The most troubling part (imho) is still the wash sales. The SEC/CFTC isn't worried about fake orders or no-bid situations since there's no real way to stop that - computers or not, markets are markets and this can happen. I've actually seen wash sales myself on a live data feed and it was scary what they did to the price. It begins with quote fluttering and if that's not enough, then the wash begins. Once you detect that there are almost no big players present at the bid/ask, you can obtain pretty decent leverage on moving the price, especially on the lower liquidity instruments (as an HFT). I dislike HFTs but have learned to live with them and understand them fully so they don't pickpocket me so much, but saying that raising transaction costs would fix the issue is like swatting a fly on the market's face with a baseball bat. I'd rather take the HFTs than FTTs, thank you. Also, i suspect HFTs will re-invent themselves in the next few years and will become a fad again if volatility picks up. It's a cycle.
"Automated Trading Desk" ha ha ha ha ha. Citi bought this, their IT strategy geniuses promptly f*cked it, and they still can't figure out how to make any money appear. Head ATD quant left as soon as his contact allowed it. A nice trade. On his end. Nothing's broken when it comes to HFT profitability. There is an issue at the incumbent firms, with an emerging legacy code base & morons managing increasingly bloated tech departments. The best developers have left in disgust, while mediocre wannabes argue about frameworks & light their farts on fire. It's a people problem.
I am not an expert on this one as I have never been a market maker but placing a better bid and a better ask at the same time, is that a wash sale? Isn't this what market makers have been doing for many years? Please accept my apologies if I failed to get this but afaik a wash sale is essentially two opposite open positions in the same instrument at the same time, not two opposite quotes.
A wash trade is trading with yourself. Happens easily when you're running two strategies that don't talk to each other. Great for the exchanges, as they will still bill you. Kind of dumb for the firm as they could have just not done that. "Just" is months of work, however.
2.5 million for the platform that gives millisecond market info and broker origination data? That is cheap. Knowing where retail have their stops is priceless info. One wonders why they are not able to mine this info.
This was a kind of specific situation. I was watching a closed-end-fund which started trading from 1% premium to 10% in 2 days. I put the bid-ask and trades chart as well as time and sales (as well as lvl2 for multiple exchanges) on one screen and observed it for 4 hours. There was substantial volume for such a low volume stock in the past days but that day there was nothing. Maybe like 500 shares per hour. Meanwhile the bid and the ask kept changing size and price at a rate of 10 times per second. It looked like there was somebody slower bidding and another HFT fluttering the bid as well as the ask quotes. However, i believe that it was just one MM dominating that quote. Anyway, it would flutter (change bid/ask price and size) so frequently and like clockwork on T&S something like 237 shares would be traded at the bid, and then at the ask. The bid/ask at one point peaked at +6% for the day while almost no trades happened. After about 4 hours of this frenzied activity, some fat cat posted a huge sell limit order on EDGE and the party was over, they could no longer pull the ask like that. It's also made possible due to the fact that some HFTs (MMs) only place the ask so it is at NBBO, and if they are left alone at the book, they move towards the next level. This way, one HFT can manipulate another, but in this case it didn't looks like that, more like the only small other counterparty was chasing the bid and moving away as well (based on the Level2 on multiple exchanges i was seeing). Anyway, the wash sales give "credibility" to the NBBO because un-sophisticated traders or algos will think that this is indeed "market value" of a security at a given time, and may trigger their own stops (which are not normally transmitted to a broker). But it was pretty powerful seing a CEF go +16% premium in 3 days (based on NBBO). Subsequently it went back down to +4% premium 5 days later and i was short (after i saw the big EDGE order), and also luckily Interactive Brokers had just a few 100 shares left for borrowing heh... crazy stuff.
That isnt a pseudo edge, its a real edge and thats how 99% of the profitable market participants make money, the 1% that makes money gambling on direction are merely lucky (for now). There is no edge in betting on direction, because the future cannot be predicted, battles are won before they've even been fought.