An astute client asked me this question, and I was at a loss about why many of these High Frequency Trading firms want to take on the burden of making markets. After doing a little digging, I stumbled upon an excellent explanation on the website Quora.com written by Tikhon Jelvis, a code warrior in Menlo Park, California.
His answer was so elegantly structured and factually accurate that I decided to steal it and post it for my readers. I hope that by giving full credit to the author, he will forgive my brazen theft and maybe even pick up a few new followers in the process. Here’s how Mr. Jelvis answered the HFT question.
“Market-making is a particular service a trading company can provide on an exchange. A market maker helps match up buyers and sellers over time. Instead of buying or selling securities based on their underlying assets, market makers constantly maintain an offer to both buy and sell securities and make money on the difference between the two offers, known as the spread. The goal is to always match every buy with a sell and every sell with a buy in order to minimize the risk of holding onto securities for longer periods of time. If a stock FOO is trading at $100, a market maker might simultaneously maintain an offer of buying at $99.5 and selling at $100.5. If they manage to find both a seller and a buyer, they’ll make $1 on the difference in price.
Why is this useful? Well, it lets people who want to sell right now do so even if nobody wants to buy at that moment, and vice-versa. In other words, market makers provide liquidity—they make it easier to trade. For really popular stocks, this isn’t much of an issue, but it becomes really important for smaller securities.
Chances are, FOO isn’t a large company at all, and does not have that many people who want to buy it. So if you’re holding onto FOO and want to sell, you might have to wait for a while until somebody who wants to buy comes along. This increases your own risk and makes it harder to execute your strategy, or use your assets for anything else. In turn, this actually makes FOO less valuable, even if the underlying company is the same, because buying FOO stock locks your money in for a period of time. It would be in your interest to sell FOO right now at a slightly worse price instead of waiting for an opposite investor to come along and buy it, and that’s exactly what a market maker offers. The same thing happens on the opposite end: if you want to buy FOO, you can get it from a market maker for a slightly higher price right now. In return for bridging this gap, the market maker makes some money on the difference in prices.
This is a process that does not rely on automation at all—it can even be done by hand. However, for a lot of securities, it’s easier and cheaper to automate the market-making process. A program can be run on many different securities at once where a human trader might only be able to monitor a handful. So now a lot of companies end up making markets either in a completely automated or at least a semi-automated fashion, which is what you’ve heard of as “market-making algorithms”. In general, this has resulted in lower spreads, which is broadly good for the retail investor.
High-frequency trading (HFT) is a particular style of algorithmic trading characterized primarily by speed rather than sophisticated analysis. In a very broad sense, the goal is to use sophisticated hardware and software to make “obvious” trades faster than the competition. This also often involves making relatively little on every trade and compensating for this with sheer volume. The key part of HFT is not just using algorithms and computers but, specifically, optimizing for speed.
One particular strategy many HFT firms employ is market making. Here they out-compete everyone else by updating their quotes quickly and bringing the spread down even more: they’re willing to make less money each time because they can easily scale their market making operation to incredible volumes. But this is not the only sort of trading an HFT firm can do: they also use their technology for things like arbitrage (making money on small differences between related securities) or execution (breaking up large institutions’ trades to minimize market impact). I don’t want to go into much more detail here; the point is that HFT can do things other than just market-making. All that matters is speed.
On the flip side, not all market making algorithms are HFT because they don’t necessarily optimize for speed above everything else. Firms sometimes operate in markets where HFT is not present or would not work very well. They might also rely on slightly more complicated analysis which would be too slow for proper HFT, or even operate on a semi-manual basis with an actual human trader checking in occasionally—or even executing the actual trades themselves!—which is slower still. They still use algorithms, and they still do market-making, but it is not the same as HFT.”