Far back in the golden age of finance, interest-rate benchmarks were set by philosopher kings working for the greater good of the markets. Billions of dollars in fines later, we've read of traders comparing themselves to ``a whore's drawers'' and promising bottles of Bollinger in return for rigging Libor, and the sheen has come off somewhat.
That's led to pushes for less coziness. In Australia, the bank bill swap rate, or BBSW, a reference for A$2.37 trillion ($1.7 trillion) in loans, used to be calculated in a similar way to Libor: By asking major banks to submit their assessments of market pricing once a day. In 2013, regulators decided to go for something more transparent and instead take observations of live prices in the market around 10:00 a.m.
Here's what happened to trading around the rate set once dealers knew they were under scrutiny:
In the first 12 months after the Australian Financial Markets Association started compiling the benchmark in 2013, average daily turnover during the set was about A$500 million and volume dropped to zero in one out of 10 days, according to the country's Council of Financial Regulators. This year, turnover has slipped to A$225 million and volume has been zero one-third of the time.
The long-range chart above actually makes things look a bit rosier by taking a 20-day moving average. If you take trading data over the past month, only five days saw so much as a cent of three-month paper -- in theory, the most liquid variety -- traded around the rate set. Trading volumes are now so low, and potentially unrepresentative, that there's a risk investors will stop using the benchmark altogether, Reserve Bank of Australia Assistant Governor Guy Debelle told a Bloomberg Summit in Sydney this week.
Traders' shyness is understandable. The country's markets regulator has been looking into whether the BBSW is being manipulated in the manner of Libor, and has so far extracted fines from RBS, UBS and BNP Paribas. People are reluctant to take an outsized position in the market if it could result in an investigation being unleashed on them. But it's hard not to stand out in a small market like Australia, especially when the country's big four banks are the only ones whose securities are eligible for consideration in the setting. Making matters worse, the more volumes decline, the greater the influence of those still transacting during the rate set, and the greater the risk of being labeled a manipulator.
The country's financial regulators are now looking at reforming their methodology again. One option is to return to a revamped submissions-based system. Another is the hipster option: Get traders to negotiate prices for a bigger chunk of their securities one-to-one rather than with reference to the BBSW, and use that expanded set of small-batch, artisan interest rates as the raw data from which to build a better benchmark. That would slightly defeat the object of having a benchmark in the first place, which is to make things simpler for market participants, but perhaps it will help. Or everyone could just muddle along with the status quo.
The difficulties of balancing relevance and reliability illustrate how collective action problems are baked into the process of coming up with market benchmarks. In an ideal world, the number would be of paramount importance to investors and completely immune to manipulation. Down here on earth, traders are smart. Inevitably, they'll find ways to game any number that's so crucial to their lives.
This column does not necessarily reflect the opinion of Bloomberg LP and its owners.
To contact the author of this story:
David Fickling in Sydney at firstname.lastname@example.org
To contact the editor responsible for this story:
Matthew Brooker at email@example.com