This paper poses a few fundamental questions regarding the attributes of the volume
profile of a Limit Order Books stochastic structure by taking into consideration aspects
of intraday and interday statistical features, the impact of different exchange features
and the impact of market participants in different asset sectors. This paper aims
to address the following questions:
1. Is there statistical evidence that heavy-tailed sub-exponential volume profiles
occur at different levels of the Limit Order Book on the bid and ask and if so does
this happen on intra or interday time scales ?
2.In futures exchanges, are heavy tail features exchange (CBOT, CME, EUREX, SGX and
COMEX) or asset class (government bonds, equities and precious metals) dependent and
do they happen on ultra-high (<1sec) or mid-range (1sec -10min) high frequency data?
3.Does the presence of stochastic heavy-tailed volume profile features evolve in a
manner that would inform or be indicative of market participant behaviors, such as
high frequency algorithmic trading, quote stuffing and price discovery intra-daily?
4. Is there statistical evidence for a need to consider dynamic behavior of the parameters
of models for Limit Order Book volume profiles on an intra-daily time scale ?
Progress on aspects of each question is obtained via statistically rigorous results
to verify the empirical findings for an unprecedentedly large set of futures market
LOB data. The data comprises several exchanges, several futures asset classes and
all trading days of 2010, using market depth (Type II) order book data to 5 levels
on the bid and ask.