Uncertain liquidity ratios

Like their counterparts elsewhere, South African banks are bracing themselves for a round of changes to Basel II rules. But it is the implications for liquidity and not capital that most concern market participants.

It's been a whirlwind 12 months for the Basel Committee on Banking Supervision. Faced with the wreckage of the financial crisis, regulators published a raft of proposals last December designed to strengthen global capital and liquidity regulations. Across most of the globe, reaction to these changes has focused on the capital implications. In South Africa, though, discord has centred on measures designed to combat liquidity mismatches, with bankers claiming the proposals will be too hard on the country’s banks.

In response to the liquidity squeeze suffered by global financial firms during the crisis, the Basel Committee has proposed two key ratios banks will have to maintain: a liquidity coverage ratio and a net stable funding ratio. The liquidity coverage ratio is designed to ensure banks have enough liquid assets to survive a one-month period of acute market stress, and is measured by dividing a bank’s stock of unencumbered and high-quality liquid assets by its net cash outflows during a 30-day period. Meanwhile, the net stable funding ratio is supposed to address long-term structural liquidity mismatches by establishing a minimum amount of stable funding, based on the liquidity of bank assets over a one-year time period. It is defined as the net amount of stable funding banks have available over one year divided by the net amount of stable funding they require. For both ratios, the proposed minimum is 100%.

While dealers say they agree with the aims of the proposed ratios, the way they are defined will make it difficult for local firms to comply. “The principles of the two ratios are solid and we certainly subscribe to them. However, if you look into the current way the two ratios are written into the consultative paper, they are very punitive on local banks,” says Mike Davis, general manager of group asset-liability management and capital management at Nedbank in Johannesburg.

Due to the small size of South Africa’s bond market, the rules would effectively force local banks to hold a far larger amount of government paper than banks in Europe and the US. This is because only cash reserves and government debt are likely to qualify as high-quality liquid assets, Davis says. On the funding side, the rules allow banks to split their deposits into stable and less stable categories. But for retail deposits to qualify as stable they have to be backed by a deposit insurance scheme. This would also make it harder for South African banks to meet the requirements, he adds: “The South African market does not have a deposit insurance scheme, and as a result we would not benefit from being able to split retail deposits into the stable funding category.”

The biggest problem with both rules is the structural liquidity mismatch of local banks. South African banks get up to 40% of their funding from wholesale sources, predominantly money-market funds, says Voyt Krzychylkiewicz, financial analyst at Deutsche Bank in Johannesburg. At the same time, the assets of local banks are typically fairly long-dated – approximately 75% of retail loans are mortgages, he adds. Despite the fact local banks held up relatively well during recent havoc in global financial markets, this imbalance would cause South African banks to look shaky if viewed through the lens of both the proposed ratios. “The interbank market here is actually very stable. But the way these ratios are calculated means South African banks will come across quite badly,” he says.

Issuing longer-dated paper might seem like a solution, but South African money-market funds are not allowed to buy instruments with maturities longer than 12 months and are obliged to maintain an average maturity of no longer than 90 days. Krzychylkiewicz explains: “Effectively, South African banks get a large proportion of their funding from these entities that can’t buy instruments longer than 12 months and they can’t have a longer average maturity than 90 days at any point in time.”

In a research note on January 20, Krzychylkiewicz estimates South African banks would score approximately 60% on the Basel net stable funding ratio and 40% on the liquidity coverage ratio. Based on the liquidity coverage ratio, local banks are around 480 billion–600 billion rand ($62.12 billion–77.65 billion) short of liquid assets, he says. Adopting the framework would increase collective funding costs for South African banks by 10 billion–24 billion rand a year, he claims.

Local dealers say complying with the rules would be a tall order. Even if South Africa’s retail deposits could be classified as stable, the lack of a sizeable middle class in the country makes it difficult to attract deposits in enough size to make a difference. Raising large amounts of long-dated funding in the local market could be almost as challenging, suggest dealers.

Andries du Toit, Johannesburg-based group treasurer at FirstRand, believes South African banks would not be able to comply with both ratios without structural reform, requiring changes by policy-makers, rather than banks: “The supply side of money within South Africa is a structural problem. You’re going to need a reform on the savings side and on what kind of instruments funds can hold,” he says.

The effect of the liquidity ratios and the other proposals outlined in December is due to be assessed by banks during the first quarter of this year, with the results of the impact study submitted to national regulators. Sources say the South African Reserve Bank has requested local banks to submit estimates and data on the impact of the liquidity ratios in advance of the other components of the tests. While the central bank has typically remained faithful to the letter of the Basel Committee’s rules when implementing them in the local market, there’s speculation the regulations could be tweaked if the proposal passes in its present form.

In an emailed statement to Risk, the Reserve Bank said it was far too early to comment on the effect of the December proposals with any degree of accuracy: “In general, we think South African banks will fare comfortably with the capital guidelines, while the liquidity guidelines might be somewhat more difficult.” However, a number of other countries were in a similar position to South Africa regarding the liquidity guidelines, it claimed. While the Reserve Bank would always enhance regulation where appropriate, it would not blindly follow regulation designed for developed countries, the statement said. Instead, the central bank would “ensure the proposals are indeed appropriate for our system and add value in terms of enhanced risk management without being overkill and a business inhibitor”.

