Last Updated: 17 May 2021
The developing scandal over Greece’s foreign currency swaps is another reminder that the data we all rely on when investing can’t necessarily be trusted.
For those who’ve missed the story, the Greek public debt office entered into a complicated currency swap with Goldman Sachs in 2002 in order to reduce its official public debt figures at a time when the country was struggling to comply with the Maastricht treaty debt-to-GDP limit, a prerequisite for joining the euro.
The deal was not secret – Risk magazine reported on it at the time, while also pointing out that Italy had conducted several similar transactions from 1996 onwards. Italy, by the way, was also famous for using hedge fund LTCM to buy its government bonds and ensure “convergence” of the yield with lower-yielding German bonds ahead of the whole monetary union process. But I digress.
The deals with Greece were highly lucrative for Goldman, which earned US$300 million for arranging the swap, according to James Saft of Reuters, and then another US$1 billion underwriting Greek government bond issues over the subsequent eight years, according to Bloomberg. In the prospectuses for the bond issues arranged by the bank there was no mention of the previously arranged currency swap, Bloomberg reports, something that investors in the country’s debt might have considered material.
But what’s new? According to Carmen Reihart and Kenneth Rogoff, who have written the authoritative work on the subject, finding accurate data on countries’ domestic public debt is notoriously difficult. Governments hide liabilities as much as they can and are notorious for conducting off-balance sheet transactions of the type Greece and Goldman cooked up earlier last decade.
If you add back in the off-balance sheet transactions to official statistics of public debt you end up with a very different (far worse) picture of solvency. Dylan Grice of Société Générale has recently produced the chart below, showing that Greek total net liabilities are around 800% of GDP, while the picture is little better across the rest of the EU and in the US.
These figures have been widely circulated and are getting the attention they undoubtedly deserve.
But isn’t there a more general problem here of accounting transparency, one that extends beyond the public sector?
Andrew Lapthorne, also of Société Générale, wrote recently that the gap between US companies’ pro forma reported earnings (which typically omit certain write-offs and non-cash charges) and reported operating earnings is currently at 20-year highs, with the former exceeding the latter by 60%.
And, if earnings figures can’t be trusted, can other accounting metrics like book value (which sums the shareholders’ original capital and the earnings retained over time) be trusted either? Buying stocks that appear cheap in comparison to their book value is one of the key strategies adopted by the father of value investing, Ben Graham, and his subsequent followers like Warren Buffett.
For this reason, it’s worth remembering that index-based investment strategies which use accounting metrics to screen and select stocks cannot be foolproof. At least Research Affiliates’ fundamental indexation strategy (to give one example) uses three other metrics apart from book value to select stocks: sales, dividends and cash flow (which is, in theory, harder to manipulate).
The shenanigans in Greece remind us that while ETFs are a transparent way of investing in financial indices, what underlies the index can be a great deal murkier.