What's the harm? | Fin24
  • Moving Target

    What if FlySafair were to run embattled SAA? asks Ferial Haffajee.

  • Air Quality

    Activists acknowledge their court action is a 'baptism of fire' for Barbara Creecy.

  • Fin24’s newsletter

    Sign up to receive Fin24's top news in your inbox every morning.


What's the harm?

Feb 19 2017 06:02
Dewald Van Rensburg

Johannesburg - For all the outrage directed at banks this week, it is difficult to ascertain the exact damage that collusion among currency traders would have done – and to whom.

The banking industry has been under political attack from an array of parties, ranging from government – in the form of President Jacob Zuma and Minister of Mineral Resources Mosebenzi Zwane – to the Gupta family and the Progressive Professionals’ Forum.

READ: Zwane feels vindicated by bank collusion probe

This week, trade unions and political parties – notably, the ANC and the Economic Freedom Fighters – issued uniformly scathing attacks on the banks, after allegations of price-fixing and other irregularities were made by the Competition Commission.

The collusion case, involving 17 banks – three of which are major local banks Absa, Investec and Standard Bank – will involve a massive burden of proof to show exactly what damage was inflicted between 2007 and 2013.

READ: How banks allegedly colluded on currency trades

But experts and one insider say the damage is likely to be a fraction of what South Africans seem to believe.

On Friday, the Competition Tribunal released the full referral it had received from the Competition Commission, spelling out the accusations in detail and naming the individuals involved at each bank in each instance of alleged collusion.

The referral suggests that the cooperating banks and traders have provided records of their communications with traders at other banks to fix prices and divide the market.

There are no allegations that do not involve Absa, suggesting that the bank’s records have been central to the case being made.

The commission is also relying on foreign instances, where guilt has already been established, to support a substantial part of its local investigation.

A single trader, American Jason Katz, appears to be the sole reason for at least two of the banks appearing in the local probe.

Katz is cited as having taken part in the collusion on rand exchanges when he worked for BNP Paribas SA as well as at Standard New York (a subsidiary of Standard Bank Group) and Barclays.

Last month, in a New York court, he pleaded guilty to participating in a conspiracy to manipulate emerging market currency trades, and was subsequently barred from participating in the banking industry.

Following suit a week later was Christopher Cummins of Citibank. He is another US trader cited in the commission’s case who pleaded guilty to manipulating exchange rates.

READ: Absa let off collusion fine as it helped regulators - source

Citibank has applied for leniency in South Africa and will be cooperating with the commission alongside Absa, its parent company Barclays and fellow subsidiary Barclays Capital.

The man most cited for colluding at Absa is Duncan Howes, who was suspended in 2015, according to Bloomberg reports.

The commission is not seeking any penalties against Absa, Barclays, Barclays Capital or Citibank. It is, however, pursuing the maximum 10% annual turnover fine against all the other banks, which would amount to an astronomical sum.

The damage done

It is impossible that the collusion alleged in this case could have caused the rand to trade at a meaningfully different value against the dollar than it would have anyway, said a source with intimate knowledge of the matter.

“The client does not get harmed in the way people think. He would not have paid a different exchange rate; he would simply have got less of the discount that large clients usually get,” said the source.

“The volumes involved are not enough to push the rand. The harm is not on the exchange rate we all see.

“Bonuses are the driver.”

READ: Price fixing: Rogue traders or bank collusion?

Theoretically, collusion could have seen banks’ clients either over- or underpaying for the rand.

The referral by the commission accuses the traders of moving the rand in both directions.

More importantly, the actual scale of their fraud would likely have been tiny, compared with the market for rand trading.

Exchanges of rands for dollars amount to R51 billion a day, according to the affidavit supplied this week by the commission’s Mfundo Ngobese, who heads its cartel division.

“It [the case] will take years and the people involved will be gone by the time it concludes, said Ian Cruickshanks, chief economist at the SA Institute for Race Relations and a former veteran of Nedbank Capital.

“You get these traders who get bonuses based on their own trading profit, so maybe they take a chance. I don’t know, but wherever you have a market, you will have someone trying to beat it.”

Cruickshanks also expressed scepticism regarding how much influence colluding traders could have on the rand.

“The primary factor governing the value of the rand is still confidence,” he said.

Stuart Theobald, chair of economic research agency Intellidex, also noted that “it [traders’ influence] is only ever going to be a temporary move”.

“One thing the Competition Commission is not saying is that the absolute level of the rand was manipulated,” he wrote on the website of his company, which has a reputable track record in financial and competition research.