Deutsche’s Krzychylkiewicz says one idea would be to reclassify wholesale funding from money-market funds as more stable. “Wholesale funding is effectively as sticky as retail funding in South Africa. Assuming 100% of wholesale maturing in the next 30 days is to leave a bank is a very onerous thing to apply here because it’s extremely unlikely to happen,” he says. Due to foreign exchange controls, a flight of liquidity from the local market is not a plausible scenario, say market participants.

Meanwhile, Davis at Nedbank suggests haircutting the requirements to account for the structural mismatches facing South African banks. “If South African banks have to buy a huge chunk of government bonds to comply with the ratios and it is practically not possible to do so, then we’re not going to be able to comply, and I don’t believe that’s the intention of the committee,” he says.

The rest of the Basel Committee’s proposals are less controversial in South Africa. The changes started last July, with finalised rules for bank trading books introducing a number of new components to the capital calculation. In particular, regulators focused on beefing up the capital held against credit risk-bearing instruments through the incremental risk charge (IRC) and a new regime for securitisations and resecuritisations. Overlaying these elements is a requirement to work out how value-at-risk figures could change in malfunctioning markets – the so-called stressed VAR measure.

South African banks should be fairly well insulated from the trading book changes, largely because they tend to have fairly conservative trading businesses, say dealers. “The amount of prop trading and market risk at South African banks is a lot smaller than our global counterparts. If you look at our balance sheet, it would suggest we are a vanilla bank. So we think the consequences would be moderate from a capital perspective,” says Nedbank’s Davis.

As a result of the IRC changes, the International Swaps and Derivatives Association has claimed global dealers would have to increase trading book capital by three times, with the effect on some firms expected to be even higher. South African banks should escape that kind of hit, says Davis. “A number of organisations globally will see some horrible numbers. But if you look at South African banks, we’re involved in vanilla lending while prop trading is only done to facilitate client flows,” he asserts.

Bigger questions hang over the December package, which includes the hated liquidity rules. Those proposals also seek to raise the quality of bank capital – for example, by disqualifying minority shareholdings and deferred tax assets as contributors to Tier I capital. In addition, the proposals seek to introduce a leverage ratio and tame counterparty risk by raising capital charges for counterparty credit exposures. By applying a multiplier of 1.25 to the asset value correlation of financial firms and a zero risk-weight for qualifying central counterparties, the rules effectively attempt to push over-the-counter derivatives towards central clearing. While the committee hasn’t given up on the idea of a counter-cyclical capital buffer, the December package contained no specifics on methodology, and a separate proposal could be ready for publication around the middle of 2010.

Banks everywhere have been bracing themselves for the changes, and South African banks are no different in that respect, says Anthony Walker, senior director for financial institutions at Fitch Ratings in Johannesburg. “Most of the South African banks were raising capital last year. That was in response to the economic environment and global crisis, but also in anticipation of regulatory reform,” he says.

Here, the outlook is a mix of good and bad. Following the release of the December proposals, analysts at Credit Suisse looked at how South African banks would be affected by the proposals on capital quality. Due to their low leverage and scant use of hybrid capital, the research concluded the “healthy capital buffers” of South African banks would not be compromised. But some of the changes would have a material impact – in particular, the exclusion of minority interests as a component of Tier I capital.

The proposals mean that if a bank owns 90% of a foreign firm, for example, the remaining 10% cannot be used to absorb losses – a measure that could be problematic for local banks with African subsidiaries, such as Standard Bank, says Walker. “Standard Bank would be most affected by that, because it owns several institutions in Africa where it has a majority stake,” he says.

According to Credit Suisse, minority interests represented 10.4%, 4.2%, 3.8% and 2.71% of equity Tier I capital for Standard Bank, Nedbank, FirstRand and Absa, respectively, as of June 30, 2009. The elimination of deferred tax assets from regulatory capital would also have a modest impact, with analysts pointing to FirstRand in particular. The bank’s deductions would be roughly equivalent to 3.8% of equity Tier I capital as of June 30, 2009, according to Credit Suisse. For Standard Bank, Absa and Nedbank, the deductions would be 1.5%, 0.85% and 0.6%, respectively.

At the moment, Basel II allows banks to make some deductions from regulatory capital on a 50:50 basis between Tier I and Tier II. The removal of this provision  – a measure that also looks likely – would erode 7.3%, 5%, 2.8% and 1.3% of equity Tier I capital for Standard Bank, Absa, Nedbank and FirstRand, respectively, as of June 30, 2009.

When everything is stacked up, the proposals seem likely to cause the capital ratios of South Africa’s five largest banks to decline, while simultaneously increasing the amount of capital they have to hold. But it is hard to know exactly how big the final effect will be, says Fitch’s Walker. In Europe and the US, the trading book rules and December proposals were greeted with dismay. There, bankers complain the counterparty risk changes, for instance, could double the amount of capital needed to support their exposures (Risk January 2010, page 8).

The impact on South African banks could be more balanced. For a start, they may be better placed to deal with an increase in capital requirements thanks to low levels of leverage compared with banks elsewhere. “If you compare their gearing levels with some of the international institutions, they are fairly conservative,” says Walker.

South African banks are already subject to a minimum core Tier I capital ratio of 5.25%, which compares favourably with the 4% minimum typical elsewhere. In any case, the largest South African banks sit comfortably above the regulatory minimum, analysts note. And their capital ratios may well be reinforced over the coming years if the local economy recovers slowly from the global financial crisis and lending volumes remain depressed.