This article was first published in the 4th-quarter 2015 edition of Personal Finance magazine.

In the past five years, the Satrix 40 exchange traded fund (ETF), which can be seen as a proxy for the JSE, has risen strongly, powered by glamorous and fleet-footed growth stocks. It has left two competing ETFs that select shares from the same universe plodding behind.

The Rafi and eRafi ETFs, which offer investments in value-orientated shares, have under-performed the Satrix 40 over one, three and five years.

The reason, says Daniel Wessels, a financial planner at financial planning firm Martin Eksteen Jordaan Wessels in Cape Town, is that value investing has not been a winning strategy in the past five years.

“Market preferences for stocks and sectors or industries move in cycles, but those are not predictable in advance,” Wessels says. “The emphasis in the Rafi is on stocks that recently lagged the market performance. As the market moves through cycles, the value style – and implicitly the Rafi – will start to out-perform once again.”

The attraction of a basic ETF that tracks the FTSE/JSE Top 40 Index is that it gives you broad access to the market. In one investment, you are exposed to 40 shares that make up about 83 percent of the JSE’s market capitalisation. The key term is market cap. The Top 40 Index is weighted by market cap, which means that companies with a high share price and many shares in issue have the biggest presence in the index. This is clearly seen in the holdings of the Satrix 40 ETF, which tracks that index. At the end of June, more than 30 percent of the ETF’s holdings were made up of just three counters: Naspers, SABMiller and BHP Billiton.

“The fundamental assumptions of a market-cap-weighted index is that the market is efficient, the market is pricing instruments correctly and, the bigger the stock in the market, the bigger the holding should be,” Wayne Dennehy, the head of ETFs and passive investments at Barclays Africa Group, says. “In a nutshell, it puts the market at the centre of the universe.”

An ETF or a passively managed unit trust fund that has the market as the centre of its universe uses what is known as a pure beta strategy; it does not aim to do anything but follow the market.

“Any conventional strategy that tracks indices that are weighted by market cap accepts the market price as the best judge of the value of businesses,” Wessels says. “However, systematic errors in Mr Market’s evaluations do creep in, especially towards the end of a strong bull market, when there is maximum optimism, and during periods of maximum pessimism. These periods are not easily identifiable ahead of time.”

An attempt to create a methodology that does not accept market price as the best judge of what is going to happen in the future led to the creation of the Rafi and eRafi indices. They are designed to weight companies based on fundamental values, rather than purely on market cap (refer to “How shares are selected for the Givi”, below). These indices use an enhanced beta strategy, or a smart beta strategy, because the weighting is based on something other than market cap.

Investopedia defines fundamental value as “the actual value of a company or an asset based on an underlying perception of its true value, including all aspects of the business”. It says value investing is a strategy for selecting stocks that trade for less than their fundamental value.

Licensing issues caused the creators of the eRafi methodology, Research Affiliates of the United States, to withdraw from South Africa the licence to use the eRafi indices, Dennehy says. The NewFunds eRafi 40 ETF and three sector eRafi ETFs (industrial, financial and resource sector ETFs) have been replaced by Absa’s NewFunds S&P Givi series.

The licence for the Rafi index remains in place, but there is just one ETF left that uses a Rafi strategy: the Satrix Rafi 40.

The JSE’s Rafi 40 Index is made up of shares of 40 companies that are selected and weighted based on their fundamental value. Value is assessed mathematically on the quality of the companies’ sales, cash flow, book value and dividends.

This leads to significant differences in the weighting of companies in the index, which can affect investment performance materially,” Nerina Visser, a director of etfSA, says. There are also some differences between the companies that make up the Rafi Index and those in the Top 40 Index. Naspers, which is the largest component of the Satrix 40 ETF, did not even make it into the top 10 holdings of the Rafi 40 ETF in June. It is not surprising that this was no longer a value play after the share price bounced around R1 900 in April and May. Instead, MTN dominated the Rafi portfolio, with 9.48 percent, compared with a ranking of fifth in the Satrix 40 ETF, making up 6.92 percent of the portfolio.

Research Affiliates created the Rafi strategy in 2005. It then fine-tuned the value filter by adding two more measurements: the quality of profits and the risk of financial distress. This enhanced Rafi methodology was called eRafi.

Four Absa NewFunds ETFs that track eRafi indices were launched in 2009. In South Africa, Absa’s NewFunds eRafi ETFs were available until the end of May, when they were replaced by Absa’s NewFunds S&P Givi series of four ETFs (the S&P Givi SA Top 50 and three sector funds).

Givi stands for “global intrinsic value indices”, and these indices were created by S&P Dow Jones. The Givi ETFs in South Africa are value-style investments, although they define value quite differently from the eRafi indices.

