Sentana and Wadhwani published a seminal study in 1992 examining the presence of feedback trading in stock markets and the extent to which such behaviour is linked to the level of arbitrage opportunities. They found that when volatility is low, daily (and hourly) returns exhibit positive autocorrelation, but when it is high, returns exhibit negative serial correlation. They also found an important asymmetry – negative serial correlation is more likely after price declines which is consistent with price declines being more likely to induce positive feedback trading

Recent work has applied their findings to the context of Carbon emission, Coal, Oil and Gas energy markets in Europe.  Most notably, Chau, Kuo and Shi (2015) find evidence of feedback trading in coal and electricity markets, but not in carbon markets where the institutional investors dominate.

This finding is consistent with the notion that institutional investors are less susceptible to pursuing feedback-style investment strategies. The most recent work shows that the intensity of feedback trading is significantly related to the level of arbitrage opportunities, and that the significance of such relationship depends on the market regimes.


Economists have long debated the impact of feedback traders on equilibrium market prices, especially after the dramatic rises and falls of stock markets in recent years. Some argue that their existence is destabilising, causing inefficiency and instability in asset prices whilst others recognised that the presence of trend-following investors can be beneficial as they provide market participants with liquidity. Numerous papers have been devoted to the study of feedback trading activities in global markets with the literature being focused primarily on positive feedback strategy whereby investors buy (sell) when prices rise (fall), i.e., chasing the trend.

Evidence of this type of behaviour is found in both individual and institutional investors and also in a wide variety of markets. When it comes to commodity markets, however, there is no clearly identified evidence of the feedback trading, despite the increasing use of commodities as an investment tool by the fund industry.

In recent years, with the historically low interest rates and meltdowns of financial markets, many institutional investors and portfolio managers had turned to commodity markets as a way of meeting their investment objectives and, to a lesser extent, as a means of controlling risk.

“Investment fund activity in commodities is currently at 330 US$ billion…9 times higher than a decade ago, when this activity started becoming a popular investment vehicle within the financial community.”

Despite the growing popularity of commodity markets in strategic asset allocation, scarce evidence exists in the extant literature on the trading behaviour of commodity investors, and in particular there is little research examining the presence of feedback (trend-following) behaviour in these important markets. This is somewhat surprising given the nature and design of commodity futures markets (i.e., the low cost of trading, absence of short-sales constraints, and high leverage opportunity) can appeal to several feedback-style investment strategies such as portfolio insurance, short selling, and margin trading.

Previous empirical investigations have generally assumed the behaviour of feedback traders is not impacted by the level of arbitrage opportunities in financial markets. However, it is widely recognised that arbitrage activities and rational speculation are among the most significant factors contributing to feedback trading and there is growing evidence that the arbitrage opportunities – as measured by the spot-futures basis or convenience yield – have a predictive value in future price variations. (Gorton et al., 2013). Thus. it seems overly restrictive to assume that the behaviour of feedback traders is unaffected by the level of arbitrage opportunities.

Noise Trader: The term used to describe an investor who makes decisions regarding buy and sell trades without the use of fundamental data. These investors generally have poor timing, follow trends, and over-react to good and bad news

Literature review

Whether noise trader in general and feedback trader in particular affects prices is a question of long-standing interest to economists. Shiller argues that social norms or fashions can influence asset price movements and Black introduces the concept of noise traders and offers a formal definition of ‘noise trading’ as trading on noise (or non-information) as if it were information. Sentana and Wadhwani (1992) demonstrated that trading between rational arbitrageurs and feedback traders gives bubble-like patterns where positive feedback traders reinforced by arbitrageurs’ jumping on the bandwagon leads to positive autocorrelation of returns at short horizons. Eventual return of prices to fundamentals, accelerated by arbitrage, entails a negative autocorrelation of returns at long horizons. Since news results in price changes that are reinforced by positive feedback traders, prices overreact to news and exhibit excessive volatility of a destabilising fashion. Moreover, they also found an interesting result that returns switch from being positively autocorrelated to negatively autocorrelated as volatility increases, predicting a negative relationship between volatility and autocorrelation.

In subsequent investigations – and consistent with the existence of positive feedback traders – a negative relationship between autocorrelation and volatility has also been found to be the feature of returns in both mature and emerging stock markets, foreign exchange markets, index futures markets , ETF markets and crude oil market.

Feedback trading can be the result of various motivations. De Long (1990) argues that rational speculation and arbitrage are among the most important factors contributing to feedback trading. Interpreted within the context of futures markets, spot-futures arbitrage is a trading strategy that rational investors pursue to profit from the deviation of futures price from its underlying spot price (Chung, 1991). This is also the central mechanism in maintaining the linkage between two markets and to contribute to price discovery.

