Tragedy of the Commons - Price Formation Edition
[rant]
Imagine Deutsche Bahn (or Amtrak, or National Rail, ...) was required to let competitors use their rail infrastructure for free. Competitors would obviously accept this gift - and only run their own trains on the few truly profitable routes. All the while, the national provider would still be responsible for the maintenance of the entire network as well as servicing the less/non-profitable routes. Worse than that, the competitors' trains are VIP-only and not marketed to the general public, no timetables or fares are published.
Now replace "rail network" with "price formation", "Deutsche Bahn" with "Deutsche Börse", "profitable routes" with "liquid stocks" and you get the state of play for European equities. It is the classical tragedy of the commons but with price formation instead of fish.
[/rant]
BMLL's Classified Trades files are now available in the Deutsche Börse datashop and Fig. 1 below is solely based on them. It shows the fragmentation of trading in German equities across trading venues vs. the stock's liquidity. The vertical axis is the market share (by traded notional) for each stock. Stocks are ordered from most liquid to least liquid along the horizontal axis which shows the average daily notional. The left panel considers all market models while the right panel only breaks down lit (including lit auctions) volumes. Note that the Xetra auctions are lit in this classification as indicative prices, volumes, and imbalances are published throughout the call phase.
Observations:
The top-4 venues make up >90% of the volume. All others just add complexity without adding much value.
A lot of volume trades in the dark ("VIP trains"). This volume does not contribute to price formation; it only benefits from the price formation provided by lit venues ("train tracks").
If the rationale behind dark venues was offering the ability to trade large sizes with limited market impact, then one would expect a larger market share of dark trading for less liquid names where market impact is more significant. The data shows the opposite.
More generally, the non-primary venues' market shares are close to zero for less liquid securities. They pick the raisins and only serve the popular rail routes.
Xetra is responsible for around 70% of the price formation (if one uses lit volume as proxy; also see here) for the most liquid names - significantly punching above its weight.
PhD Researcher in Market Microstructure at KU | European Stock Market Fragmentation
2moAlways like to read your analysis! Regarding your take on dark pools and illiquid stocks, I wonder about i) executing time since filling and finding a counterparty might take more time for illiquid stocks, and ii) whether the price impact is really less for liquid stocks adjusted for similar proportional trade sizes (front-running etc.). What is your view?
Capital Markets Execution Consultant for Public Companies | Share Buy-backs | Group Exco | Global Head Execution Services | Algorithmic and Program Trading | Governance | Risk Managment | SM&CR Significant Person
2moStefan Schlamp - really interesting perspective and points. Tough challenge- as our perception is increasing competition drives efficiency and better outcomes for consumers- but I think there are arguments within your points which suggest that those better outcomes might really for the VIP train providers. The more I think about market structure and what delivers the best outcome for the investors and issuers (the markets end purpose) is that anything that trades away from the lit should probably have a pretty large “LIS” minimum- maybe something like 10% of ADV. Thoughts?