Absa’s S&P Givis are smart beta ETFs, which means that the shareholdings and proportions of shares are not determined by market cap.

Satrix’s Top 40 ETF and Rafi 40 ETFs and Absa’s NewFunds eRafi Overall SA Index ETF each track (or tracked, in the case of the eRafi Overall) a different index, so comparisons of performance may not be entirely fair. The funds track, respectively, the JSE Top 40 Index, the Satrix Rafi 40 Index and the eRafi Overall SA Index. Nonetheless, the comparison is instructive, because it shows how an investment in either of the two smart beta funds has under-performed the basic Top 40 ETF.

An investor looking at the returns over the past five years may wonder: why bother with an enhanced beta strategy? Why try to look for value when things are going so well? In the quarter to June this year – the last one with figures available for all three ETFs over periods up to five years, because the eRafi Index was no longer in effect after that date – the Top 40 ETF consistently beat the enhanced beta funds (refer to the table – link at the end of this article).

The methodology of the eRafi Index meant that Absa’s NewFunds eRafi Overall ETF was heavily invested in resource stocks, which have under-performed recently, Dennehy says. “The returns in the past year were particularly poor.”

Wessels explains how the composition of the Rafi 40 Index at the end of June 2015 reflected a value tilt

compared with the Top 40 Index (and therefore the ETFs that track them): “There was a large weighting (36 percent) of basic materials in the Rafi, versus 20.6 percent in the Top 40.” He points out that consumer services and goods are significantly underweighted in the Rafi Index (21 percent) versus the weighting in the Top 40 (41 percent).

Zack Bezuidenhoudt, the head for South Africa and sub-Saharan Africa at S&P Dow Jones Indices, says that investors need to be aware of the various styles of value investing and understand that not all definitions of value are the same.

“Value investing is a good long-term strategy that can complement a market-cap strategy, but investors need to understand that no factor adds value all the time – just like active managers don’t out-perform market-cap indices all the time.”

Visser agrees that value investing has been a very poor performer recently. “Active managers that follow the value style – RE:CM, Foord, and so on – have under-performed the general market recently, as investment styles also follow cycles.”

She says value investing is the antithesis of momentum investing.

Investopedia defines momentum investing as “a strategy that aims to capitalise on existing trends in the market. A momentum investor believes that large increases in the price of a security will be followed by additional gains.”

Visser says the under-performance of value investing “won’t persist in future; then the Rafi will out-perform, as it has in the past”.

She says that if you add value investing to your investment portfolio in the form of a smart beta ETF, it will not be a silver bullet to get you through different market cycles unscathed. However, adding a value-tilt investment to a market-cap-weighted investment adds to diversification, which is always a good idea.

Wessels asks whether it is useful to have funds in your portfolio that track smart beta indices such as the Rafi, and answers his question by saying it depends on how quickly the market begins to re-appraise those relative under-performers.

“A value style works well when turnaround times between cycles are relatively short, but in recent times that has not been the case. The winners have remained winners and the losers stayed losers, because the momentum style was dominant,” Wessels says.

So the ugly duckling will turn into a swan, and Cinderella will go to the ball. But trying to fix a date for the transformation would be foolish. Visser says investors need to be aware of cycles and what they mean for your money.

The Rafi and eRafi were introduced in South Africa just after the 2008 financial crisis and the resulting market downturn. Visser says it was a time when defensive shares had done well, and it was a good time for value shares. But market lift-off began in 2009 and the momentum accelerated from there.

“There is a risk if you invest based only on past performance,” Visser says. “How can you expect an investment to perform the same under different market conditions?”

She says that investors need to understand where a particular performance has come from. “If you don’t understand the origin of a particular performance, it is like looking at your garden at the end of summer and expecting it to always be green.”

The benefit of using smart beta strategies is diversification, Dennehy says. “To me, the key with all the smart beta strategies is that they offer something different for investors to the pure beta strategies. They both follow a mathematical formula … but one weights holdings by size, one by something else.

“I think too many people focus on whether one is better than the other. They certainly are very different. And, in the context of putting together a diversified investment portfolio, we like different,” he says.

Wessels doubts whether smart beta strategies are anything other than a form of active management, “although at lower management costs than conventional active management”.

Wessels is not saying that smart beta strategies cannot work and that they should not be used. But, in his opinion, investors must be clear that they are a type of active management, not part of a conventional passive strategy.

“What I do like about the smart beta concept is the fact that it can provide an active management style at relatively low cost,” he says.


Recent market darling Naspers dropped out of the portfolio faster than Tiger Woods dropped from the top of the rankings when Absa’s NewFunds eRafi Overall SA Index Exchange Traded Fund (ETF) was replaced by the NewFunds S&P Givi SA Top 50 ETF. The reason for Naspers getting the chop was not that the share had become too expensive relative to its underlying value, says Wayne Dennehy, the head of ETFs and passive investments at Barclays Africa Group; it was because the price was so volatile.

