Where you are. You know how a price forms, what a market maker is paid for, and what a prime broker lends against. This lesson takes the app most readers actually use and asks the question all of that machinery makes answerable: the trade said zero commission, so where did the money come from?
Zero, and somebody is still paid
You buy a hundred shares. The app charges nothing. No commission, no ticket fee, no monthly charge.
Meanwhile the broker has staff, systems, regulators, marketing and investors. It is not a charity, and the trade was not free to execute.
Lesson 12 has already told you where to look. Between the best bid and the best ask sits a gap that every crossing order pays, and it does not appear on any statement. A business model that does not charge you a fee can still be paid handsomely out of that gap - and can even hand you part of it back while keeping the rest.
The idea in one paragraph
Zero-commission brokers do not execute your order on an exchange; they route it to a wholesale market maker, which pays the broker for the order. The wholesaler is willing to pay because retail orders are, on average, the good kind of flow to trade against - the opposite of lesson 12’s informed trader - so it can quote better than the exchange’s spread and still profit. Some of that improvement reaches you as a better price than you would have got at the touch, some goes to the broker as a rebate, and the rest is the wholesaler’s edge. Nothing is free; the payment simply moved from a visible commission into the price you received, where it is much harder to see and much harder to compare.
Splitting one order three ways
The exercise makes the arithmetic explicit. The spread at the touch is two cents, so buying at the ask costs you one cent against the mid. The wholesaler executes at half a cent better than that.
That half-cent of improvement is real: you paid less than you would have on the exchange. Of the remaining half-cent the wholesaler captured, a fraction goes to your broker as a rebate, and the rest is the wholesaler’s own margin. Add the three and they equal exactly the cent you would otherwise have paid.
Over ten thousand orders of a hundred shares, the exercise prints all three columns. The customer’s share is genuine money and so is everybody else’s.
Where the rules currently stand
The divergence is the interesting part rather than any one rule. The same practice is treated as an irredeemable conflict in one jurisdiction and a manageable, disclosable arrangement in another, on the same evidence. Where regulators genuinely disagree, this course says so rather than picking a side.
The same cast, new venues
Crypto-native trading desks are worth reading through this module’s vocabulary rather than as something unprecedented. There are market makers quoting both sides and earning spreads, brokers routing orders, lenders financing leverage, and custodians holding assets. The names are new and the venues are new; the roles are the ones you have spent this module learning.
Two differences are real. Trading runs continuously rather than in sessions, which changes what “the close” means and removes the overnight gap that traditional risk systems assume. And several roles that regulation deliberately separates in traditional markets - exchange, broker, custodian, market maker - have often sat inside a single firm, which is a concentration of exactly the conflicts the rest of this lesson has been describing. Module 7 rebuilds these venues properly.
Check yourself
1. The app charged no commission. In what sense did the trade still cost you something?
In the price. The order was routed to a wholesaler rather than an exchange, and although you received a better price than the touch, part of the spread you would have paid was retained by the wholesaler and part paid to your broker. The cost moved from a line on a statement into the execution price, where it is invisible and hard to compare.
2. Why is a wholesaler willing to pay for retail orders when lesson 12 said trading against others is dangerous?
Because the danger in lesson 12 was adverse selection - trading against people who know more. Retail orders are largely uninformed, so on average they are safe to trade against. That safety is what makes the flow valuable enough to buy, which is lesson 12’s mechanism running in the profitable direction.
3. Best-execution rules exist because of a structural tension. What is it?
The broker chooses the routing, and that choice determines both the customer’s price and the broker’s revenue. Those two move in opposite directions, so the conflict is built into the business model rather than arising from occasional bad behaviour, which is why it is regulated rather than left to reputation.
4. Two regulators look at the same practice and reach opposite conclusions. What should a reader take from that?
That the evidence does not compel a single answer: one side treats the conflict as impossible to manage and bans it, the other treats it as manageable through disclosure and best-execution duties. Where regulators genuinely disagree, the honest response is to state both positions and their dates rather than to present either as settled.
Do this
Split the spread and see all three columns.
python3 code/pfof_economics.pyThe starter leaves you split_per_share: the customer keeps the price improvement, the broker keeps the rebate, and the wholesaler keeps whatever remains of the half-spread. The assertion that matters checks the three parts sum to exactly the half-spread the customer would otherwise have paid - nothing is created, it is divided. Success prints the per-share and total figures across ten thousand orders and the line:
the trade was free and it was not costless: the customer's saving, the broker's rebate and the wholesaler's edge are one spread, split three waysThe completed version is in solutions/pfof_economics.py.
What you can now do. You can say precisely who pays for free trading and how much, read the divergence between jurisdictions without taking a side, and recognise crypto venues as this module’s cast in new clothes. That completes the cast. The project assembles the whole module into one running market.