Ernie.QT
patreon
Common Alpha 008: Market Impact Analysis
🕑 Added 2024-12-27 16:10:28 +0000 UTCMarket Impact Analysis
Introduction
This content is to discuss how to measure & conduct analysis on market impact in an execution. For any electronic desk & any algo trading strategies, it is crucial to know how much transaction cost is since a minor change in cost can result in turning a profitable strategy into a losing one.
Transaction Costs = Explicit (commissions and fees) + Implicit (slippage)
The explicit cost is kind of fixed & known before you trade, while the latter one is not.
The implicit cost, slippage, refers to the difference between the expected price of a trade and the actual price at which the trade executes. It occurs when the market moves rapidly, or when there is low liquidity, between the time you place your order and when it is processed.
Transaction Cost Impact on a portfolio
Here I will take a mid-frequency portfolio as an illustration. In crypto, assuming NAV as $1mio, with average daily turnover of 1.25x, transaction cost as 6bps (i.e. 0.06%) per trade, the transaction cost yearly will cost $273,750 (i.e. 27.375% annualized).
To improve the performance of the portfolio via transaction cost minimization, there are 2 ways;
the 1st one is to reduce the commission fees, that includes renting a higher VIP level accounts or
the 2nd way is to use some advance algos mentioned in last content, like using more limit orders to execute the order, since maker order is cheaper than taker order in majority of the crypto-exchanges.
Execution Trade-off
Timing Risk: risk of price change and information leakage as interval between arrival mid quote and last fill increases
Market Impact: risk of moving the market by shortening the interval between arrival mid quote and last fill.
In short, if you would like to get your order done quicker, assume market order the significant full size, you will suffer from slippage brought by ‘market impact’, on the other hand, if you have patience to wait to execute your order gradually, the market might move, usually against you, that is the ‘timing risk’.
There are a few of the models that I can find and they are trying to model the market impact, I attach one here with python scripts, readers can try to replace the setup and assumptions in crypto trading. Here we will introduce one:
Almgren Chriss Model
This model assumes the initial order, X, is completed at a uniform rate of trading over a volume time interval T. That is, the trade rate in volume units is v = X/T, and is held constant until the trade is completed. Constant rate in these units is equivalent to VWAP execution during the time of execution.
Temporary Impact: This component represents the immediate price change caused by executing part of the order. It is a transient effect that decays after the order is completed. The temporary impact is generally modelled as a function of the order size and market liquidity.
Permanent Impact: This component represents the lasting change in the asset's price due to the execution of the order. Unlike the temporary impact, this effect is irreversible and reflects the information conveyed to the market by the large order.
The essence of the Almgren-Chriss model lies in balancing the trade-off between minimizing market impact costs and controlling execution risk. Executing a large order quickly can reduce the risk but increase the market impact while executing it slowly can minimize the impact but heighten the risk. Readers can reference the python script on each components.
