The 25th anniversary of Inquire Europe was celebrated at the Hotel Grande Bretagne in Athens (see image above) on 4-6 October 2015 with a seminar on “Liquidity and Trading” which brought together 85 participants including practitioners and academics.
Markets are neither efficient nor inefficient. They are both!
The key note speaker was Lasse Pedersen from the Copenhagen Business School, New York University, Centre for Economic Policy Research and AQR Capital Management, who presented his recent book “Efficiently Inefficient” where he defends the view that markets are neither efficient nor inefficient but simply fall in between. In his view markets must be inefficient enough so that fund managers can be compensated for their costs through the profits of their trading strategies but at the same time efficient enough so that the profits after costs do not encourage additional active investing. His thesis is built upon previous work from Grossman and Stiglitz who defended that markets cannot be fully efficient since then no one would have an incentive to collect information, making it logically impossible for both markets and asset management to be fully efficient.
Illiquidity and information as sources of performance
In Lasse’s view, beating the market can be achieved either via compensation for liquidity risk or via compensation for information. Compensation for liquidity risk may arise from market liquidity risk, funding liquidity risk (leverage) or providing liquidity to demand pressure whereas compensation for information may arise from production of information, access to information or slow diffusion of information.
Pierre Collin-Dufresne from the Swiss Finance Institute at the Ecole Polytechnique Fédérale de Lausanne investigated the trades of identified informed trades by activist shareholders, e.g. Icahn Capital LP plans to force the breakup of Chesapeake’s board and the installation of new directors nominated by Icahn Capital LP and other leading shareholders. In the US such trades must be filed with the SEC (13D fillings) within 10 days of acquiring more than 5% of any class of securities of a publicly traded company if the buyer has an interest in influencing the management of the company. Pierre found that such informed trades are significantly profitable and that activists implement such trades over a number of days where liquidity is abundant and use limit orders to reduce the adverse market impact.
Illiquidity based strategies
Philippe Mueller from the London Business School discussed how measures of illiquidity based on noise in government bonds yields can be used not only to explain stocks returns but also to create profitable strategies. In particular he finds that the stock alpha increases with local illiquidity and decreases with global illiquidity. In turn, Nils Friewald investigated corporate bonds and found that simple strategies based on the holdings in dealer inventories can earn significant positive risk-adjusted returns. The limited capacity of dealers to bear risk and share risk are key determinants for the risk they take in their inventory and, thus, for bond prices.
High frequency and algorithmic trading
Robert Korajczyk from the Northwestern University found that high frequency traders (traders using electronic trading platforms to trade in milliseconds or even microseconds, typically buying and selling many securities per second) supply liquidity to markets, contributing towards market making activity much like market making dealers, but tend to back off from stressful trades. Curiously he also found that high frequency traders, at least in Canada, lose money on their trading activity. It seems that the profits of high frequency traders in Canada derive essentially from liquidity rebates (commissions earned on each trade) which tend to exceed the losses on their trading activity.
In turn Elvira Sojli from the Erasmus University of Rotterdam investigated algorithmic trading, which may or may not be high frequency. Market making algorithms provide liquidity and lower trading costs. Elvira found significant positive abnormal returns in stocks with lower number of quotes when compared to the number of actual trades executed, i.e. stocks for which when investors place an order and get a quote, more often than not they end up pursuing the transaction.
Dark pools versus lit markets
Increasingly institutional investors use dark pools for trading securities. A dark pool is a private electronic network for trading securities where trades remain confidential and outside the purview of the general public. Trading in dark pools is often offered by broker dealers but there are some independent venues too. Lit markets are the opposite of dark pools. Primary exchanges are examples of lit markets.
Private trading networks are often criticised for lack of transparency. Some worry that the presence of dark venues reduces the incentives for liquidity provision in lit markets, potentially widening bid-ask spreads and threating business interests of public exchanges. Ingrid Werner from The Ohio State University addressed this question. She developed a theoretical model to capture the key features of the real market and allow traders to use different types of trading venues. Her results suggest that when traders have access to a dark pools the lit market bid-ask spreads tend to widen and investors restricted to trading in lit markets face higher trading costs as a result.
Hans Degryse from Ku Leuven investigated to what extent investors can use hidden orders on lit trading venues instead of dark pools since both may act as substitutes. He finds that dark pools appear to be a better substitute for hidden order trading in lit markets than the other way around. In particular he found that hidden order trading is preferred over dark pool trading on high volume days, when visible depth is small, the quoted spread is narrow and fewer traders employ smart order routers. Hidden orders in lit markets do not offer an adequate opaque alternative to dark pools.
Trading venues are sorted from dark pools with low cost (narrower bid-ask spread, smaller price impact and information leakage) but low immediacy (slow execution), to lit markets with high cost but high immediacy. Bart Zhou Yueshen from INSEAD investigated the “pecking order hypothesis” predicting that investors prefer to trade in low cost and low immediacy venues but move trading into higher-cost and higher-immediacy venues when their trading needs become more urgent. This hypothesis was tested in U.S. markets and the results seem to confirm it.
Lasse’s thesis keeps the door open for active management. As others put it before, the efficient market hypothesis entails a paradox as the market can only be efficient if enough investors believe the market is inefficient and trade according to their beliefs of fair prices. Lasse goes further defending that the markets reach the equilibrium between efficiency and inefficiency where active investors are not discouraged to leave but not everyone can win. And if that is not the case then markets breakdown. But investors and active managers should worry about implementation all the way down to trading. The trade-off between low cost and immediacy in their investment strategies is an important factor for the way their trades are implemented. Finally, investors should pay attention to how fund managers aim at beating the market, in particular to what extent their strategies derive performance from information advantage or exposure to illiquidity.