Estimating Volume Limits & Trading Capacity

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
Full Member
Posts: 23
Joined: Sat May 28, 2005 8:03 am

Estimating Volume Limits & Trading Capacity

A salient issue in system development is the twin issue of realistic volume limits and trading capacity; that is:

1. How much volume can you trade in a particular market and

2. Given the above for each of the markets in your portfolio what is the maximum asset capacity for your trading strategy.

These two issues are of course heavily impacted by market venue (pit or electronic exchange) and market order type.

For the course of this discussion lets consider the following assumptions:

1. Order Type: Limit Orders only

2. Market Venue: Electronic exchanges/Interbank FX

3. Trading Times: all markets trade for 22-24 hours a day/5 days a week

4. Markets: Electronic Indices, major FX pairs, Interest Rate Futures, Crude Oil, Gold,

Ideally we should have:

A. VOLUME LIMIT (VL) = A percentage of average daily volume and

B. TRADING CAPACITY (TC) = derived from VL and your money management algorithm. Let's use a simple money mgmt algo where you limit your margin usage to 20% of account equity (MER).

In this case TC = Sum of VL * Average Daily Volume * Margin for each market divided by MER

My thoughts are VL should be no more than 1% of average daily volume.

What are your thoughts on this issue?

Whatâ€™s the ideal/most realistic VL estimate?

What other methods exist to estimate market impact and volume limits?

Lets have an intense, focused and insightful debate on this important issue.

Many Thanks.

sluggo
Roundtable Knight
Posts: 2986
Joined: Fri Jun 11, 2004 2:50 pm
There's a pretty good book by Larry Harris (link) that discusses strategies for getting your orders filled when you know your size is big enough to move the market. And then there are "Algorithmic Execution" bots that chop up big orders into very small pieces and dribble them out a little at a time, hoping for less market impact. A well known example of algorithmic execution is the "Iceberg" order, in which only a small piece (the tip of the iceberg) is made visible at any one instant. There are a couple of nice surveys of Algorithmic Execution (here) and (here).

Algorithmic execution uses bots to get better (we hope) fills for orders supplied by traders. There's another area, called algorithmic trading, in which bots generate the orders and work them for best fills.

stancramer
Roundtable Fellow
Posts: 60
Joined: Sat Feb 07, 2004 10:34 am
Location: Washington DC