You will be able to estimate the full cost of a trade, including the spread and the price moved by your own order.
Marcus once bought a small SGX stock at what the app said was S$1.60, sold it two weeks later at what the app said was S$1.60, and lost money after commission. The price hadn't moved. He'd bought at the ask and sold at the bid, and the gap between them, on that stock, was wider than his commission twice over. The cost never appeared on any statement as a line item. It was built into the prices.
The commission is the visible cost of a trade. The spread, market impact and slippage are the invisible ones, and on small stocks they're usually bigger.
The spread is the ask minus the bid. On its own, two cents says nothing. As a percentage of the price, it tells you what a round trip costs before commission.
Use the midpoint between bid and ask as the fair reference price. Buying at the ask costs you half the spread above the midpoint, and selling at the bid costs the other half below it. A round trip, buying and later selling, costs about one full spread.
Here are three made-up quotes taken at the same time of day. Marcus's local bank shows a bid of S$31.99 and an ask of S$32.00: a spread of one cent, about 0.03% of the price. His US chip designer shows US$149.98 and US$150.02: four cents, also under 0.03%. Larkspur, the small SGX supplier from lesson 5.1, shows S$1.59 and S$1.61: two cents, or 1.25% of the price.
So a round trip in Larkspur costs about 1.25% before commission, against about 0.03% for the bank. Trade Larkspur in and out four times a year and the spread alone takes about 5% of the money involved. That's more than many years of fund fees, and it never shows on a contract note.
Spreads change through the day. They're usually widest just after the open, can widen before results or news, and are often narrower in the middle of the session. They also widen in a sell-off, when market makers need more pay for the risk, which is the moment many investors decide to trade.
The spread only tells you the cost for an order that fits inside the best bid or ask. A larger order eats through several price levels. The extra cost of that is market impact.
Take the made-up Larkspur book again: 4,000 shares offered at S$1.61, 6,000 at S$1.62 and 10,000 at S$1.63. A market order for 5,000 shares takes all 4,000 at S$1.61 and 1,000 at S$1.62. It costs S$8,060, an average of about S$1.612, which is S$60 above the midpoint value of S$8,000, or about 0.75%.
Now make it 20,000 shares. The order takes every share on offer at all three levels: S$6,440 plus S$9,720 plus S$16,300, for S$32,460. The average is about S$1.623, and the cost against the midpoint is S$460, about 1.4%. Quadrupling the order nearly doubled the cost per dollar traded. And the book you see isn't the whole story, because once sellers notice a large buyer, some raise their prices or pull their orders.
Impact is tiny for large, busy stocks. On Marcus's bank shares an order of a few thousand dollars barely touches the first level. On SGX small caps it can be the largest cost you pay. A rough check: compare your order size with the depth at the best price and with the stock's average daily volume. If your order is more than a small slice of either, expect impact.
Slippage is the gap between the price when you decided to trade and the price you actually got. It includes the spread and impact, plus any move in the price while you hesitated, waited for the open or let a limit order sit.
Say Marcus decides on Friday night to buy Larkspur at its last price of S$1.60, and his market order fills on Monday at an average of S$1.623. His slippage is 2.3 cents a share. Some of that is the book and some is the market moving over the weekend. The split matters less than the habit: write down the decision price before you trade, so you can measure the gap later. Without it, you'll always remember the price you got as the price you meant.
Slippage can be negative. A limit order that sits below the market and fills after a dip gets a better price than the decision price. That's the point of a limit, and the cost is the trades that never fill.
The standard way to reduce impact is to break a large order into smaller limit orders placed over time, so each one fits within the depth available. For the 20,000-share Larkspur order, four limit orders of 5,000 shares over several days, each at or near the ask, might each cost close to the spread rather than climbing the book.
That isn't free. The price may move away from you while you wait, and some of the orders may not fill at all. Professional trading desks spend a lot of effort on exactly this balance, and you face a simple version of it: if you're adding to a long-term holding, a few days' delay rarely matters as much as paying 1.4% in impact. If you're reacting to news everyone else has seen, the price may run before you finish.
The closing auction from lesson 5.2, Order types beyond limit and market, and when each helps, is another option, since more opposite orders arrive at once.
Marcus wrote one line in his workbook after doing this sum: any order in an SGX small cap larger than the shares at the best price goes in as limit orders, split or into the close. Now check the same thing on your own screens, at one time of day, for three very different stocks.
Compare the spread as a percentage for an STI stock, a small SGX stock and a large US stock at the same time of day.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).