You will be able to set maximum weights for single positions and sectors and explain why they matter.
In the risk dashboard from lesson 2.8, Build a risk dashboard for your portfolio and two benchmarks, Marcus wrote that his bank and chip holdings added risk he hadn't chosen deliberately, and that he'd come back to it when he wrote position limits, which is what this lesson does. A limit is a decision made once, on a calm day, about the most damage any one idea or theme is allowed to do. Without one, the size of each holding is set by whatever happened to rise, which is how his account ended up where it is.
Marcus's holdings are the made-up figures used throughout the course, about S$202,000 in all. Other figures are made up too.
A position limit is the largest weight any single holding may have, as a share of the whole portfolio. It's the simplest risk rule there is, and the most useful, because every single company can fail in ways no analysis catches. Lesson 6.3, The auditor's report, key audit matters and going-concern wording, opened with Hyflux for that reason.
Marcus's weights start with the world ETF at S$91,000, about 45.0% of the portfolio, followed by the SGD bond fund at S$30,000 or 14.9% and the STI ETF at S$20,000 or 9.9%. Then come the bank at S$16,000 (7.9%), the chip designer at S$15,000 (7.4%), the S-REIT at S$12,000 (5.9%), the T-bills at S$10,000 (5.0%) and Larkspur at S$8,000 (4.0%). Shares of all kinds add up to about 80%, as his written allocation says they should.
Using the sizing method from lesson 11.1, Position sizing: equal, conviction and volatility weights, he set a target of 5% of the portfolio for each single stock, or 4% for any stock with volatility above 30%. On S$202,000, that's S$10,100 and S$8,080. Lesson 8.6 found Larkspur could lose about a third of its value in a bad cycle, so at 4% the bear case would cost about 1.3% of the portfolio, a loss he can absorb.
Against those targets, his bank at 7.9% and his chip designer at 7.4% are both over, by about S$5,900 and S$6,900. Larkspur at 4.0% sits on its target. The S-REIT, at 5.9%, is just above.
Positions don't stay put. A stock that doubles while the rest of the portfolio stands still roughly doubles its weight, and nobody decided that it should. A limit with no rule for winners is a limit that holds only until something goes up.
So the rule needs a trigger and an action. Marcus wrote: "When a single stock rises more than one percentage point above its target, I trim it back to target at the next review." A 5% target trims at 6%, a 4% target at 5%. The one-point gap stops him trading every small wobble. By that rule, the bank and the chip designer need trimming now, and the S-REIT, at 5.9%, doesn't trigger on its own. Lesson 12.2, Rebalance a mix of ETFs and single stocks to a written rule, turns these breaches into trades.
The second kind of limit applies to sectors and countries. It catches concentration that's invisible when you look holding by holding, such as owning several bank stocks, or a bank stock plus a fund full of banks.
To count it properly, look through your funds to what they hold. The STI gives large weights to the three local banks, as lesson 4.6, Singapore's exposure as a trade and financial hub, noted. Say, with a made-up figure, that 40% of Marcus's STI ETF sits in banks. His bank exposure is then his S$16,000 bank holding plus 40% of S$20,000, or S$24,000, about 11.9% of the portfolio. His world ETF does the same thing with technology. It already holds the large US technology companies, his own chip designer among them at a made-up weight of 1%, and with a made-up 8% of the world ETF in semiconductor companies, his chip exposure is the S$15,000 chip designer plus S$8,000 of Larkspur plus about S$7,300 through the fund, some S$30,300 or 15.0%.
Marcus set a sector limit of 15% of the portfolio, counted through funds. Banks pass at 11.9%. Semiconductors sit right on the limit, which is a warning in itself.
He set a country limit for Singapore shares too, because his benchmark holds only 10% in the STI and Build and run an ETF portfolio covers the pull of familiar markets in lesson 5.1, Home bias: holding too much of what you know. His Singapore shares are the STI ETF, the bank, the S-REIT and Larkspur, S$56,000 in all, or about 27.7% of the portfolio. He set the limit at 25%, so that's a breach too.
There's no correct number for any of these limits. They come from three questions you answer for yourself. How much would you lose in one holding before it changed your plans? How confident are you in your own analysis, measured by your record rather than your feelings? And how much do you want your portfolio to differ from your benchmark?
Marcus's answers gave him 5% for a single stock because a total loss there would cost about a year of his savings, which he could survive. The sector limit of 15% stops any one industry carrying more than about a fifth of his shares. The Singapore limit of 25% allows a home tilt well above his benchmark's, but not without end. Your numbers can differ. What matters is that they're written down before you need them, and that every limit has a stated action when it's broken.
His summary line read: "Three breaches: bank and chip designer over the single-stock limit, and Singapore shares over the country limit, with semiconductors sitting on the sector limit. No trades until I've written the cluster limit in lesson 11.5."
Now get your own holdings and weights in front of you, with the latest factsheets for any funds, ready to set and check each limit.
Write a maximum weight for single stocks, sectors and countries and check your current portfolio against each.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).