On commodity price limits

• Author: K. Geert Rouwenhorst,Xiao Qiao,Rajkumar Janardanan
• Published date: 01 August 2019
• Date: 01 August 2019
• DOI: http://doi.org/10.1002/fut.21999

Received: 18 January 2019

|

Revised: 18 January 2019

DOI: 10.1002/fut.21999

RESEARCH ARTICLE

On commodity price limits

Rajkumar Janardanan

1

|

Xiao Qiao

2

|

K. Geert Rouwenhorst

3

1

SummerHaven Investment Management,

LLC, Stamford, Connecticut

2

Paraconic Technologies US Inc,

New York, New York

3

Yale School of Management, Yale

University, New Haven, Connecticut

Correspondence

Rajkumar Janardanan, SummerHaven

Investment Management, LLC, Stamford,

CT 06902.

Email: rjanardanan@summerhavenim.com

Funding information

J. P Morgan Center for Commodities at

the University of Colorado Denver

Abstract

This paper examines the behavior of futures prices and trader positions around

the occurrence of price limits in commodity futures markets. We ask whether

limit events are the result of shocks to fundamental volatility or the result of

temporary volatility induced by the trading of noncommercial market

participants (speculators). We find little evidence that limits events are the

result of speculative activity, but instead associated with shocks to funda-

mentals that lead to persistent price changes. When futures trading halts price

discovery migrates to options markets, but option prices provide a biased

estimate of subsequent future prices when trading resumes.

KEYWORDS

circuit breakers, commodity futures, commodity options, price limits, speculation, speculative

trading

JEL CLASSIFICATION

G13, G14

1

|

INTRODUCTION

Exchanges use a variety of mechanisms to curb volatility, such as circuit breakers, price controls, and price limits. In

particular, commodity futures markets primarily use price limits which restrict prices from rising above or falling below

prespecified levels. Limits can impact market microstructure, as well as affect real outcomes by restricting trading when

incentives and benefits to reallocating resources are potentially large.

Price limits in commodity markets have a long history that goes back to the first commodity futures exchange in the 18th

century: The Dojima rice exchange in Japan curbed the price movements of rice futures during a time when prices were

falling (Moser, 1990; Schaede, 1989). The stated purpose was to avoid excess volatility. Price limits were first introduced in

the United States during World War I when the New York Cotton Exchange restricted the daily price change of cotton

futures to three cents per pound on August 27, 1917 (Howell, 1934).

1

The Cotton Exchange’s justification for imposing limits

was ''to avoid abnormal fluctuations of price created by the European war, and injurious speculation incident thereto''

2

—a

sentiment that was shared by news reports which referred to ''abundant evidence of hoarding and speculation'' but ''no real

scarcity of raw materials.''

3

In 1922, the US Supreme Court ruled to support price limits:

''It was shown that a continually fluctuating, and not a stable, market is the desire of speculators. Such a market is

against the interests of the producer; he must have stable prices in order to market his crop to best advantage. A market

without wide and frequent price fluctuations would greatly benefit the producer.''

4

J Futures Markets. 2019;39:946–961.wileyonlinelibrary.com/journal/fut946

|

© 2019 Wiley Periodicals, Inc.

1

Soon after in December, limits were imposed in Hog markets. Grain markets had absolute price limits rather than limits on price changes.

2

Cotton Fluctuations Limited, Wall Street Journal, August 16, 1917.

3

The New York Times of August 19, 1917 notes: ''There is no real scarcity of the raw materials, but there is abundant evidence of hoarding or manipulation. Take the case of cotton, for instance.''

4

Senate Report No. 212, 67th Congress.

This brief history of price limits illustrates the view that limits are a tool to reduce volatility caused by speculation. In

this view, volatility is perceived to have a large transitory component caused by uninformed traders or speculators who

push prices away from fundamentals. Slowing down price changes provides informed traders an opportunity to enter

the market and steer prices towards fundamental values. Speculative activity is expected to decrease around limits:

Speculators reduce their long positions after limit ups and reduce their short positions after limit downs, thereby stop

pushing prices away to more extreme levels. Price changes are likely to reverse and volatility is likely to decline

following limit events (Ma, Rao, & Sears, 1989).

