Trading across multiple markets

By Scot Blythe | June 3, 2009 | Last updated on June 3, 2009
5 min read

Not so long ago, only a decade, retail investors were wringing their hands about outages on the Toronto Stock Exchange, which sometimes failed to provide high-speed access for day-trading in Nortel stock. They envied the U.S., with its multiple exchanges, including electronic communications networks (ECNs) that provided trading in Nasdaq stocks.

Ironically, the Nasdaq stocks that propelled day-trading never really recovered. And Nasdaq faltered too, as ECNs ate into its volume. In the end, Nasdaq became, in competition with the New York Stock Exchange, a major acquirer of stock exchanges, in the U.S. and in Europe. For its part, the NYSE has also stepped up to the acquisition plate, not only in Europe, but in buying its own ECN.

Now, electronic execution of trades has finally taken purchase in Canada, years after it transformed the American capital markets landscape. Five competing alternative trading systems (ATSs) are vying for stock volume with the TSX, along with an institutionally focused “dark pool” of liquidity that facilitates trading among pension and mutual funds of large blocks of shares.

The annual FPL Canadian Electronic Trading conference in Toronto, now in its fifth year, is a barometer of these changes, and the speed at which they are happening. Five years ago, there were no equity ATSs in Canada, and the NYSE conducted trades primarily on a trading room floor, to take two examples.

But the key feature is not just electronic execution; Canada was a pioneer in replacing open outcry auction on the floor of a stock market, and along with it, phoning in an order to a floor broker. Instead, it is the emergence of alternative marketplaces where speed of execution and the prospect of lower trading fees is altering a trading landscape that remained almost unchanged for two centuries — until the beginning of the decade, says Tom Caldwell, chairman and CEO of Caldwell Financial and a keen investor in stock exchanges.

Now he suggests that Microsoft ought to consider buying a stock exchange. It’s no longer about listing fees, although a primary listing market for publicly traded stocks remains essential, says Randee Pavalow, a former director of capital markets at the Ontario Securities Commission and now head of business opaerations at Alpha Group, which runs an ATS. But she expects that the Toronto Stock Exchange’s share of trading to drop to 50% as alternative venues such as Pure, Alpha, Chi-X and Omega gain momentum.

Nor is it about trading fees. Canada once had amongst the highest exchange fees in the world; now it is among the lowest, thanks to price competition from the alternative marketplaces. With the advent of new marketplaces, he notes, “as competition was working its magic, it was driving commissions down.” But marketplace operators failed to notice that because volumes were rising, often thanks to “high-velocity” traders — former professional traders acting a bit like hedge funds with rapid-fire trades to take advantage of price discrepancies across markets.

While the higher volumes breed liquidity, which in turn calls forth more liquidity, as traders find it easier to fill orders, the old mutual, non-profit ownership structure no longer suffices to maintain a franchise. Indeed, there was a time, Caldwell says, when the TSX gave its market data away for free.

Now selling market data is part of a diversified set of revenue sources, which include listing fees and trading commissions but also derivatives and fixed income trading. These, he says, feed each other’s volume.

Still, market data consolidation remains a problem in Canada. “We find providing data doesn’t make it accessible,” says Pavalow.

While ATS’s in Canada are trying to take advantage of a market opportunity, they also complicate the regulatory picture. Retail investors — and the brokers who serve them, be they discount or full service — have operated on one basic principle: best price.

That is enshrined in the trade-through rules: brokers should seek the best price for clients. In the U.S., that meant moving among various market-makers and ECNs, in view of Regulation NMS, the idea that there was one stock market, but many venues where stocks were traded. And, indeed, to fulfill Regulation NMS, the 10 formal, self-regulating exchanges in the U.S. send client orders, through smart order routing, to the venues where they can be best executed — that is, where there is the best bid and offer.

As those concepts migrate north, just as alternative market venues have, they have put pressure on Canadian investment dealers. On the regulatory front, the trade-through rules clash with new norms of best execution, tailored to institutional investors. For an institutional investor, like a mutual or pension fund, best execution encompasses a variety of dimensions, not just best price, but speed and certainty of execution, the cost, the market impact, as well as a number of other factors.

Best execution, following the U.S. experience, has come to Canada. Still, the trade-through edict complicates things: how to keep best price foremost, yet take into account all the rest of the investor’s need, mostly on the institutional side. Sonali GuptaBhaya, legal counsel for market regulation at the Ontario Securities Commission, contrasts the two this way: “Best execution is a fiduciary obligation to the client,” she says, while “trade-though is an obligation to the marketplace.”

But more complications follow. Best execution will probably require that investment dealers are linked to as many ATSs as possible. In Canada, the big-bank dealers have already done that, says consultant Tim Thurman. But, of the 125 or so investment dealers in Canada, only 75 have done so. It’s too expensive for the rest. As a result, some see the boutique houses and big brokerages surviving, with a more opaque future for the mid-tier.

And that heralds a new era of consolidation, and, at the same time, specialization. Brokers will have to decide whether they are better at research, or execution of trades, or can do both. As the buy-side undertakes transaction cost analysis, it will become more discerning of who does what best.

Meanwhile, block-trading, which used to be a major threat to the TSE, as institutional investors had their trades internalized by the bank investment dealers — the dealers executed the trade against their own inventory, rather than sending the order to the trading floor — has changed in nature.

For one thing, there are now “dark pools” of liquidity, such as Liquidnet, price and depth are not displayed on public or “lit” markets, but instead are filled in a crossing network, where bid and offer are matched, and then reported after the trade. That takes away from price discovery — a part of the market is hidden from subscribers to real-time data feeds.

But such “upstairs” trading seems to be in decline, as algorithmic trading, facilitated by both a multiplication of stock-trading venues and the refinement of smart order routers that send trades electronically to the most attractive marketplace — in milliseconds — chop up a trade into tiny sizes. Five years ago, the average trade size on the TSX was 1,800 shares. It is now 600.

For all that, “there’s nothing new under the sun,” says Jeromee Johnson, vice-president market development at BATS trading, an upstart electronic exchange that in two-and-one-half years has come from nowhere to capture 10% of all equity trading volume in the U.S. “The way people trade today is not fundamentally different from five years ago.”


Scot Blythe