|
An Evaluation of
the Influence of Large Reporting Traders
on Futures Market Performance
Executive Summary
 Background:
This
study was commissioned in May, 2008 by the CME Group for the benefit
of a consortium of futures exchanges that included the CME Group,
Kansas City Board of Trade, Minneapolis Grain Exchange, New York
Mercantile Exchange and the Intercontinental Exchange. Work was
completed in October, 2008. The distribution of this study on
Informa’s website has been authorized by the participating
exchanges.
Purpose:
The
intent of this study was to take a detailed look at the data
available from the CFTC on large reporting traders and, in
particular, index funds and managed money, to evaluate their impact
on market liquidity, volatility, price discovery and convergence.
Data:
The CFTC
gave the CME Group temporary access to the core data used in this
study for the specific and limited purpose of facilitating the
accuracy of the study’s findings. The CFTC provided the data after
its staff determined it to be necessary for the furtherance of the
regulatory responsibilities of the CFTC and the participating
exchanges. The CFTC did not participate in or endorse the design,
execution or formulation of the findings of the study.
The data provided was a subset of aggregated information
regularly published in the weekly Commitments of Traders
report. The CFTC data did not disclose individual positions or the
identities of traders. In addition to the normal categories of
commercial and non-commercial traders, the data also identified
accounts that were associated primarily with index trading and
managed money. The dataset began on January 1, 2005 and ended on
June 30, 2008.
General
Finding:
After
a lengthy and detailed analysis of the data provided, Informa found
very little evidence that the trader groups of interest, index funds
and managed money, were routinely detrimental to any of the studied
markets. All of the trader groups displayed instances of
non-optimal behavior (including small traders), but none were
consistently harmful to the studied markets.
Specific Findings:
Part 1:
Participation Levels
-
The most consistent pattern across
most markets was the well-known building of positions by index
traders about 75 days prior to expiration and the rapid
liquidation of these positions around 25 days to expiration.
-
Market presence by money managers
was much more erratic than any of the other trader types.
-
Commercial traders entered most
markets early and maintained relatively large positions
throughout.
Part 2:
Liquidity
-
Liquidity, measured as the ratio of
volume to open interest increased over the study period
(2005-2008) in the corn, Chicago wheat, natural gas and crude
oil markets
-
Liquidity appeared to decline in the
Kansas City wheat, Minneapolis wheat and cotton futures.
However, abnormal market conditions in early 2008 may have
contributed to this result.
-
In most markets, gains in liquidity
could not be consistently linked to any particular trader group.
Weak relationships were detected between market presence of
index funds and liquidity in Kansas City wheat and crude oil.
Part 3:
Volatility
-
For some commodities, the presence
of index traders and money managers tends to be associated with
higher volatility (corn, wheat, cotton) but for other
commodities this association is not as clear (crude oil, natural
gas).
-
In most of the agricultural markets,
the presence of index traders and money managers is somewhat
correlated. We suspect this is because markets trended higher
during the study period, causing trend-followers (money
managers) to adopt the same position as index traders (long).
-
No persuasive evidence emerged that
would suggest index traders or money managers caused increased
volatility.
Part 4:
Price Discovery
-
A measure of the amount of daily
price pressure that each trader group exerted in the market was
developed and then categorized as beneficial if the group’s
pressure moved toward fundamental value or detrimental if the
group applied pressure that moved price away from fundamental
value.
-
All groups exhibited both types of
pressure. Index traders tended to have a higher
beneficial-to-detrimental ratio, but that may have just been
because the market was coincidentally trending in the direction
of their preferred position (long).
-
Granger causality results held
little evidence that any group has a sustained and significant
influence on price. This is consistent with previous findings.
-
A Vector Autoregression model
confirmed the Granger causality results.
Part 5:
Convergence
-
The price pressure methodology from
Part 4 was repeated with a focus on the last 20 trading days.
-
No trader group was found to apply
pressure in routinely helped or hindered the convergence of the
futures to cash.
-
In most commodities, prices had a
tendency to trend upwards during the final days of a contract’s
life, suggesting that prices were not consistently inflated
above fundamental value.
-
Cotton and natural gas exhibited
routine over-pricing. However, this could have been a function
of the limited study period.
Executive Summary in pdf
format
Powerpoint
Presentations
in pdf format
(please either open the following links
to view online or right-click,
save to your computer
and then open for faster viewing...)
Large Trader Study, Part
1
Large Trader Study, Part 2
Large Trader Study, Part 3
Large Trader Study, Part 4
Large Trader Study, Part 5
|
