Fabrizio Lillo,J. Doyne Farmer
Abstract:For the London Stock Exchange we demonstrate that the signs of orders obey a long-memory process. The autocorrelation function decays roughly as $\tau^{-\alpha}$ with $\alpha \approx 0.6$, corresponding to a Hurst exponent $H \approx 0.7$. This implies that the signs of future orders are quite predictable from the signs of past orders; all else being equal, this would suggest a very strong market inefficiency. We demonstrate, however, that fluctuations in order signs are compensated for by anti-correlated fluctuations in transaction size and liquidity, which are also long-memory processes. This tends to make the returns whiter. We show that some institutions display long-range memory and others don't.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the long - memory characteristics of the financial market and its impact on market efficiency. Specifically, the author studied the order flow in the London Stock Exchange (LSE) and found that the sign of the order (buy or sell) exhibits significant long - memory characteristics. This characteristic means that past order signs have a significant impact on future order signs, which theoretically leads to the predictability of market prices, thus challenging the market efficiency theory.
### Main problems:
1. **Long - memory characteristics of order signs**: The paper shows that the order signs (buy or sell) in the London Stock Exchange have long - memory characteristics, that is, the time series of order signs shows a power - law autocorrelation function \(\tau^{-\alpha}\), where \(\alpha\approx0.6\) and the corresponding Hurst index \(H\approx0.7\). This means that future order signs can be predicted relatively accurately from past order signs.
2. **Contradiction of market efficiency**: If the order signs have strong long - memory characteristics, then the changes in market prices should be predictable, which contradicts the market efficiency theory. The market efficiency theory believes that market prices reflect all available information, so there should be no significant predictability.
### Solutions:
To reconcile this contradiction, the author further studied how other market mechanisms compensate for the long - memory characteristics of order signs to maintain market efficiency. Specifically:
- **Reverse adjustment of trading volume and liquidity**: The author found that when buyer orders are more likely, the ratio of the relative size of buyer market orders to the trading volume at the best ask price will become smaller, which means that the probability that a given buyer order penetrates the best price and triggers a price increase will be reduced. Similarly, when seller orders are more likely, the opposite is true. These adjustments make price changes more random (i.e., "white noise"), thus maintaining market efficiency.
- **Differences in the behavior of different institutions**: The author also analyzed the order behavior of different institutions and found that some institutions exhibit obvious long - memory characteristics, while others do not. This indicates that long - memory characteristics may be related to specific trading strategies.
### Conclusion:
Through detailed data analysis and statistical tests, the paper proves the long - memory characteristics of order signs and explains how the market compensates for this characteristic through other mechanisms (such as the adjustment of trading volume and liquidity) to maintain market efficiency. This research provides a new perspective for understanding the complex dynamics in the financial market.