I first read Flash Boys by Michael Lewis on warm summer evenings by Lake Geneva, back when I was single, kid-free, and had the time to think deeply about things like latency arbitrage and fragmented liquidity. Simpler times.
I remember being fascinated by the rise of high-frequency traders (HFTs). They were like financial ninjas, slicing milliseconds off trades to make profits in the slivers between market inefficiencies. At the time, they felt like villains. Now, I’m not so sure.
Like most things in finance, it’s complicated.
Liquidity Loves Company
Imagine if all trades for a stock happened in one place. You’d get maximum liquidity—buyers and sellers finding each other faster, tighter spreads, better pricing.
But that also means giving one venue a monopoly. And monopolies don’t have the best track record for customer service or competitive pricing.
So instead, we have multiple exchanges. More competition, more fragmentation. That’s where HFTs came in. They connected the dots, arbitraging tiny price differences between exchanges and making money doing it.
In theory, that helped bring liquidity to the broader market. In practice? It depends.
When Arbitrage Turns Into Market Making
HFTs started out by exploiting inefficiencies. Over time, they realized they were already acting like market makers—providing constant quotes and smoothing out pricing. So they went all-in.
But unlike traditional market makers, these are profit-driven firms with no obligation to stay in the game when things get messy. When volatility spikes, they often pull back. Exactly when liquidity is needed most.
So are they good or bad?
Well, if they disappeared tomorrow, markets would definitely be less liquid. But their presence can also create the illusion of liquidity—until it evaporates at the worst moment.
Where the Profits Come From
Some HFT profits come from fair value creation. By arbitraging across venues, they help equalize prices. That’s a good thing. Trades happen that wouldn’t otherwise happen.
But some of the profits come from less friendly practices—like front-running large orders.
You might think you’re playing a fair game by simply asking for a price. But behind the scenes, the system is trying to figure you out. If you’re a Swiss exporter who regularly sells USD to buy CHF, the algorithms know your playbook. They can anticipate your next move and trade ahead of you.
If you’re a retail trader with no pattern, no edge, and no market-moving potential, you’re less dangerous to them. So they offer you better pricing. Tighter spreads. More “liquidity.” Because they are less afraid of adverse selection. Market Makers cannot set an infinite safety margin because they are engaging in a competitive activity: if their risk management is too tight, someone else would pick the trade.
That’s the irony: the more predictable (and unthreatening) your flow is, the better deal you get.
In the corporate world, banks offer so-called dark pools, a service that essentially removes the “tag” from your order, and therefore the risk of being front-run. Also, you have to trust this bank in providing the best possible price, since you are not putting them in competition with others.
Retail Traders and the House Advantage
Retail platforms offering CFDs know their clients lose money—consistently. So they often don’t even bother hedging the positions. They act like a casino running its own book.
Why hedge when the odds are overwhelmingly in your favor?
Retail options are another type of “CFD”, in the sense that the Market Maker can hedge (or not) their book and therefore have an incentive in acquiring retail orders.
Internalise Trades
This is another way in which Market Makers can make a profit. If they acquire a flow that is selling and another one that is buying, they can match them, keep the bid-ask spread as a profit and not pay any fee to any exchange.
Brokers do this as well and, in theory, should rebate some of those cost savings to their customers.
Information = Edge
Market makers love information. They pay for order flow because knowing who’s trading—and why—is valuable.
That’s what “payment for order flow” (PFOF) is all about. It’s not just about capturing spreads. It’s about buying insight.
That model came under fire in the EU. ESMA raised concerns in 2021, and the model is expected to be phased out by mid-2026.
So what do you do if your business relies on PFOF and that revenue stream is disappearing?
You build your own exchange.
Scalable Capital: If You Can’t Beat ‘Em, Become ‘Em
Scalable Capital, a European neo-broker, launched the European Investor Exchange—where it acts as a market maker.
In theory, this could reduce trading costs for their clients. Tighter spreads, better execution. Everyone wins.
But vertical integration also introduces risks.
When Everyone Works for the Same Company
Scalable’s co-founder said they’re “taking trading, clearing, settlement and custody into our own hands.”
So did FTX.
When a single firm controls the full stack—execution, custody, settlement—there are fewer checks and balances. If something goes wrong, there’s no outside party to raise a red flag. We have a sample =1, so we cannot draw any conclusion but the risks embedded in the model should be clear. Sure, Scalable is regulated by BAFIN, but BAFIN is the same regulator that dismissed many claims by FT journalists that Wirecard was a fraud…until Wirecard turned out to be a fraud. Not only did they not do their job, they hindered people from doing their job. The presence of a regulator doesn’t necessarily mean they are going to police.
If a broker starts to send funds to venture capital from the client account, which has to be separated from their corporate account, someone might spot it and raise their hand. No one could in the case of FTX because everything was done in-house.
Madoff did the same thing.
“Firstly, investors who invested directly with Madoff did not buy shares in an offshore company or partnership interests in an onshore limited partnership. They opened a trading account. The assets of the account were then managed, executed and held in custody by Madoff. Typically, hedge funds will:
• utilise independent brokers to execute trades,
• engage a fund administrator to independently calculate the NAV of the fund, and
• retain independent custodian(s)/prime broker(s) to custody the fund’s positions.
The segregation of these functions is typical of the hedge fund industry and, amongst other things, plays an important role in reducing the risk of fraud. While most feeder funds into Bernard L. Madoff Investment Securities utilised independent fund administrators and auditors, in this instance, these service providers had no other choice but to verify the existence and accuracy of the trading information back to Madoff, rather than to an independent broker.”
Clients thought they had independent oversight. They didn’t. No separate custodian, no third-party broker, no admin calculating returns. Just one guy running the show.
We don’t need to assume bad actors are everywhere. But the structure matters. When you eliminate external oversight, the risk of failure—or worse—increases.
Your broker is your gateway to the outer world: you do not want to tie your hands by dealing only with them because moving from one to another is complicated. Trading conditions might be good today, but it is also important to think about how they might be tomorrow. To move your business away is surely time-consuming, and it might also be costly if, for example, you are forced to sell your positions and pay capital gain taxes.
Transparency Beats Illusion
Look, brokers need to make money. Nobody expects them to work for free.
The problem isn’t with making money—it’s with pretending they don’t.
If you’re not paying a commission, you’re paying somewhere else. Maybe in wider spreads. Maybe in poor execution. Maybe in being front-run.
There’s no such thing as a free trade.
So ask yourself: Would you rather pay $5 upfront and know the cost—or feel good about “zero commissions” while unknowingly giving away more?
Sometimes, transparency is the real value.
What I am reading now:
Follow me on Bluesky @nprotasoni.bsky.social
0 Comments