[throwawaymath]

I'm going to upvote your comment because I don't think it deserves to be downvoted, and at the time of writing it's grayed out for me.

That being said - you're incorrect about about competition between active investors and HFT. That's a common misconception. HFT primarily occupies a marketing making role, which means they try to play both sides of the spread very quickly for a very, very small profit on each trade. There are elements of valuation here, but what's really much more important is very small holding times and low latency turnaround. The ideal goal of an HFT operation is a trading strategy which earns a profit 51% of the time and trades very frequently.

In contrast, active investors - whether quantitative, fundamental or some mix thereof - care more about being correct on fewer bets, which have more money behind them and which are held for longer periods of time (hours, days, weeks or months). These funds are not competing with HFT: HFT only competes with HFT. This is because HFT activity and active investing activity are completely alien to one another. HFT has a material impact on the profit margins (slippage), volume and liquidity available to active investors, but strictly speaking they don't actually compete (except in the narrow sense that you "compete" with a car salesman to buy a car for a better price).

HFT is a relatively tiny portion of the financial industry which gets outsized attention. It's generally more accurate to think of HFT firms as financial utility providers rather than investing firms.

[jandrese]

HFT is basically a tax on each transaction that gets applied because you don't have as accurate a view of the market as the guy who is down on the wire. If you're strategically buying shares in a company and holding them then HFT is not your competitor.

If you're a day trader trying to flip stocks by holding them for a couple of seconds at a time HFT is why you're bankrupt.

But it's also not true that HFT folks create markets. To create a market you need to sit on shares and offer them for sale. HFT leeches off of existing markets. It's true they offer share for sale, but only ones they bought a few nanoseconds earlier for the original price.

[throwawaymath]

Your second and third paragraphs are incorrect.

Day traders flipping stocks every few seconds won't lose money because of HFT firms, they'll lose money because of trading fees. Trading every few seconds is a wildly unrealistic strategy for most people to pursue on their own. On an average, per-trade basis the fees associated with buying and selling are several orders of magnitude higher than the profit margins of any HFT strategy. The only way your fees will even come close to the profit margins of an HFT are if your volume is such that you've become a market maker yourself. This is a very basic and fundamental tension that precludes HFT from being a competitive force to other traders engaging in speculation and investing.

Your final paragraph strikes me as ideologically bent, particularly with your use of the word "leech." It's an uncontroversial fact that HFT firms facilitate market making. HFT firms do sit on shares and offer them for sale. Most often they do this quickly, but occasionally they have holding times with longer horizons. What's more important than the turnaround time is the low latency with which they execute orders. Definitionally, HFT is engaging in market making because when someone wants to purchase a share, an HFT is ready to sell it to them. Likewise when someone wants to sell a share, an HFT is ready to buy it from them. This is quite literally, "making a market."

In point of fact, your hypothetical day trader would not be capable of buying shares every few seconds if it weren't for HFT (inadvisable though it may be). How do you propose they'd achieve the same kind of liquidity otherwise? By calling a broker? There are far fewer market makers than there are active investors. Passive investing activity with index funds also dwarfs the scale of HFTs. The straightforward conclusion that follows is that fewer, faster parties must exist to make markets for the many, comparatively slower investors.

This is an extremely well-studied subject; when you peel back the pomp and PR about HFT as an industry, you'll encounter an incontrovertible reality. There is no way to service modern trading activity happening every second without the HFT activity that happens every microsecond. By calling that latter activity "leeching", you read more like someone delivering an opinion rather than a cogent, well-informed and substantive criticism.

[jandrese]

At the end of the day how much position does a HFT firm hold? If nothing has gone wrong it is zero.

Just because they're stuck holding the bag on trades that end up being cancelled sometimes and have to wait for them to unload doesn't means they're making markets. If a market doesn't already exist the HFT firm is not going to create it.

They absolutely do increase the volume figures on markets, and that can be interpreted as making markets, but it's not an accurate representation of the big picture.

[throwawaymath]

> At the end of the day how much position does a HFT firm hold? If nothing has gone wrong it is zero.

This is also incorrect. HFT firms routinely hold positions for days and weeks. It would be inefficient and strange to literally deplete all positions and begin fresh anew each day. Executing orders with extremely low latency is not equivalent to be perfectly symmetric in buy and sell orders.

I encourage you to learn more about the topic you're talking about, because what you're saying is substantially at odds with how the industry actually works. At this point I'm curious how you would define market making, because it's very ironic for me (and anyone else reading) to hear you say these things and talk about what is or is not "an accurate representation of the big picture."