跳到主内容
@wquguru
精选80Stat Arb研究与分析

做市商返佣机制推高交易成本,惩罚竞争性做市商

Maker rebates effectively mandate that market makers earn a minimum amount of sp…

原文
发到 X

Maker rebates effectively mandate that market makers earn a minimum amount of spread and overall increase the effective cost to traders (in my view).

Assume there is some fair spread for an asset in a market that does not have fees. This might be 0.5 bps for MSFT shares.

If I am an exchange, and instead of charging everyone a 1 bp fee (bring the total effective cost paid to 1.5 bps), I instead do a 1.7 bps taker fee and a -0.7 bps rebate, then what happens is that because the market for MSFT cannot have a negative spread, it trades 1 tick wide.

Let's assume a fairly small tick size, so we'll say that it's basically at 0 (reality is not often this way, so in reality it is worse than this often).

Now the trader has paid no spread and 1.7 bps of fees instead of paying 1 bp in fees and 0.5 bps of spread and paying 1.5 bps total.

Who gets, that extra 0.2 bps? It's the market maker, it enforces a minimum income, and they get it. Worst of all - it punishes competitive market makers.

When spread is able to vary more, better market makers get the good fills, and the smart guy wins. But when spread is stuck at 1 tick wide then firms have to compete on who is the fastest to get to the front of the queue, and who is willing to wait in the queue.

We see this in 1 tick wide markets like stablecoin/stablcoin where it is very non toxic and it takes over an hour for the size on top of book to cycle once. That's a lot of time to wait in the queue! In the USDT/BUSD market on Binance this is caused by tick size (in fact this is why they make the maker fee is 0 bps in the stablecoin markets since a rebate would make the problem 10x worse), and we can see on Coinbase which has tighter tick sizes for USDC/USDT and does not have this issue.

更进一步:量化金融体系

看懂新闻只是起点——沿量化金融路径,把它变成能交付的工程能力

进入量化体系 →

相似阅读

另一事件,读法相近