Multi-Stock Basket

Typically 5-15 stocks for effective diversification. Below 5 provides limited diversification; above 15 adds complexity with diminishing benefits, and Singapore's market is narrower than larger markets so quality names run out faster. For beginners, 6-10 stocks is a good starting point. Because SGX has only three local banks, blend across sectors and include S-REITs to diversify properly.

Yes, but stock selection is limited by board lots of 100 shares. A high-priced stock like DBS at around S$45 needs about S$4,500 for a single board lot, which would swallow most of S$5,000. Focus on lower-priced names - many S-REITs trade around S$1-3 (one lot is S$100-300), Singtel around S$3, ComfortDelGro under S$1.50 - so you can build a diversified 4-5 name basket. Some brokers also offer regular-savings or fractional plans, and odd lots trade on the Unit Share Market.

For most baskets, buy all stocks within a short window (same day, or 2-3 days) to maintain your intended allocation. Staggered entry over weeks changes your risk exposure and complicates tracking. On SGX, place orders after the 9:00 AM open and use limit orders. Exception: large positions in thinner mid-caps may need phased entry to limit impact cost.

Use a spreadsheet with columns for stock name, shares/units owned, entry price, current price, distributions/dividends received, P&L, and current weight. Most broker apps (DBS Vickers, OCBC Securities, UOB Kay Hian, POEMS, Tiger, Moomoo) also show portfolio performance. In Singapore, be sure to include tax-free distribution and dividend income in your total return. Update at least weekly and compare to a benchmark such as the SPDR STI ETF.

This is exactly why you hold a basket. If one stock of 10 falls 50%, your basket is only down about 5% from that name. Review whether the fundamentals have changed - for a REIT, a distribution cut or dilutive rights issue is a red flag; sell if the thesis is broken, hold if it is a temporary setback. Individual stops (20-25%) can limit single-stock damage.

Equal weight gives smaller companies the same impact as large ones and, importantly, dilutes the banks. Market-cap weight lets the largest names dominate and closely mimics the STI - but the STI is roughly half banks, so market-cap weighting is effectively a concentrated bank bet. If you want broad Singapore exposure without that concentration, equal weight (or a capped scheme) is usually preferable.

Quarterly works well for most investors - frequent enough to maintain weights, infrequent enough to minimize fees. Use a 5% drift threshold alongside the calendar. A real Singapore advantage: there is no capital gains tax, so rebalancing never triggers a tax bill - the only cost is brokerage plus SGX fees plus 9% GST. Just avoid over-trading, which racks up fees and, in the extreme, risks being treated as a trader by IRAS.

Correlations typically spike during market stress - names that normally move independently start falling together as risk-off selling hits everything. In Singapore this is acute because the index is bank- and REIT-heavy. Plan for correlation breakdown by keeping some cash, considering an SGX index-futures hedge, and accepting that diversification helps least in extreme events.

Yes - in Singapore this is often the smartest move. S-REITs offer high tax-free yields and, crucially, tend to respond to interest rates in the opposite direction to banks. Pairing rate-winners (banks) with rate-losers (REITs) smooths returns across the cycle and adds income. Just ensure the REITs are liquid and watch gearing and DPU sustainability.

Dividends and REIT distributions: add to your return calculation; reinvest or withdraw (and enjoy that they are tax-free for individual investors). Rights issues (common among S-REITs raising acquisition capital): decide whether to subscribe, as ignoring them dilutes your stake. Scrip dividend schemes: choose cash or shares. Splits and bonus issues: share count rises and price adjusts, no action needed. Track all actions for accurate performance.

Use 3-5 years of monthly returns to estimate expected returns, volatilities and correlations. Apply a solver (Excel or Python scipy.optimize) to find the minimum-variance or maximum-Sharpe portfolio, using the low SGS/T-bill yield (recently around 2.5-3%) as the risk-free rate. Apply constraints (e.g. 0-15% per stock, 0-35% per sector, and a cap on combined bank weight) and robust/shrinkage estimation to handle estimation error - essential given how easily the optimizer would otherwise pile into the banks.

Target 3-5% tracking error versus the STI (benchmarked via the SPDR STI ETF or Nikko AM STI ETF). Lower than 3% offers limited alpha; higher than 6% creates behavioural risk from extended divergence. A key Singapore-specific goal of smart beta is reducing the STI's ~50% bank weighting. An information ratio (alpha divided by tracking error) of 0.5+ is good; 1.0+ is excellent.

Limit pair size to 10-15% of basket value and size for dollar or beta neutrality (e.g. long OCBC / short DBS). On SGX the short leg requires borrowing via the SBL facility or trading a CFD, the order must be marked as a short sell, and aggregated short positions are reported weekly. Account for borrow fees or CFD financing. Use a stop if the spread moves to roughly 3 standard deviations.

There is no tax-optimal sequence to engineer, because Singapore has no capital gains tax - the order in which you sell has no tax consequence, and tax-loss harvesting achieves nothing since there are no gains to offset. Sell based purely on conviction (exit broken theses) and cost (minimize brokerage and SGX fees plus GST). The only tax-related discipline is to keep turnover reasonable so IRAS does not treat you as a trader, which would make gains taxable as income at personal rates (up to 24%).

Use walk-forward testing: optimize on 3-4 years of data, test on the next year, and roll forward. Include realistic transaction costs - SGX trading fee 0.0075% plus clearing fee 0.0325% of contract value, 9% GST on fees, plus brokerage and slippage. Because there is no capital gains tax, you can skip after-tax return and tax-lot modelling entirely. Be mindful of the small SGX universe (the STI has only 30 names) when defining factor buckets, and report out-of-sample Sharpe (using the SGS risk-free rate), maximum drawdown and turnover.