Abstract
This thesis examines the impact of closing price manipulation (“marking the close”) on several dimensions of equity market quality, including both the efficiency and integrity of equity markets. Marking the close is an illegal practice that involves attempting to influence the closing price of a stock by aggressively executing orders at or near the end of the trading day. Although this practice affects a security’s market price over only a short period of time, often seconds before the close of business, it has significant harmful consequences because of the importance and widespread use of equity closing prices. For example, many financial institutions use the closing price as a reference price to calculate investment fund performance over a given time frame, or to determine fund manager compensation, or to determine the price or payoff to a derivative security.
Current understanding of marking the close manipulation is poor due to the scarcity of data relating to demonstrated cases of manipulation. This thesis describes the curation of a dataset containing all prosecuted closing price manipulation cases reported by the U.S. Security and Exchange Commission (SEC) for the period 1995 to 2018. This dataset is subsequently used to characterise manipulation, investigate how manipulation affects trading, and investigate the overall motivation(s) for manipulation. The results show that the majority of prosecuted and confirmed cases of manipulation were relatively illiquid small companies with manipulation often occurring to satisfy requirements to remain listed on an exchange and/or to avoid margin calls.
Following the characterisation of market manipulation, this thesis investigates the impact of regulation on market quality in both liquid (U.S.) and comparatively illiquid (New Zealand) markets.
In U.S. equity markets, a natural experiment was performed to investigate the impact of the 2011 SEC naked-access ban (Rule 15c3-5) on high-frequency trading (HFT) participation, market efficiency, and market integrity. The resulting insights shed new light on previous understanding of HFT as one of the largest drivers of efficiency and integrity of the U.S. equity market. The results dispel concerns that HFTs harm market quality by creating an uneven playing field and potential for market manipulation. On the contrary, higher HFT participation in the U.S. markets appears to lead to overall higher market integrity and more efficient markets. HFT activities, especially arbitrage strategies of trading against manipulators, increase the cost of manipulation by providing liquidity, minimising their trade impact in the market, and significantly decreasing adverse selection.
To contrast the behaviour of the liquid U.S. markets, the thesis examined the New Zealand Stock Exchange (NZX) trading characteristics and liquidity over the period 2007-2018. This analysis has shown some important, and possibly unique, characteristics of the NZX and the possibility of closing price manipulation associated with large negotiated off-market trades. The thesis highlights that approximately two-thirds of the total dollar turnover of the NZX in the past 10 years is negotiated off-market trades and that it is common for off-market trades to be executed at the closing auction price once that price is determined. Off-market volume is not typically part of the closing auction mechanism and, in many cases, the volume of off-market trades executed just after market close is much larger – sometimes 100 times larger – than the volume absorbed by the closing auction. Therefore, an off-market trader who wants to trade a large volume at the closing auction price could plausibly manipulate the thin pre-close auction using relatively low volume so that their large-volume off-market order is executed at the manipulated closing price just after the close.
To measure the possibility and breadth of this kind of closing price manipulation, a rule-based manipulation index was developed based on the Financial Markets Authority (FMA) v Warminger judgment report (New Zealand’s first and, at present, only manipulation case to come to trial). The Warminger judgment report provides detailed insights regarding the machinery of the exchange surveillance systems and the regulators’ investigation process in the New Zealand context. This information became available to the public for the first time as a result of this case. Using the proposed manipulation index, trading in the S&P/NZX 50 component stocks over a two-year period was analysed to measure the possibility and breadth of this type of manipulation on the NZX. This study presents the first detailed empirical analysis of stock price manipulation in New Zealand markets. Feedback from both the NZX and the FMA suggests that the approach used in this thesis is based on the rational assumptions, effective in terms of structure, and arguably the best possible method for replicating a stock market surveillance system using publicly-available data.