An alternative view is that futures price volatility mostly reflects news about fundamentals, which are persistent over

time. Having price limits in place merely postpones inevitable price changes and slows down price discovery. Because

limits dampen fundamental volatility rather than speculative volatility, they are not effective at restricting speculation.

In fact, price limits may even negatively alter the trading behavior of market participants, encouraging them to speed up

trading to avoid being closed out just before price limits occur (Kyle, 1988; Lehmann, 1989). Along this line of

reasoning, speculators should not change their trading behavior on limit days compared to nonlimit days. Following

limit events, prices are likely to continue in the same direction and volatility should not decrease.

The above views are not necessarily mutually exclusive: It is likely that volatility has both temporary and permanent

components. The net impact of limits on market participants, as well as the behavior of prices and positions, is an

empirical question. It is against this background that our paper attempts to answer three questions, using a large sample

of more than 5,000 limit events in 12 commodity futures markets over a 25‐year period.

The first question we address is whether price limits mitigate speculative activity of market participants. Using

positions data from the Commodity Futures Trading Commission (CFTC), we find direct evidence against reduced

speculation after price limits. Noncommercial traders, who are an important source of speculative capital, do not

change their long positions following limit up events. Instead, they significantly reduce their short positions. This

behavior, rather than decreasing, increases their net long positions following limit up events. Similarly, following limit

down events, noncommercials do not change their short positions and reduce their long positions, resulting in an

increase of net short positions. Our finding is robust to different regression specifications controlling for determinants

of noncommercial position changes.

The second question we ask is what causes limit events to occur? Price limits are more likely to occur when futures

price volatility is elevated. High volatility could be symptomatic of increased speculative activity or could reflect

changing fundamentals. We provide evidence that the elevated volatility does not appear to be associated with higher

speculation, as the behavior of noncommercial traders does not appear to lead to limits: Long and short positions of

noncommercial traders do not materially change before limit events. Instead, high price volatility appears to be related

to low levels of physical commodity inventories. Just before limit events, the front end of the futures curve is often in

steep backwardation. For storable commodities, a steeply backwardated futures curve reflects low inventories, which

lead to increased stock‐out risk (Deaton & Laroque, 1992). Higher stock‐out risk is associated with high futures

volatility, which in turn raises the probability that price limits are hit. Our results suggest that the conditions leading up

to limit events are fundamental in nature rather than speculative.

If limit events occur due to a shift in commodity fundamentals, any related price effects will likely persist. We

investigate price behavior around limits and find large and persistent price continuations across almost all commodities

we examine. We treat limit days as ''events'' and trace out average returns after these events and find large price

continuations do not reverse even after 2 weeks. Previous studies placed emphasis on the day immediately after limit

events (Evans & Mahoney, 1996, 1997; Hall, Kofman, & Manaster, 2006; Reiffen & Buyuksahin, 2010; Reiffen,

Buyuksahin, & Haigh, 2006). However, speculation could subside in the days following limit events, but not necessarily

on the day after. Tracking returns over longer horizons allows us to separate persistent price effects from return reversal

after several days.

Since price limits occur for fundamental reasons and do not stop speculation, do they affect price discovery in

futures markets? This is our third question. If market participants can easily switch between commodity futures and

other financial instruments, we may expect price discovery to migrate to other related markets. Options markets are

closely connected to the futures markets but currently do not have price limits. Historically, price limits on options had

been significantly looser than limits on futures and were almost never hit. Therefore, options provide an ideal

laboratory to understand price discovery around limit events and to gauge what unrestricted futures prices would be if

price limits did not exist. We explore this idea and compare option‐implied futures prices at the time of limit events to

subsequent prices when trading reopens in the underlying futures markets. We find that these option‐implied futures

prices are biased but informative predictors of subsequent futures prices. In forecasting regressions, a 1% increase in the

JANARDANAN ET AL.

|

947