A DV E RTO R IAL
0 0
100
100
1 1
500
600
500
500
600
500
2 2
500
500
400
500
500
400
3 3
4 4
5 5
100
100
6 6
400
500
100
400
400
500
100
400
7 7
300
300
8 8
200
200
200
200
9 9
500
100
300
600
500
100
300
600
##
100
100
400
300
## ith the current 100
400 how to measure
300 the
focus on transparen- 100 Let’s first consider
##
## cy in equity markets, many different ability of a broker to successfully capture
metrics
are being used
to assess broker
passive liquidity that is sitting in a lit order
##
200
##
200
performance, including spread capture and
book. The average trade size will depend on
##
their ##
ability to interact effectively with both two things: the size of the active order sent
##
lit and
by the broker to access the passive liquidity
## dark order books. One such metric
being
used
to
assess
how
brokers
interact
and the structure of the orders that are
##
##
with the market is the average trade size
sitting passively in the order book.
##
300
400
## This metric
300
600 For example,
400
achieved.
is used to proxy for 600
consider a broker that would
##
liquidity
is simple to calculate;
like to immediately execute 1,000 shares.
## capture and 500
500
however
the problem
is that it is also a
The broker sends a marketable order for
##
500
##
500
simplistic measure.
1,000 shares to an order book displaying
##
## it comes to 600
600 capture, there 1,000 shares. At one extreme, the order
When
liquidity
0 0
100
100
1 1
1600
1600
2 2
1400
1400
3 3
4 4
5 5
100
100
6 6
1400
1400
7 7
300
300
8 8
400
400
9 9
1500
1500
10
900
10 and have
900
one trade of 1,000 shares
an aver-
11
age trade size of 1,000
11 shares. In the second
case, the broker will 12
execute 10 200
trades for
12
200
100 shares each resulting in an average
13
13
trade size of 100 shares. In both cases the
14
broker captured 1,000
14 shares of liquidity
at once but the resulting
average price de-
15
15
pends on the structure of the order book at
16
16 their 1300
1300
the time the broker sends
order.
The broker has no 17
control
the order
17 over 500
500
book, yet if we use average
fill size
as our
18
500
18
500
proxy for performance, we would come to
19
19 about 600
600
very different conclusions
how the
Size matters: A better than
average fill metric
W
are three dimensions to consider: what can
be captured by removing liquidity from an
order book, what can be captured from sit-
ting passively in a lit order book, and what
can be sourced in dark pools.
Figure 1
book could consist of a single passive order
for 1,000 shares while at the other extreme,
the order book could consist of 10 passive
orders of 100 shares each.
In the first case, the broker will execute
broker did. If we had an effective metric,
we should see no difference between these
two outcomes – in both cases the broker
got us our 1,000 shares.
We can easily fix this problem by simply
Avg
Avg fill
fill size:
size: 360
360
1800
1600
1400
1200
1000
800
600
400
200
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Time
Figure 2
Avg
Avg fill
fill size:
size: 771
771
1800
1600
1400
1200
1000
800
600
400
200
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Time
8
THETRADETECH DAILY
In Figure 1 we
show a chart that
stacks individual
fills of various
sizes for each time.
The average size
is shown in the
title and is a result
of averaging all
the individual fills
without aggregat-
ing by time. In Fig-
ure 2 we aggregate
the all the volume
that is stacked and
then compute the
average. If there is
more than one fill
in a time window,
the average on the
right will be larger
than the average
on the left. This
method can be
used to compute
the average fill size
of active volume to
measure a broker’s
ability to capture
passive liquidity.