When spot-futures basis (a widely used signal for arbitrage opportunity) increases beyond a threshold level, arbitragers can simultaneously buy futures and sell the underlying asset to benefit from these price deviations (Kumar and Seppi, 1994). Intuitively, to the extent that the presence of arbitrage opportunities motivates more investors to trade, the level of feedback trading is also expected to increase as a result of enhanced rational speculation and arbitrage activities. While the profitability of arbitrage (Chung, 1991) and spot-futures mispricing (McMillian and Philip, 2012) have been extensively studied, the issue of whether arbitrage opportunities affect feedback trading activities is yet to be explored.

Positive Feedback : A self-perpetuating pattern of investment behavior. The herd mentality that causes investors to sell when the market is declining and buy when it’s rising is an example of positive feedback. Positive feedback is the reason why market declines often lead to further market declines and increases lead to further increases. It is also a source of market volatility. When a cycle of positive feedback continues for too long, it can create an asset bubble or a market crash

Feedback and arbitrage in commodity markets

In recent years the commodity markets have become increasingly important in tactical asset allocation. The trading strategies of commodity investors attract considerable attention in academic research. Miffre and Rallis (2007) shows that both momentum and contrarian strategies are profitable in commodity markets and discussed at length on ‘Trading on mean reversion‘ on this website. Marshall et al. (2008) suggests that certain technical trading rules can generate abnormal returns in commodity markets. However, to date, there exists a limited research on feedback trading strategy in commodities. Cifarelli and Paladino (2010) examines feedback trading in the U.S. crude oil markets using weekly data, but the vast of majority existing research utilises either daily or intraday prices. There is a benefit of using a high frequency dataset because feedback traders usually adopt short-run computerised strategies to capture the observed trends which tend to vanish quickly and the use of weekly data may fail to detect these feedback trading activities.

Growth in European electricity trading
Growth in European electricity trading

Motivated by the forgoing discussion, Chau, Kuo and Shi (2015) sought to examine the presence of feedback trading in carbon emission and energy futures markets and the extent to which such behaviour is linked to arbitrage opportunities. Specifically, they developed and estimated several feedback trading models in which the behaviour of feedback traders is conditional on the level of arbitrage opportunities.

The latest results show that there exists significant feedback trading in coal and electricity markets, but not in carbon, natural gas and crude oil markets. As the vast majority of investors in carbon markets are institutions, these results are consistent with the notion that institutional investors are not particularly susceptible to feedback-style investment strategies. In further analysis, they show that the intensity of feedback trading is significantly related to the level of arbitrage opportunities, and that the significance of such relationship varies across market regimes.

Overall, the results show that the degree of positive (negative) feedback trading decreases (increases) as the lagged basis becomes widen. The lagged spot-futures basis provides useful indicator for feedback trading, who can use this as a signal of ‘channel breakouts’ in technical analysis. When the basis is within certain thresholds, feedback traders expect that the current trend of futures prices will persist and adopt a trend-following positive feedback trading strategy. However, if the basis is wide enough, the current channel will be broken out by arbitrageurs; as a result, negative feedback trading becomes more profitable.

This is consistent with the findings of Marshall et al. (2008) who concludes that channel breakouts trading rules are consistently profitable in the U.S. commodity markets; supporting the argument that many rational arbitrageurs tend to jump on the bandwagon themselves before eventually selling out near the top and take their profit.


Taken together, the latest findings add to the body of literature that studies the role of behaviourally biased feedback trading in commodity markets and the effect of arbitrage on such investment behaviour. The work of Chau, Kuo and Shi highlights investors’ trading behaviour and investment strategies in commodity markets, particularly on the newly opened carbon emission market where they find no evidence of feedback trading and that arbitrage opportunities do not affect the demand by feedback traders.

Moreover, the results suggest that arbitrage opportunities have a significant influence on feedback traders’ demand in electricity and natural gas markets. This finding is consistent with the view that the behaviour of feedback traders tend to vary depending on the level of arbitrage opportunities in these markets. Additional analysis indicates that the response of feedback trader to past return or arbitrage opportunities depends on market regimes.

There are however some important issues remain in the extant literature that requires further investigation. For instance, future research in this area may seek to identify the reasons why feedback trading is found in coal and electricity, but not in oil and gas.


  1. Sentana and Wadhwani, 1992 – Feedback traders and stock return autocorrelations: evidence from a century of daily data – Econ. J., 102 (1992), pp. 415–425
  2. Chau, Kuo and Shi (2015) – Arbitrage opportunities and feedback trading in emissions and energy markets – Journal of International Financial Markets, Institutions and Money
  3. De Long et al., 1990 – Positive feedback investment strategies and destabilizing rational expectations – J. Financ., 45 (1990), pp. 379–395
  4. Gorton et al., 2013 – The fundamentals of commodity futures returns – Rev. Financ., 17 (2013), pp. 35–105
  5. Chung, 1991 – A transaction data test of stock index futures market efficiency and index arbitrage profitability – J. Financ., 46 (1991), pp. 1791–1809
  6. Marshall et al., 2008 – Can commodity futures be profitably traded with quantitative market timing strategies? – J. Bank. Financ., 32 (2008), pp. 1810–1819

I'd love to hear what your thoughts are...please feel free to leave a reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.