What is Market Making?
You have probably heard that a market maker (MM) captures the bid-ask spread in return for providing liquidity. But where does the spread come from? And given that market prices fluctuate all the time, how can the bid-ask spread be enough to compensate for the risks involved?
MMs profit by:
- Trading at prices where they have a theoretical edge.
- Hedging some or all of the risk of their trades so that they retain that theoretical edge as profit.
- Being highly optimised and efficient at identifying and capturing these opportunities.
- Receiving incentives such as reduced fees in return for providing liquidity and trading large amounts of volume.
Put simply, the investor buys an ETF because they want exposure to the “risk” that the price goes up, hopefully by many percentage points. The MM is happy to sell the ETF because they have calculated the price to be (for example) one basis point higher than its NAV at exactly this instant, and they have the ability to hedge the exposure and keep the profit.
The best MM is the one that can calculate exactly what the asset is worth (Pricing), can hedge the risk correctly (Risk Management), and can do it both quickly and cheaply (Execution).
A market maker calculates a ‘theo price’ – their best estimate of the exact fair value of a financial asset right now – according to whatever method makes the most sense for their trading strategy.
For example, option prices depend on the volatility of the underlying asset between now and expiry. However, if today is the 1st of January, then the price of options expiring in March and options expiring in April clearly have some component of volatility in common – the first 3 months out of 4. The MM must factor this in to ensure that March and April option prices are fair relative to one another.
Option prices (and the fair price of the OPTI-ETF, for that matter) also depend on the “base price” – i.e. the price of the underlying asset. This price changes all the time – hence the emphasis that theoretical pricing must be correct right now.
But what exactly is the base price? Imagine the order book looks like this:
…what price will you plug into your pricing model?
- The midpoint of the bid and ask?
- A weighted average of the bid and ask orders?
- The most recently traded price?
How does your answer change if someone lifts 999 lots from the best offer right now?
Think: what are the practical implications of these decisions?
You have calculated a theoretical price, but what orders should you place into the market?
What happens if all of your orders trade at the same time? Will you still be able to hedge your risk? Maybe you have been caught out in an adverse market move, and you traded on ‘stale’ information before you were able to adjust your quotes?
Maybe your theoretical price is perfect, but your Autotrader is slower than your competitor’s one, so they get the trade before you do?
The design of a MM’s Autotrading algorithm needs to be take all of these risks into account to ensure that it sends the orders they want, and deletes the orders they no longer want, ASAP. Remember: fair market prices fluctuate all the time.
It is therefore important to identify practical risks and constraints, set appropriate limits on the size of your positions, test and verify your Autotrader’s behaviour, and always be aware of all the ways that your theoretical pricing and execution could go wrong.
Obligations & Incentives
- maximum quoted spread width (e.g. < 0.3% difference between the bid and offer)
- minimum volume quoted (e.g. at least $30,000 worth on both the bid and ask sides)
- proportion of the market open hours that the MM must quote (e.g. at least 80% of the time).
In return, MMs are offered incentives such as reduced exchange fees. As MMs trade a large amount of volume for a small amount of edge, fees can add up very quickly!
This is one of the many factors that influences market prices.
Think back to our Telstra example in the liquidity section (Liquidity); let’s now extend that example:
- If you buy $1,000 of Telstra shares, nobody will notice; you will probably have no “market impact” at all.
- If you buy $1,000,000 of Telstra shares over the course of one day, it will still probably go relatively unnoticed; that’s 1% of the daily volume.
- If you buy $20,000,000 of Telstra shares in one trade on exchange, you push the market price against you; i.e. you won’t be able to buy all of that size at the best offer.
- If the market thinks you’re an informed investor, the market price will probably remain higher after your trade, possibly even move further upwards if others speculate that you might want to buy even more shares.If the market thinks you’ve made an error or simply executed your trade in a rush, the market prices will probably rebound quickly.
It’s not necessarily predictable exactly how market prices will respond to any given order, trade or pattern of trades. However, what is true on average is:
- The larger the volume you need to trade, the worse your average traded price will be.
- The more the market believes you are an informed investor, the more likely it is market prices will be permanently moved by the trade.
MMs continuously provide buy and sell orders, whereas investors can choose when and whether to buy or sell from the MM.
In many cases, the investors might know more about a particular asset than the MM does – e.g. if they specialise in researching technology companies.
This situation is called information asymmetry (i.e. the investor is acting on better information than the MM), which in turn gives rise to what we call adverse selection: the principle that there is a > 50% chance that the “true” fair price of the asset is worse (from the MM’s perspective) than what the MM thought it was, but you can only know whether it was too high or too low after someone has chosen to trade against you.
If that happens to you, you have been adversely selected!
How can a market maker sell an asset that they don’t own?
If an investor wants to buy an ETF from a MM, does the MM have to hold a stockpile (or “inventory”) of ETFs ready to sell?
The precise answer to this depends on the rules and structure of the specific market – not every market is the same – but in many cases MMs are permitted to “short sell” – i.e. sell an asset before buying it back later.
This is easier to understand if we think of “financial assets” as being purely represented by numbers in an account or written on a piece of paper, which means we can have flexibility in the timing of when “buying” and “selling” actually occurs – as long as all counterparties involved eventually receive the money and assets they are entitled to.
Algorithm (“algo”): you can search the internet for a formal definition, but informally this just refers to the logic of a computer program. i.e. “it’s a really good algo” means “this computer program does a really good job of fulfilling its purpose in life”.
Basis point (bp): 1 basis point (“1bp”) is equal to 0.01% – i.e. 1.2345 is 1bp of 12345.
Hedging: the act of doing one trade in connection with another trade to reduce or eliminate the risks with the first trade.
Informed: an investor that has a very good idea (or simply knows) that the true value of an asset is higher or lower than current market prices is said to be “informed”.
NAV: stands for “Net Asset Value” and refers to the “per unit” value of assets held by an ETF.
Quotes: refers to the order prices displayed (“quoted”) by a MM (much like getting a “quote” for car insurance, for example).
Theoretical edge: the difference between the trade price and the theoretical fair value; e.g. a MM buying an asset for $1.00 when they think it’s truly worth $1.0001 has 1bp of theoretical edge.