Optimal Fee Pricing

Roberto Riccó, Barbara Rindi, Duane J. Seppi

Discussion by Katya Malinova

DeGroote School of Business

McMaster University

EFA 2023 Annual Meeting

Amsterdam

Research question: in the title!

  • Exchange trading fees
    • Equilibrium?
    • Socially optimal?
    • Need for regulatory intervention?
  • Key take-aways
    • optimal fees (fees vs rebates) depend on investor valuation dispersion/gains from trade
    • regulatory fee caps when exchange can extract rents

Building on Yong, Yao, Ye (2018): rebates \(\to\) finer tick size

Sophie: V = 20.0325

has the asset

Katya: V = 20.0375

price = 20.04

price = 20.03

Gains from trade \(\checkmark\)

but: the grid is too wide

Trade @p=20.03,

fee of 0.005 on Katya, rebate 0.005 to Sophie

Trade @p=20.04

fee of 0.005 on Sophie, rebate 0.005 cent to Katya

As if trade @20.035

This paper: investor valuations play a role!

Sophie: V = 20.0325

has the asset

Roberto: V = 25.0375

500 price levels between 20.04 and 25.03 to trade at

No need to rebate anybody for a trade to occur

Trade occurs even for large trading fees

\(\to\) exchange rent extraction! 

2-period limit order book, monopolistic exchange

Maker + taker = total exchange fee

Taker fee

Maker fee

Exchange profit

Range of investor valuations

Rebates only for  small ranges of valuations = low gains from trade

transaction probability

Fees oscillate with changes in valuation ranges:

  1. want makers to post the highest bid 
  2. want more takers to trade when the posted bid worsens by a tick

2-period limit order book, social planner

Welfare

Taker fee

Maker fee

Set the total exchange fee to 0

Questions:

  1. Why "cap the access fee" and not "cap the total" as the regulatory intervention?
  2. If the total fee can be capped, would the rebates re-emerge for valuations?
  3. Cheap shot: exchange competition?

2-period limit order book: Questions and Comments

  • Practical implications of "oscillating" fees?
    • challenging to implement even if valuations are proxied for

 

  • Restrictions and limitations of 2 periods and/or sequential entry and/or no intermediation?
    • One maker & one taker out of the set
      • Fees must incentivize trading among these two
    • Back to the example
      • Maker: Sophie V = 20.3025, has the asset
      • Takers: Katya V = 20.3075 and Roberto V = 25.3075
      • Want Sophie to trade with Roberto!
    • With >2 periods,  may want the fees such that Katya does NOT trade?

Side comment 1: model solved for limit buys (?), do the values of the optimal fees depend on this? 

Side comment 2: authors interpret valuations range = "trading demand"

To me: large demand = lots of orders/frequent arrival

3-period model, monopolistic exchange:

barely any oscillations (?)

Question/suggestion:

  • Do multiple periods "smooth" the fees?
  • If so, focus on >2 periods?

Optimal Fees = huge task

  • Yong, Yao, Ye (2018):
    • rebates help resolve frictions re: tick size
  • But re: optimal:
    • is the tick size the only friction?
    • other reasons for rebates?
    • market segmentation?
    • exchange competition?
    • optimal fees vs. optimal market structure?

A step back: why trading fees & "rebates"?

Traditional approach

  • Traders:
    • buyers & sellers

An alternate view

  • Traders:
    • liquidity suppliers & liquidity demanders
  • Exchange = service provider
    • facilitates trading
    • collects trading fees
  • If anybody gets paid
    • \(\to\) "rebate"
  • Service providers = liquidity suppliers
    • must be compensated 
      • inventory costs, adverse selection costs
  • \(\to\) Liquidity demanders compensate liquidity suppliers
    • bid-ask spread
    • fees (aka "maker rebates")

A step back: why trading fees & rebates?

An alternate view

  • Traders:
    • liquidity suppliers & liquidity demanders
  • Service providers = liquidity suppliers
    • must be compensated 
      • inventory costs, adverse selection costs
  • \(\to\) Liquidity demanders compensate liquidity suppliers
    • bid-ask spread
    • fees 

DeFi trading (AMM):

  • asset owners = liquidity suppliers
  • no tick size
  • no bid-ask spread
    • \(\to\)liquidity suppliers must be compensated through fees 
  • Malinova&Park (2023)
    • propose AMM implementation approach in equities
    • \(\to\) estimate savings of $-billions in trading costs ...

One big friction = exchange?

Summary

  • Excellent paper!

  • Cool new -- yet, close enough to the workhorse classic -- methodology!

  • Read it, use it, cite it!

@katyamalinova

malinovk@mcmaster.ca

slides.com/kmalinova

https://sites.google.com/site/katyamalinova/