee defaults. Judgment is needed to interpret the term substantially all and other aspects of the criterion that the terms of a repurchase agreement do not maintain effective control over the transferred asset. However, arrangements to repurchase or lend readily obtainable securities, typically with as much as 98 percent collateralization (for entities agreeing to repurchase) or as little as 102 percent overcollateralization (for securities lenders), valued daily and adjusted up or down frequently for changes in the market price of the security transferred and with clear powers to use that collateral quickly in the event of default, typically fall clearly within that guideline. The Board believes that other collateral arrangements typically fall well outside that guideline.
This is the crucial paragraph that explains why Repo 105 worked, at least according to Lehmans management. Think about it. Lehmans had put up securities that had a market value of $52.5 billion. They had only received cash worth $50 billion. If their counterparties had gone bust, Lehmans did not have enough cash to buy equivalent securities in the open market. They would have been 5% out of pocket. Where the assets are at risk beyond a minimal amount, US GAAP says it should be treated as a sale.
Lehmans needed a legal opinion that sale treatment was appropriate, and none of the US law firms were prepared to give this opinion, given the complexity of the judgements required and, presumably, the bad press if they got it wrong.
The principle is right, even if the threshold between sale and non-sale treatment doesn't look right in the case of Lehmans. Imagine you went to a pawnbroker with your Rolex again. If you only borrow £10 using your expensive collateral, it can be argued you have lost control of your asset. If the watch is stolen, you would be unable to buy it back. Having such an asset on your balance sheet in these circumstances might be misleading.
Consolidation ain't complicated
Consolidation is like a jigsaw puzzle. The company has a nice straight edge it presents to the outside world. Meantime, you don't have to worry about the jagged shapes of the pieces inside; you just focus on the big picture.
It's obvious that you exclude intra-group trades and balances. My brother might owe me £1,000 even while we, as a family, have no external borrowings. It would be hopelessly misleading to describe our family arrangements as having any external borrowings in this case.
It's true that profits and losses may be recorded in the consolidation alone, or balance sheets moved. And auditors should audit the consolidation process. On audits I managed, we spent a lot of time understanding how the consolidation built up. Every adjusting entry was examined and understood. You might sample routine transactions but you will never sample consolidation journal entries. You just have to test the lot.
Conclusion
It's extremely unlikely that Lehmans UK recorded the Repo 105 transactions as a non-sale in the UK pack submitted to head office. They would have recorded them under Lehmans GAAP as sale transactions. Accordingly there would have been no need for consolidation or other journal entries to adjust them at head office level.
Even if there had, they should have been audited by the head office audit team in any case.
Ritchie's thesis is a complete non-starter.
. And, as it happens, I did see some shocking examples of bad accounting (a story for another day, perhaps). But I'm certain that my team would have expected - and would have been expected - to identify a case of window-dressing that was as egregious as the Lehmans report makes out.
