What Should You Actually Trade

Rank instruments by what a round trip costs as a share of what the instrument moves in a day. The ones that move most charge most, and they charge more than proportionally, so the ranking runs almost exactly opposite to the one a chart suggests. The index ETF below hands over a fifth of one per cent of a typical day’s movement for the right to trade it. The micro-cap hands over 52 per cent.

Prerequisites: Lesson 10, for the four charges and the ratio they go into, and lesson 11, for why the same order costs differently in two instruments.

Five instruments, one column

Take the top row first, in full. An S&P 500 ETF at $500 quoted a penny wide: the spread is $0.01 ÷ $500 × 10,000 = 0.2 bps, and that is also what the round trip costs. It moves about 1 per cent on an average day, which is 100 bps. So the friction ratio is 0.2 ÷ 100, or 0.2 per cent, and 99.8 per cent of the day’s movement is still there for you to be right or wrong about.

Now the bottom row, which is the micro-cap lesson 1 priced, still quoted 1.48 and 1.55: $0.07 ÷ $1.50 × 10,000 = 467 bps for the round trip. It moves about 9 per cent on an average day, which is 900 bps. The ratio is 467 ÷ 900, or 52 per cent.

Instrument	Price and spread	Round trip (bps)	Daily range (bps)	Friction ratio
S&P 500 ETF	$500, 1¢	0.2	100	0.2%
Mega-cap tech	$200, 1¢	0.5	200	0.25%
Mid-cap	$40, 5¢	12.5	250	5.0%
Small cap	$5, 5¢	100	500	20.0%
Micro-cap	$1.50, 7¢	467	900	51.9%

These are categories rather than quotes for any particular security, and the point of the exercise is that you replace them with your own. What the shape of the column is telling you, though, does not depend on the exact figures. Top to bottom the ratio rises by a factor of 259, and it rises because the two quantities do not scale together: the micro-cap moves nine times as much as the ETF, and charges more than two thousand times as much to cross as a fraction of its own price.

That is the whole of “trade what moves”, and it is why the advice is expensive. The instruments that look most attractive on a chart are the ones whose charge grows fastest, and both facts have one cause: there is less resting size. Less resting size widens the quote, and it also lets a given order push the price further, which is the movement you were admiring.

Read the bottom row once more as a requirement rather than a cost. At 52 per cent you must be right about more than half of the day’s entire movement before you have covered the price of showing up. No edge in this course survives that. The micro-cap is not a harder version of the same game.

What this does not settle

That the cheapest instrument is the right one. A ranking is not a decision, and three filters sit on top of it. The instrument has to produce the behaviour your method needs — a mean-reversion method wants something that overshoots and comes back, and a single stock in a news cycle does not. It has to trade when you can be at a screen, which is a hard filter and no amount of discipline substitutes for being awake. And your account has to fund one sensible unit at the stop the chart demands, not the stop that makes the position fit. Any of the three can eliminate the top of your list.

Nor is the spread ratio your total friction. It is the floor, and commission, financing and impact all sit on top of it, which means adding them can only move an instrument down your list and never up. That is what makes the floor safe to rank on; it is not what makes it complete.

And a low ratio is not an edge. It tells you how much of the day’s movement is left after the tollbooth, not whether you can predict any of it. Ninety-nine per cent of nothing is still nothing.

The range in the denominator is also an average, and both of its inputs move together. In a violent week the spread widens and so does the range, and nothing here says they widen in proportion, because they do not have to. A ratio measured across sixty calm sessions is a description of sixty calm sessions, and the weeks you most want to be right in are the ones it describes worst.

And range is not the same thing as opportunity, which is the weaker joint in this whole method. The denominator counts how far an instrument travelled, not how much of that travel anybody could have held on to. Two instruments both moving 900 bps, one in a straight line and one by thrashing back and forth, get the same ratio out of this table and are not the same proposition at all. Nothing on this page measures the difference, and nothing in the rest of module 2 does either.

One evening with a spreadsheet settles a constraint that will sit on every trade you take for years. Almost nobody spends the evening.

Problems

1. Two rows, from live quotes. Pick the most liquid instrument you would consider and the least, and fill in both rows for real: price, spread watched live during the hours you would actually trade, and sixty sessions of average range. Compute both friction ratios. State the factor between them, and whether it surprised you.
2. Find where your own line falls. The two ratios connect through one number: your stop is some fraction of the daily range, and if you write k for that fraction, then lesson 10’s s is this lesson’s friction ratio divided by k. So measure your own k, take the win rate your setup actually has, invert the breakeven formula to find the largest s it can still carry, and multiply by k. That is the friction ratio at which your edge is exactly gone. Mark the line on your table: above it you are not taking a harder trade, you are taking a different one.
3. Where does your instrument sit? Rank eight to twelve instruments you have traded or been tempted by, then find the one you actually trade in your own ranking. If it is not near the top, write down the reason you chose it — and then decide whether that reason outranks the number next to it.

Sources. Joel Hasbrouck, “Trading Costs and Returns for US Equities” (Journal of Finance 64, 2009), which estimates effective cost across the whole cross-section and finds it rising sharply as capitalisation falls. Yakov Amihud, “Illiquidity and Stock Returns” (Journal of Financial Markets 5, 2002), on the ratio of price movement to volume as a measure of how expensive an instrument is to be in. Hans Stoll, “Friction” (Journal of Finance 55, 2000), the presidential address that treats cost as the central fact about a market rather than a detail of it.

You can now choose an instrument instead of inheriting one. The next lesson takes the four things a retail account can actually buy and shows that they are four different legal objects, with four different sets of risk hiding in the difference.

Educational only. Trading involves substantial risk of loss. Not financial advice. Past performance does not guarantee future results.