It might surprise some people to learn that more than a third of turnover in shares in the US now take place in the “dark” rather than on a public stock exchange. In Australia, at one point about 30 per cent of trading volume was outside the “lit” market.
New rules have scaled dark trading back both here and in Canada, while European regulators are about to place their own limits on what trades can and can’t take place in the shadow of the traditional public sharemarket.
So what is dark trading, and should we be concerned about its growth?
“Dark trading is high on the agenda of many regulators,” says Professor Talis Putnins, of the Finance Discipline Group at UTS Business School, who is a leading researcher in the area. “There’s no doubt it’s topical – some would say controversial.”
That’s partly because of the sheer scale and pace of its growth – which has raised question marks about whether it has moved too far away from its original purpose – but also because of recent headlines. Dark pool operators in the US found to have misled clients or misused client information have incurred multimillion-dollar fines.
In addition, dark trading has become conflated in the headlines with high-frequency trading (HFT), which uses extremely sophisticated and high-speed computer programs to generate automated trades that occur in milliseconds. While this can increase efficiency and liquidity, it can also make for more volatile markets and, in extreme cases, what is known as a “flash crash”.
Professor Putnins explains that on a conventional, public stock exchange – the “lit” market – there is high visibility of supply and demand. The “lights are on” as far as who’s offering to buy and who’s offering to sell, and at what prices. The market is open to everyone, and all buyers and sellers are there because they want to trade.
Dark pools, however, may not be open to all, offering exclusive access only to some traders. There is no pre-trade transparency: “You see nothing about the trading intentions of potential counter-parties until the trade is executed,” he says.
Sometimes parties “ping” a dark pool with small trades not with the clear intention of transacting but to obtain information before they go to the lit market.
So how do parties agree on a price? Prices are taken from the lit market, he explains, often using the midpoint of the spread between the best bid (buy) price and the best ask (sell) price.
Trades that go ahead become part of the public record, possibly with a bit of a delay, he adds.
One of the main reasons institutional investors use dark pools is so they don’t show their hand, unduly causing the market to move, when buying or selling large blocks of securities.
There has been substantial growth in the market share of dark trading in the past 10 years, Prof Putnins says. That’s been driven by technology, the regulatory view that competition is good, and investor demand because of dark trading’s privacy, flexibility and lower fees.
But in parallel with that growth occurring, trade sizes on the dark market have been declining – suggesting that dark trading isn’t always being used for its original role as a facility for executing very large trades.
“Dark trades used to be substantially larger than on the stock exchange, but today they’re almost identical to the average trade size you’ll see in lit markets,” Professor Putnins says.
Why is that concerning? Because some of the orders that would have gone to the public stock exchange are now being executed in dark markets, he says, and that affects liquidity on lit markets such as the Australian Securities Exchange.
It could also inhibit “price discovery” on the public market – in other words, making it harder for investors to determine the right price for a stock based on factors such as supply and demand.
Such concerns led to changes to the rules governing dark trading in Australia in 2013, when the Australia Securities and Investments Commission (ASIC) decreed that dark trades below the size of a “block trade” would have to build in “price improvement” on lit market prices. (A block trade was redefined at the same time as either $1 million, $500,000 or $200,000, depending on the liquidity of the stock in question). In other words, those who chose to trade in the dark would have to pay a slight premium for the privilege.
As a result of these changes the market share of below-block-size dark trading fell virtually overnight from 18 per cent to 11 per cent of overall share trade volume. This decrease is seen as good news by Australian regulators, who cite research co-authored by Professor Putnins that shows high levels of dark trading can harm markets, while low levels have neutral or even beneficial effects.
The European Union’s reforms of dark trading include a rule that caps it at 8 per cent of the overall market.
While the decrease in the level of dark trading is consistent with the objectives of the Australian regulatory changes, some other hoped-for results did not materialise. There was no increase in liquidity on the public sharemarket, for instance.
Professor Putnins says this is because the measures had the unintended consequence of moving many trades from one part of the dark market to another. “Not all dark trading has the same effects on market quality,” he says. “The new rules increased the share of dark trading at the bid-ask midpoint at the expense of some beneficial types of dark trading.”
Two years on from the changes, ASIC has released a review of high-frequency trading and dark trading which concludes its reforms were “largely adequate and effective”, directing dark trade back towards its original purpose. “There are fewer small dark trades and they are now fairer,” the review says.
However, ASIC also noted “inconsistent practices” for managing confidential information and conflicts of interest in dark trading and signalled that its “surveillance focus on predatory trading will continue”.
While no further regulation was required, it said, “we will continue our forward-looking review of the purpose of market and their fundamental role in an environment of rapid change.”