The insider source emphasised that only a small number of individuals were culpable.

“Traders get paid their bonuses in March and you see a lot of staff turnover immediately afterwards. Maybe someone is unhappy with the R2.5 million bonus they got, and they sign up somewhere else for another R2.5 million sign-up bonus,” said the source.

“The bonuses are premised on volumes. Volumes are what matter. If I have high volumes and lots of volatility, that is where collusion comes in.”

The commission alleges five different types of collusion, amounting to different forms of price-fixing and market division.

The purpose of price-fixing would ultimately be to widen the gap between the price at which the traders buy a currency and the price at which they sell it.

READ: Treasury touts Twin Peaks bill amid banking scandal

Apart from raising profit this way, colluders would also want to ensure certain volumes go to each of them, which describes market division.

Traders would direct big trades among themselves by pulling out of the trading platform they used at the right time, or by agreeing that only one of them would offer a slightly better rate.

“The client feels happy because they got a discount, but actually no one was competing in the first place,” explained the source.

Of the 17 banks being probed, only eight are so-called authorised dealers in South Africa – regulated by local authorities.

The altogether 25 authorised dealers in South Africa had already been subject to a review by a Foreign Exchange Review Commission (FXRC), set up by the SA Reserve Bank and the Financial Services Board in 2014.

The FXRC issued a report in October 2015, saying it found “some evidence of inappropriate sharing of confidential client information”, but also found “no evidence of malpractice or serious misconduct”.

It concluded that the local foreign exchange market was competitive and that banks were kept honest by “multibanked corporate clients”.

About 60% of the total trade in the rand takes place outside South Africa.

The terms of reference for the FXRC noted in 2014 that the rand was especially vulnerable to manipulation as it traded globally, but without having the depth and liquidity of the major international currencies.

The rand was “able to be moved by large players in any particular direction”, it said.

The clients of currency traders potentially affected by collusion are overwhelmingly non-resident.

What would a 10% fine be?

Somewhere along the line, the commission and the banks are going to argue about what harm the collusion actually caused. This will, however, only determine one part of the penalty, if there is one.

The extent to which the accused cooperate also weighs on that decision.

“Quantum [the total figure] is usually an estimate. Often it is a ballpark figure. What will happen is that they [the banks and the commission] will have rival economists make cases,” said a source familiar with the workings of the competition authorities.

It comes down to theory as much as hard measurement.

If the commission is arguing that rand/dollar exchanges were manipulated extensively enough to have undervalued the currency in a long-term and meaningful way, a massive claim could amount – tied to the supposed overpayment for imports into South Africa.

The 10% fine, which the commission is pursuing, is highly unlikely to be imposed, not least because not all the accused participated in the alleged collusion to the same extent.

Even if the maximum fine is imposed, it could be relatively small. This explains the non-reaction by banks’ share prices to the commission’s referral this week.

In the theoretical case against Absa, the trading division implicated is responsible for 15% of the bank’s non-interest revenue – which totalled about R3.5 billion in 2012.

The bank’s trading pool consists of about 60 people, of which five, at most, are implicated.

If Absa was not already indemnified by its cooperation with the commission, it could very easily argue that the fine should be on the revenue of the implicated traders, meaning 10% of a fraction of the R3.5 billion.

Even the most stringent fine would then amount to only R29 million.

Standard Bank had a trading revenue of R8.8 billion in 2012, and only one accused employee, so a similarly small fine could be argued for, even at 10%.

Even in relatively simple industries such as cement and construction, rival economists have come up with wildly different estimates of the damage done by cartels.

  • This article was amended on February 23 2017 to reflect the correct relationship between Standard New York and the Standard Bank Group. In the initial article, we stated that Standard New York was unrelated to Standard Bank.
Read Fin24's top stories trending on Twitter:

forex  |  banks  |  price fixing  |  collusion  |  currencies


Read Fin24’s Comments Policy

24.com publishes all comments posted on articles provided that they adhere to our Comments Policy. Should you wish to report a comment for editorial review, please do so by clicking the 'Report Comment' button to the right of each comment.

Comment on this story
Add your comment
Comment 0 characters remaining

Company Snapshot

Money Clinic

Money Clinic
Do you have a question about your finances? We'll get an expert opinion.
Click here...

Voting Booth

SARS has lifted the tax return threshold from R350 000 to R500 000 per year. What are your thoughts on this??

Previous results · Suggest a vote