Read a fund's reported returns and pick out the assumptions behind them before you compare it with anything.
Two funds in the same category can publish five-year returns a point or so apart. Every figure in a performance table rests on decisions made by whoever prepared it: the fees taken off, the share class reported, the currency, the period, the set of funds included and the benchmark chosen. Any one of those decisions can shift a figure by more than the difference between the two funds you are weighing. Before comparing, write down the decisions behind each number and restate both funds so they rest on the same ones.
Run the checks in a fixed order, which catches most problems: fees and share class first, currency and period second, survivorship third, benchmark last.
Marcus had a friend who wanted a second opinion on two Asia equity funds. Fund A's factsheet reported 6.9% a year over five years and Fund B's reported 5.8% (example figures), so Fund A was ahead by 1.1 points. The friend had already settled on Fund A. After twenty minutes with the two factsheets, Marcus concluded the published numbers could not be set side by side as they stood.
Your first question is what has already been subtracted. In Singapore, most performance tables are worked out from the fund's net asset value (NAV), which has the yearly management fee and the fund's other running costs deducted already. Charges you pay directly usually sit outside the table:
the initial sales charge when you buy any platform or wrap fee switching or redemption charges
A few factsheets add a second performance line with the initial sales charge deducted, and plenty leave it out. Find out which line is in front of you and add whatever it does not include.
Next, look at which share class the number belongs to. A share class is one version of a fund among several. Every class owns the same portfolio, and the classes differ in fees, in the currency they are priced in, or in how they pay out income. A low-fee institutional class reports better returns than the retail class a private individual can buy, because a higher fee on the same portfolio lowers the return by roughly that fee. Use the class you would really buy, normally retail, and add the direct charges the table omits.
Marcus found this first. The 6.9% for Fund A came from its institutional class. The retail class carries a fee about 0.8 points a year higher and returned about 6.1% (6.9% minus about 0.8).
A comparison like this needs only the rule. The full treatment of charges and share classes is in Module 2, "Find every charge you pay", of the course "Unit trusts, robo-advisors and managed money compared".
When two funds are priced in different currencies, convert one so both are in the same currency. For a Singapore investor that currency is SGD. If the US dollar fell during a year, converting a USD fund's return into SGD makes that year's return lower.
Fund A reported in USD and Fund B was priced in SGD. Marcus converted Fund A's return to SGD with the method from Lesson 1.3, "Returns in Singapore dollars when the asset is priced in US dollars". Over those years the US dollar had weakened, which took more off Fund A's lead.
Dates need the same attention. A factsheet dated three months after another covers a different five-year window, and Lesson 1.5 showed how much results can change with the window chosen. Either find factsheets that end on the same date or work out both funds' returns yourself from their published prices.
Fund platform rankings, league tables and lists of "top performing funds" include only funds still running. Funds with poor results are often shut down or folded into stronger funds, and their records go with them. The survivors therefore look better than the typical fund an investor could have chosen at the outset. This upward distortion is called survivorship bias.
Suppose ten funds start on the same day. Across five years, three perform badly, averaging minus 4% a year, and are closed or merged. The other seven average 6% a year. A table built from the funds alive today shows an average of 6%. The typical fund investors actually held returned nearer 3% a year: seven tenths of 6% gives 4.2%, three tenths of minus 4% gives minus 1.2%, and 4.2% minus 1.2% is 3%.
Each surviving fund's own record stays accurate. What the bias distorts is your sense of how likely you were to pick a good fund. Fund house marketing has the same problem, so if a manager tells you "our funds have beaten their benchmarks", ask how many funds the house ran five years ago and how many are still open. Lesson 12.7, "Survivorship bias, SPIVA and where your edge is and is not", comes back to survivorship bias when you assess your own investing record.
Each fund names a benchmark, the index or rate it compares itself with. The manager picks it and has reason to pick one that is easy to beat. A fund that mostly owns mid-sized Asian companies might cite a large-company Asian index, or a fund holding both shares and bonds might cite a cash rate plus a margin. Against a poorly matched benchmark, a fund can come out ahead in most years for reasons unrelated to skill.
The factsheet gives you what you need to test the choice: the top holdings, the split by country and the split by sector. Look at those and decide which index most resembles the portfolio. When that index differs from the one the fund names, measure the fund against both. Lesson 1.6, "Choose a benchmark that matches what you actually hold", lays out the tests for a good match.
Fund B cited a broad Asia ex-Japan index, and its holdings were close to it. Fund A cited a narrower index that had performed poorly over the five years, which flattered Fund A.
Marcus then put the two funds on one footing: retail class, in SGD, over the same five years, measured against the same index. Fund A's 1.1-point lead turned into a small lead for Fund B. The decision stayed with his friend, as it should, and it was now based on two figures that measured the same thing.
Now take any one factsheet you have and work down it as Marcus did. Before you look at any comparison, write down each convention it uses: what the return is net of, which share class it shows, its currency, its end date, and how its benchmark compares with its holdings.
Take one fund factsheet and list how it reports returns: fees, currency, share class, benchmark and period.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).