The first hurdle that shares have to clear to be included in the Givi 50 Index is volatility, or rather the lack of it. The index is made up of 50 shares (compared with the eRafi’s 40), and the way they are selected is different, too. This is how the filters on the Givi 50 work:

* All shares listed on the JSE except preference shares are considered.

* The first filter lets through 70 percent of the least-volatile shares.

* Shares that make the cut are assessed for their intrinsic value. En route, stocks with a market capitalisation of less than R10 billion and an average trade of less than R15 million a day are discarded.

* No single share makes up more than 10 percent of the index.

The Givi is based on the book value of a company, adjusted for expected earnings, Zack Bezuidenhoudt, the head for South Africa and sub-Saharan Africa at S&P Dow Jones Indices, says. This differs from the selection of the eRafi Overall Index, which began with a universe of the top 100 listed companies. To arrive at the 40 companies in the index, companies were filtered for the quality of their sales, cash flow, book value, dividends, net operating assets and debt coverage ratio.

The NewFunds S&P Givi 50 has 10 more counters in its portfolio than the eRafi Overall had. The filtering process and the additional space in the portfolio mean that 12 of the shares in the FTSE/JSE Top 40 Index have been discarded, Dennehy says, so the Satrix 40 ETF and the NewFunds S&P Givi 50 ETF have only 28 counters in common.

A peek into the components of the Givi 50 shows how it is different in other ways. The low exposure to mining shares in the Givi 50 is striking. There is no sign of Anglo American and BHP Billiton or the troubled platinum producers. Mining shares make up just 1.4 percent, compared with 12.9 percent of the Satrix Top 40 and 27.2 percent of the Satrix Rafi 40, at the end of July. The Givi 50 fills up the space with much larger holdings in British American Tobacco and the banks than the other two ETFs.

“I believe that Anglo and BHP were also removed for the same reason as Naspers in the current holdings,” Bezuidenhoudt says.

“If the stocks have high betas, meaning they are volatile relative to the rest of the market, then it’s highly likely that they will be removed in step one of the constructions of this multi-factor index.”

In addition to the flagship Givi 50, there are three sector ETFs in the Givi series: industrials (25 stocks), resources (10 stocks) and financials (15 stocks). The filters applied for the selection of shares to make up the sector indices is similar to that of the Givi 50, except the maximum weight per stock is 30 percent.

No eRafi funds have been shut down; just the benchmark indices have changed. As a result, the performance history of each eRafi fund stays intact under its new name, although they are not entirely compatible. The historical data shows eRafi returns up to May 31 and Givi returns from there on. The returns for the three months to August 31 are true Givi returns, but the longer-term performance is a blend of Givi and eRafi.

“That is unfortunate for us,” Dennehy says, “as the Givi-only return is far better, but the effect of the eRafi will wash out. Investors just need to be careful when looking at past returns, and they need to look at the actual Givi Index over different periods to assess how the Givi strategy actually did perform.”

Because the funds have stayed intact, they reflect the total expense ratios (TERs) of the past year (while the funds were tracking different indices). The TERs for the four ETFs range between 0.1 and 0.14 percent a year.

The TER is the total investment cost as a percentage of total assets, including management fees, the cost of administering the portfolio, bank charges, taxes and fees for share custody, trustees and auditing.

At a maximum of 0.14 percent a year, the Givi TERs are very low, although they are not the only fee you pay. You also pay to access the ETF through Absa Capital, etfSA or a stockbroker.

Dennehy says Absa subsidised the eRafi fees in the past, but Absa wants the Givi funds to stand on their own feet, so the TERs should start to move higher in the future. “I think they will drift close to between 0.2 and 0.25 percent over time. These levels are a better reflection of true costs.”

Plexus Asset Management was the original holder of the licence from Research Affiliates for the eRafi indices and the original asset manager of Absa’s eRafi funds. Plexus merged with Grindrod Asset Management in 2012, making Grindrod the asset manager. Dennehy says Research Affiliates gave Grindrod notice that it intended to stop the use of the index this year. When Absa Capital switched the benchmark for the four affected ETFs to Givi indices, it also moved the asset management function to Absa Alternative Asset Managers from Grindrod so the asset manager would be part of the group. The management company that issues and manages the process and the ETFs is NewFunds, which is ultimately owned by Absa.

Bezuidenhoudt says NewFunds has exclusive rights to use the Givi indices for 24 months for the ETFs but not other funds, such as unit trusts, although the index can be used as a benchmark by other unit trusts.