This publication is intended for intermediary use
The conundrum
Our clients appoint us to manage their equity portfolios to generate above-market returns over time. While investment performance is easily measured, the conclusion around whether we have met or exceeded our clients’ objectives can sometimes be confusing. In particular, the choice of benchmark against which our performance is measured could result in differing conclusions as to whether we have helped our clients achieve their goals.
For us the choice of benchmark is irrelevant in the construction of our clients’ portfolios. As we follow an unconstrained or non-benchmark cognisant approach we do not use the benchmark as a starting point to construct portfolios.
In choosing a benchmark, clients need to carefully consider any potential unintended consequences which may be hidden in the benchmark itself. For example, the FTS/JSE All Share Index is heavily biased towards international large cap stocks with the Top 10 companies making up 60% of our market. The 40 biggest companies account for almost 85% of the market, although by number of stocks account for just 25% of the 165 companies in the FTSE/JSE All Share Index.
As a result of our unconstrained value approach, our performance may differ significantly from the benchmark over different measurement periods. We invest in companies whose share prices are low relative to the underlying value of their cash flow stream over time. The amount of capital we allocate to each investment depends on the size of the discount to fair value1 and liquidity. The companies in which we are invested and the size of each investment in the portfolio therefore have very little to do with the weighting of the company in the benchmark.
Furthermore, with such a large concentration in so few companies in the FTSE/JSE All Share Index, one’s relative performance is often a function of the companies you’re not invested in.
Our clients can expect their portfolio to look very different to the benchmark at times – as is currently the case. Our view is that many of the large cap companies that make up 85% of the FTSE/JSE All Share Index are expensive. In aggregate, the Top 40 companies are around 12% overvalued2 versus the smaller cap companies which still offer around 7% upside to intrinsic value. The chart below illustrates the significant PE premium that large cap stocks trade at versus their small cap counterparts: 43% versus the long-term average of around 23%.
Consistent with our investment approach we currently have almost 20% less invested in large cap stocks and 16% more in in small cap companies relative to the FTSE/JSE All Share Index.
Some of the other difficulties with a market cap weighted benchmark relate to the challenge of allocating money to companies which have performed the best and therefore have a higher weighting in the index. For example, Naspers currently constitutes almost 8% of the FTSE/JSE All Share Index. The time to have committed substantial amounts of capital would have been in 2004 when Naspers made up only 1% of the Index. Just because it’s up 2650% over the past ten years is not, on its own, a reason to allocate such a large portion of your portfolio to Naspers today. I challenge you to ask any fund manager whether they would own 8% in Naspers today, if it only made up 1% of your benchmark.
“Everything should be made as simple as possible, but not simpler.” Albert Einstein
Another way of measuring the performance of the market is to use an equally weighted benchmark. Instead of weighting the company by the size of its market capitalisation all companies are equally weighted. A R100 invested 12 years ago would now be worth almost 20% more if it was invested in the equivalent equally weighted version of the Top 40 Index instead of in the FTSE/JSE Top 40 market cap weighted benchmark. However, the outperformance of the equally weighted benchmark did not occur in a straight line. There were two periods during which the market cap weighted index performed better.
The first period was between 2007 and 2008 when the FTSE/JSE All Share Index returns were dominated by the performance of the heavyweight resource sector. During this period resource shares made up over 50% of the FTSE/JSE All Share Index. Today it’s around half of that. Due to our concerns around the significant overvaluation of resource shares at the time, our clients had substantially less invested in resource shares than was reflected in the benchmark. While performance lagged the market cap weighted benchmark during this period, the strategy ultimately proved to be correct.
The second period in which the equal weighted benchmark lagged the returns of the market cap weighted Top 40 was the period since mid-2012. It’s no coincidence that this is a period in which heavyweight industrial stocks, such as Naspers, SAB Miller and British American Tobacco, dominated market returns. Our analysis concludes that these stocks are expensive and we prefer to allocate more of our clients’ capital to opportunities where the prices are trading well below intrinsic value.
Market cap weighted benchmarks unintentionally expose clients to the “momentum” of share prices well beyond the level justified by the value of the company’s cash flow stream. While this may work in a rising bull market, investing with a manager that closely hugs the benchmark will result in full participation in the inevitable bust which will transpire when the over-inflated prices of expensive stocks correct back to their fair value.
In summary, the disadvantages of a market cap weighted benchmark are not inconsequential. To generate outsized returns for our clients we need to think and act differently to the market. This means allocating capital to our best ideas, independent of the weighting the company may have in a benchmark. We believe that comparing an investment manager’s performance with an equal weighted benchmark helps our clients to better assess and understand the manager’s performance.
Glacier Research would like to thank Ricco Friedrich for his contribution to this week’s Funds on Friday.