Out here binge-reading the FT today (#1479092, #1478548), but let's be honest: I always have been a sucker for tradfi financial magazines. (Somebody needs to take over these establishments and run them, uh, better.)
The Myth of Peak HFTThe Myth of Peak HFT
High-frequency trading was a BIG topic of concern/fear/excitement around the time leading up to Michael Lewis' Flash Boys — an excellent book in its own right.
The quintessential HFT trade is algorithmically matching thousands of orders to buy and sell a stock a few nanoseconds faster than your rivals.
For a while there was the cable race to the center of the New York and Chicago stocks/mercantile exchanges.
If you have a neat little model of market efficiency and arbitrage, there's some profitable market-making activity that pushes prices ever in line with one another... and quite some money in front-running your dumb retail order. The faster, cheaper and better, the narrower the spreads can get. And financial markets being (hyper) competitive, these arbitrage gains-from-extremely-fast trades evaporate quickly.
...except, they didn't. The deaths of the high-frequency trading firms have been greatly exaggerated. Somehow they're back and flourishing!
But it is sort of the case that that standard, beat-them-to-the-punch excessive HFT profits are gone:
with speeds butting up against the limits of physics, the era of big investments in co-located data centres, microwave towers or hollow-core fibre is probably over anyway, even as the cost of the existing infrastructure means that it’s extremely difficult for new players to emerge.
The profitability of the classic, superfast, high-frequency trading is probably still broadly as modest as it had become when [Donald] MacKenzie was researching his book [Trading at the Speed of Light]. But “traditional” HFT has become just one aspect of the industry, and mostly relevant to traditional market-making in very liquid markets like stocks.
Bitcoin and Retail Gambling to the HFT rescueBitcoin and Retail Gambling to the HFT rescue
Trading firms have also had to expand into new markets, products and asset classes, like fixed income and cryptocurrencies.
Apparently, trading crypto has been a major boost for 'em.
The head of one major firm once told Alphaville that he thought crypto was “stupid, but I’ll miss it when it’s gone”, because of the balkanised trading venues and enormous bid-ask spreads even on more liquid form of funny-money like bitcoin
SORT OF vindicates the main idea... you can make fast-money profits when the market you're trading in is inefficient, young, un(der)regulated, disorganized, and choppy.
the HFT industry didn’t enjoy the Great Tranquility of the 2010-2019 period, where volatility in virtually every asset class and market was remarkably low. Periodic bursts of turbulence — such as in 2011, 2015 or 2018 — simply weren’t enough. However, since Covid-19 there’s been a clearly different volatility regime.
...and nobody is surprised that it's the dumb-money retail feeding the beast, the dumbest of all money having entered the bitcoin-financialization craze.
This transformation of financial markets into slot machines is great deal for the house — stock exchanges and above all the traders that act as intermediaries. Despite all the reasons listed above, it is no coincidence that the prop trading upswing started in 2020, when retail trading took off again.
Last year, market structure analyst Paul Rowady offered up an interesting argument: while trading firms are now cruising, the main drivers of their success — seizing market share from banks and expansion across asset classes, geographies and timeframes — would soon run out of road
SUM:
- turning to VC means they can't see opportunity elsewhere
- new players not entering (no outsiders hungry to compete away excess profits)
- stock markets are just ~10 stocks now.
...though, these numbers are terrifying:
Jane Street will probably any day now reveal to its lenders — and therefore to the world — just how much money it made across 2025. The number is likely to be a big one — the trading haul from the first nine months of the year alone came to $24bn. Barring an epic collapse, Jane Street will have comfortably made more revenue from trading than even Goldman Sachs last year. If Jane Street’s traders reproduced the $10bn net trading revenues of its second quarter then it would have generated almost as much as Blackstone and BlackRock’s combined 2025 revenues.
How can dudez be so big?!
TL;DR, we'll continue to have HFT firms for a loooong time it seems.
archive: https://archive.md/VluEY