Has Competition Been Good or Bad for Australia’s Securities Market?
Our analysis starts with an appreciation of the core mandate given to ASIC and virtually every other securities regulator in the world today including IOSCO: The reason I am referring to the regulator’s mandate is that they are the party assigned the responsibility by national governments to oversee market design changes. Their mandate is to ensure that markets and changes thereto pass the dual tests of fairness and efficiency. If this is their core mandate, it stands to reason that ASIC and other regulators need to be able to define and measure fairness and efficiency and look at the two attributes pre and post the introduction of competition. If fairness and efficiency go up in the wake of a particular market design change then the answer is that the change was good for the market. If one attribute goes up and the other remains constant then I still believe the answer is positive. If however, one attribute goes up while the other goes down the answer is not so clear cut and the change should be rejected or amended.
To begin to address the regulatory mandate we need to start with definitions of fairness and efficiency. I define an efficient market as one in which it is cheap to trade and one in which the price reflects all available information. The cheaper, the more efficient the market. The quicker the information is impounded in prices, the more efficient the market becomes. Following from this definition, one needs to measure the cost of trading and the speed with which information is impounded in prices.
If it gets cheaper to trade after the introduction of competition and information is more quickly impounded in prices, then I would argue that competition had a positive impact on the efficiency of the market. If it increased one but left the other unchanged, I would argue that we are still better off. If one goes up and the other goes down as a result of competition, I would argue that the change hasn’t had the necessary impact on efficiency to pass the regulator’s test of enhancing efficiency.
While we can look to the impact of the same type of change in other markets (such as the introduction of Chi-X in London) the results there are not a perfect indication of the likely changes in Australia as the London market is the net result of hundreds, perhaps thousands of previous market design changes, all of which could lead to a different outcome when they interact with competition. In point of fact, the results from London suggest that Chi-X should both reduce transaction costs and make price discovery more efficient.
I define fairness (hereafter used interchangeably with integrity) as the extent to which market participants engage in prohibited trading behaviours. These prohibitions are defined in law, just like the regulator’s mandate, and most laws enshrine three types: insider trading, market manipulation and broker-client conflict (such as front-running).
None of these are directly measureable but we can use proxies to estimate them. I am not alone in arguing that information leakage is a good estimate of insider trading. It is unusual price and volume behaviour prior to a market sensitive news announcement, and is an upper limit for insider trading because not all information leakage is illegal. One needs to distinguish between expected and unexpected news announcements (e.g., between earnings and takeover announcements) with the later more likely to reflect insider trading.
After the introduction of competition, if metrics that represent insider trading, market manipulation and broker-client conflict go down or at least stay constant, then one might argue that the introduction of competition had not had a negative impact on market integrity. With a positive change to efficiency and no change to integrity, I would argue that the change has been good for the market. The problem arises if one of either Integrity or efficiency goes up and the other goes down. How might that happen? Again, we can look to London for salutary lessons. It was reasonably clear from the evidence that was supplied by academics in the wake of the introduction of Chi-X that efficiency went up. While no evidence was supplied in respect of integrity, one might reasonably hypothesise that integrity would go down because there was no centralised surveillance authority established in Europe prior to the introduction of competition. This meant that parties that previously had to trade in one market could now split their orders across multiple markets. Insider trading is a game of “use it or lose it”. The new market design made it easier to take smaller positions across different markets and harder to detect without centralized surveillance. It was now also possible to manipulate prices in one market in order to trade in another, and with no oversight of the multiple markets one might conclude that we gave up integrity for greater efficiency. This seems to me a sub-optimal position and unfortunately, it still describes the situation today.
Back to Australian competition. How might we know whether competition has been good for Australia (or any other marketplace) or bad? Or for that matter, high frequency trading, dark pools, short sale bans, circuit breakers or any other market design change? The answer is that we need to measure the proxies for efficiency (transaction costs and price discovery) and integrity (insider trading, market manipulation and broker-client conflict) pre and post the relevant market design change, leaving enough time for the change to take effect, probably 3 months.
The Capital Markets Cooperative Research Centre (CMCRC) has spent 10 years and A$40 million putting together the infrastructure necessary to measure the relevant metrics for all markets in the world over the last decade. Table 1 (on efficiency) and Table 2 (on integrity) provide an indication of where Australia sits among the world’s exchanges using one proxy for efficiency (transaction costs) and one for integrity (insider trading). These tables show Australia is currently ranked 14th in the world on efficiency and 12th on integrity with numerical estimates of each. On efficiency, at 14 (approximately 24 basis points), Australia is 4 times less efficient that the most efficient market in the world. It is 8 times less fair than the fairest market in the world with insider trading estimated at US$147 million (or approximately 5/100ths of 1%) per quarter in Australia. As an aside, note the value of insider trading as a % of turnover in the Australian and other markets in table 3. At just under 5/100ths of 1% (and less for other markets) I suggest that this does not fit with folklore about the extent of insider trading. This is s another reason why more objective evidence on such issues is urgently needed.
Has competition been good or bad for Australia’s securities market? We’ll take a first pass at answering the question with these metrics and how they change in the wake of the introduction of competition. We would want to see both efficiency and integrity improve, or one go down without affecting the other. Watch this space for the results.
(1) See http://www.asic.gov.au/asic/pdflib.nsf/LookupByFileName/cp-145.pdf/$file/cp-145.pdf. This is the consultation paper on Australian Equity Market Structure released by ASIC in November of 2010 in response to the introduction of competition.