• market making is costly => reward for posting
  attract posted orders to marketplace

• (maybe) tick size too wide in very competitive markets
  allow more aggressive posting and then attract market orders

Perfect

Markets

• m/t split is irrelevant, only net fee matters
• bid and offer prices adjust to account for net fee

• m/t fees not paid directly by investors but brokers
• m/t target specific traders (HFTs) and brokers (retail)

Imperfect

Markets

Conflicts of Interest

• post at low rebate venue: trade faster
• post at high rebate venue: earn more fees
• Battalio et al 2016: harms passive retail orders

Our paper

• brokers' routing practices (measured by e.g. the execution quality of orders)

• the level of intermediation

• standard measures of market quality

• different market participants (i.e., retail traders, institutional participants, liquidity providers)

• Role of the exchange fee split  when (some) traders only pay them "on average", through "flat" broker commissions

• Zoom onto one friction: "averaging of the exchange fees"
  • Assume: competitive brokers charge "the average" exchange fee through a flat commission.
• Abstract from conflicts of interest, routing conflicts, etc.
• Abstract from other market frictions (e.g., tick size)

• Single asset, no asymmetric information
• Investors value it at either +V (buyers) or -V (sellers)
  • Arrive asynchronously (one per period)

• Trading:
  • Post limit orders or trade with market orders
  • Several simplifying assumptions (limit orders good for one period, one limit order/one lot at a time, etc.)
  • Maker-taker exchange:
    • taker fee for market orders
    • maker rebate for limit orders

• "maker-taker": pay taker fees on market orders & receive maker rebates on limit order trades
  • e.g., because they access the exchange directly or b/c or the "tiered plan"
• "flat-fee": pay fixed/flat broker commission
  • competitive broker, charges the average of the fees
  • same fixed fee, no matter the type of order

• All traders want to either buy or sell
  • no ex ante "market makers"
  • (earlier version of the model had market makers + asym. info)

• weighted by the probabilities of "flat fee" investors trading with market vs. limit orders
• flat fees = average exchange fees incurred by the broker

• fee for limit orders
• assumed <0 (rebate)

• fee for market orders
• assumed > 0 

maker fee < flat fee < taker fee

ask

bid

ask+flat fee

ask+taker fee

flat fee buyer: spread/2 + flat fee

maker-taker buyer: spread/2 + taker fee

Ceteris paribus:

• "more expensive" for "maker-taker" traders to trade with market orders than for "flat-fee" traders

 "Flat-fee" traders accept a wider range of quotes

ask

bid

bid\(+\)flat fee

bid\(-\)maker rebate

flat fee buyer: get spread/2 - flat fee

maker-taker buyer: get spread/2 + maker rebate

Ceteris paribus:

• limit orders are more attractive for "maker-taker" traders than for "flat-fee" traders 

 "Maker-taker" traders post better quotes

\(\Rightarrow\) for a given limit order book state

1. When a "maker-taker" trader prefers trading with with a market order to submitting own limit order, so does the "flat-fee" investor.
2. When a "flat-fee" investor prefers to use a limit order, so does the "maker-taker" trader

• "Maker-taker" and "flat-fee" traders are willing to accept different prices

Earn a smaller spread but with a larger probability

Earn a larger spread but with a smaller probability

• \(\Rightarrow\) can choose to price limit orders in two ways:
  • at "narrow" spreads \(\to\) next period, all traders on the right side of the market send market orders
  • at "wide" spreads \(\to\) next period, only "flat-fee" traders send market orders and "maker-taker" traders prefer to post own limit orders

• (Almost) all traders are "maker-taker"
  • \(\to\) need to send limit orders that are attractive enough to these
  • \(\to\) all traders submit limit orders at "narrow spreads" 
  • \(\to\) all traders use market orders whenever there is an order in the limit order book on the "right side of the market"

0

\(\lambda_1\)

\(\lambda_2\)

\(\lambda_3\)

\(\lambda_4\)

1

• maker-taker traders post narrow spreads
• flat-fee traders post wide spreads
• maker-taker traders use market orders only when spreads are narrow
• flat-fee investors use market orders whenever there is an order on the "right side of the market

• all post narrow spreads
• all use market orders

Type 1

Type 3

Type 5

Type 2

Type 4

• all post wide spreads
• only flat-fee investors use market orders

mixed strategy

mixed strategy

• For larger fractions of flat-fee investors (\(\lambda>\lambda_1)\):
  • maker-taker traders do not trade against flat-fee trader limit orders (spreads are too wide)
  • \(\to\) lower volume & unrealized gains from trade
  • \(\to\) maker-taker/asymmetric fees are detrimental to welfare 

Establishing causality from historical data is challenging

• Lower fee asymmetry (i.e., smaller difference between the taker and the maker fees) \(\Rightarrow\) larger \(\lambda's\)
  • \(\Rightarrow\) larger \(\lambda_1\) \(\to\) larger "no distortion" region

• As maker-taker fees change:
  •  thresholds change
  • may switch across different equilibria types
  • \(\to\) non-monotonic effects on
    • market quality
    • trading volume

• narrow spreads
• all use market orders

Type 1

Type 3

Type 5

only flat-fee investors use market orders

• MT: market orders only when spreads are narrow
• FF: market orders for all spreads

Highest

Volume

Broker:

client limit orders

client marketable orders

Problem:

• long queue & low exec probability
• trade after Exchange 2 limit orders 
• adverse selection

• Any asymmetry between maker and taker fees may distort routing! 
  • Not just rebates!