Choose a fair benchmark for a portfolio and explain how a mismatched one gives a misleading picture.
At a family dinner, Marcus's brother-in-law says his portfolio returned 4% last year, which he says beat the market. Marcus asks which market. The answer is an SGD bond index, but the brother-in-law's portfolio is mostly bank shares.
He compared a bank-heavy account with a bond index in a year when the bond index happened to do worse. That says nothing about how well he chose his investments. A return on its own tells you little. It only becomes useful when you compare it with the return you could have had by doing something simple instead. That simple alternative is your benchmark. Choosing it badly is the easiest way to fool yourself about how you're doing.
A fair benchmark passes three tests. It is investable, you choose it in advance, and it holds roughly the same kind of risk as your portfolio.
Investable means you could have bought it through an index fund or ETF at a cost close to what the index shows. If nobody could have held it, it is a wish list. It was never a real alternative that you passed up.
Choosing it in advance means you pick it before the period starts and then keep it. If you choose after you know the results, you will end up picking one you beat. Fund managers who change benchmarks after a bad stretch are doing exactly that.
Holding the same kind of risk means that if you hold 80% shares, the benchmark should hold about 80% shares. If your shares are spread around the world, the benchmark's equity part should be spread around the world too. A mismatched benchmark measures the difference in risk and tells you nothing about skill. If you compare a share-heavy portfolio with a bond index, you look brilliant in most good years and terrible in a crash. Neither result says anything about your choices. Over time, shares tend to return more than bonds, and in exchange they fall further in bad years. The gap between a share-heavy portfolio and a bond index is mostly the equity premium. Module 2, lesson 2.7, "Risk premia: what you are paid to hold equity, term, credit and illiquidity", covers it in detail.
Cautious investors often make the reverse mistake. Someone with half their money in T-bills compares themselves with a world equity index and trails it in every rising year. They then decide they are bad at investing, when all they did was choose lower risk. Their benchmark should hold the same share of T-bills that they do.
A third version is common among Singapore investors. They compare a global portfolio with the STI because it is the index on the local news. The STI is concentrated in a few sectors, and banks carry a large weight in it, so it behaves differently from a world index. Unless your portfolio looks like the STI, the STI is not your yardstick.
Most portfolios hold more than one kind of asset, so the benchmark should too. A blended benchmark combines two or more indexes in fixed weights that match your target allocation.
Marcus wrote down his allocation in the course Build and run an ETF portfolio: 80% shares and 20% bonds and cash, with a small tilt to Singapore. His blended benchmark is 70% a world equity index, 10% the STI and 20% an index of Singapore government bonds.
Take a made-up year in which the world index returns 12% in SGD, the STI returns 6% and the bond index returns 2%. To get the blended return, multiply each index return by its weight and add the results: 0.7 times 12, plus 0.1 times 6, plus 0.2 times 2. That is 8.4 plus 0.6 plus 0.4, which comes to 9.4%. Marcus's own time-weighted return that year is 7.8%, so he trailed his benchmark by 1.6 points. This is a fair comparison because the benchmark holds the same mix he set out to hold.
Two rules keep a blend honest. First, rebalance the benchmark on a fixed schedule in your spreadsheet, for example yearly, so its weights don't drift with the markets. Second, keep the benchmark weights equal to your target allocation, whatever you happen to hold today. Say your stock picks have pushed you to 90% shares. Keeping the benchmark at 80% is what shows you that you took on extra risk.
The benchmark has to be measured under the same rules as your own figures. If it isn't, the comparison is broken from the start. Measure it:
as a total return, with dividends reinvested, using the net version where one exists (lesson 1.2 explained this) in SGD, converting from US dollars where needed with your chosen rate source (lesson 1.3 showed how) over exactly the same dates as your own figures
Index data is usually published by the index provider. If the index data is hard to get, an ETF that tracks the index is a practical stand-in. Using the ETF's total return has an advantage: it already includes the fund's costs, so it is closer to what you could actually have earned. Lesson 3.2 of Build and run an ETF portfolio, "Tracking difference is what you actually lost to the index", explains the gap between an ETF's return and its index.
Write the benchmark down once. For each part, record three things: the index, its weight and where its data comes from. After that, don't change it unless your target allocation changes, and if it does, note the date. The returns tab in lesson 1.8 will measure you against what you write down here.
Write down the blended benchmark for your current portfolio: each index, its weight, and where you will get its data.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).