I read an interesting statement here:
"If an ETF’s market price tracks its NAV well, it is likely to have a small market impact. On the other hand, if the market price is more volatile than the fund’s NAV, a large order could move the price before a market maker would be willing to step in and close that gap, leading to higher market impact costs."
Now, I am aware of the standard heuristic "$\beta$-exponent participation rate" market impact model as discussed by Almgren here and in this question.
With ETFs and OTC trading, the situation seems to be quite different depending on what your counterparty does. They can
- Keep the position open
- Buy/Sell the position at the exchange
- Match it with a trade of different direction (in-house)
- Hedge
- Trigger the Creation/Redemption process
Now some of these actions would directly impact the price while the others won't. The question is: Is there a way or a model to quantify market impact effects in this case?
I will be glad to accept specific models or empirical studies about this as an answer. Please do not hesitate to post your practical experience/findings (as a trader/market maker) in this topic as an answer as well!
