Many of the fixed interest investments, conduits and structures on offer, including the low cost housing securitisation issue from Standard Bank – the Siyakha Fund, are simply too risky for the level of yield offered, according to two local asset managers.
In terms of rating the Siyakha Fund, Fitch Ratings did this on behalf of Standard Bank, and because it was a private rating, the technical report into the offering, known as a pre-sale report, is not available for public consumption.
According to Tertius Smith at Fitch Ratings, the rating took approximately two months to complete, and the final rating was provided in early December, and was paid for by Standard Bank.
Smith says that while a public report wasn’t published, a ratings letter was presented to Standard Bank, and these ratings were discussed with the three parties involved in the transaction, although they wouldn’t have had sight of any documentation.
The three parties were Investec Bank, who joined Liberty Life and Old Mutual and invested R800m each into the R2.4bn low cost housing home loan securitisation private placement arranged by Standard Bank. According to Smith the R2.4bn issue had a spread of rated bond tranches ranging from AAA to BB. He went on to say that the structure required more credit enhancement because of the nature of the underlying mortgages.
“This wasn’t a normal Residential Mortgage Backed Security (RMBS) transaction. While not sub prime, the type of borrower is different and the properties are concentrated in the low-income sector. This meant that we had to modify our methodology, and introduce higher stress multipliers,” says Smith. This was part of a drive to help co-signatories of the Financial Sector Charter (FSC) to gain exposure to the low-income housing sector and meet their charter obligations.
At least one asset manager says that there have to be several tangible benefits before asset managers would consider investing their clients’ hard-earned capital in the structure. A low cost housing securitisation issue is considered too high risk for Kimon Boyiatjis at Trident Capital, who says that he feels strongly about not getting involved in this type of fixed interest structure for now, because the reward just isn’t there.
“We look after people’s money, including that proverbial ‘little old lady’, and I don’t want to be responsible for her eating dog food at the end of the month because of not taking enough care when investing ‘her’ money.
“The truth is that many managers were seduced by complicated structures that the banks and issuers were marketing, which are essentially over-priced,” said Boyiatjis.
In general the banks and issuers were applying huge amounts of pressure for asset managers to get involved in the many fixed interest conduits and structures – locally and internationally. According to Boyiatjis there just wasn’t enough yield to warrant taking the risk. There is a deeper underlying issue. The banks and issuers weren’t pricing in liquidity discount, either. Buying into the issue is not really the problem. But who do the buyers sell too.
It appears that generally the structures are opaque, and according to Boyiatjis, so complicated, that you would require a high-powered legal team to understand the intricacies of the offering. The opaqueness stretches to the underlying assets, their net asset value and the ‘what-if’ scenarios if something should go wrong.
An asset manager, who chose not to be named, says that some of the fixed interest issues – including the Siyakha Fund - are essentially the premium that asset managers have to pay for doing business in South Africa. He went on to say that if some of these products were presented to their investment committees, and they didn’t have a BEE flavour to them, they would be thrown out immediately, as being too risky.
Whatever the rating agency’s changed methodology, the level of risk was too high, in relation to the expected yield, for at least two fund managers, who make it their business to understand risk and return, when investing their clients’ money. Makes you think, doesn’t it?