MiFID II Intended and Unintended Consequences: What If Size Just Doesn't Show Up?

By Larry Tabb, TABB Group

Originally published on TABB Forum

Under MiFID II’s high-frequency trading rules, the EU is moving away from a faster and more electronic market. Restrictions on trading algorithms and greater risk controls may be aimed at increasing transparency and reducing risk, but they are likely to force participants to trade in larger size even as liquidity providers retreat.

The impact of high-frequency trading rules under MiFID II will certainly be interesting as the regulation layers in risk controls, cancellation restrictions, market maker liquidity requirements and costly compliance regimes. Clearly, the European Union is moving away from a faster and more electronic market while developing incentives for trading in size; but how will these rules really impact European markets?

The impact of greater risk controls, depending on the tolerance levels, could be significant or not. Market access and pre-trade risk controls should be fine; but stringent order-to-trade ratios, on the other hand, could make providing liquidity risky and more costly. If order-to-trade ratios get too restrictive, liquidity providers will need to quote in wider increments.

Increased oversight of algorithms should push a segment of the market toward more vanilla algos. But it would be odd to see less-sophisticated liquidity-providing algos, given that more-sophisticated algos tend to reduce risk and generate more revenue. As a result, depending on the level of scrutiny, vanilla algorithms likely would be used only for buy-side trading strategies.

So greater algorithmic oversight will make buy-side algos dumber, while possibly pushing scrutiny-sensitive liquidity providers out of the game.

In addition, if the definition of an algo, or what triggers vetting, is broad, it will be a pain, but not horrible; however, if the triggering definition is any programmatic change, then it will be a nightmare, as any little change could trigger a new review. Given the fact that algos change all the time, hyper-vetting would make getting anything into production a challenge.

Mandatory quoting requirements for market makers under ESMA’s latest MiFID II proposals shouldn’t be an issue for the majority of liquidity providers. Companies engaged in electronic market-making strategies would be obligated to quote for at least 50% of the daily trading hours. But how will quoting be defined? If it is defined by having a valid quote in the market, do stub quotes count? That worked out really well during the US Flash Crash (sarcasm intended). But seriously, how many ticks away from the touch will be satisfactory? Will it be measured by stock, or by firm? Again, depending on the regulatory definition of quoting, this could be either a non-issue or a disaster.

Beyond the MiFID II HFT rules there are a host of other overlapping regulations – including dark pool caps, systematic internalizer quoting obligations, and Basel III capital requirements. While the EU wants greater transparency and less HFT, it also is making it harder and more expensive for banks to provide capital.

Now, all of the rules aren’t final. But let’s try to see what these rules will mean as they are all added up.

Greater scrutiny on algos, increased quoting responsibility, greater transparency and access/risk controls (for SIs), and the imposition of order-to-trade ratios will mean fewer algos, less-sophisticated buy-side algos and less electronic market makers. Dark pool restrictions will push the buy side to trade in larger size, and blocks will be manually crossed (or will use a Liquidnet-type platform). These things point toward a wider tick size and greater crossing at the midpoint (which reduces price discovery incentives), but increased trading at the touch on lit venues (which provides greater incentive to quote). Volatility for liquid names most likely will not deteriorate; however, for less liquid names, it will be harder and more expensive to find the other side.

This will be bad for anyone that trades in between the spread but not at significant size – namely, retail flow not executed by a systematic internalizer, smaller money managers, and folks who trade smaller capitalized equities, even though less liquid smaller cap equities will have lower thresholds to fall under large-in-scale wavers.

In a wider tick-sized market, we can also assume that greater liquidity will pool at the tick (or outside the tick, if tick increments are not adjusted); but will we move away from automated market making? Probably not. That would be very risky, as while “good” market-making HFT would most likely be curtailed because of order-to-quote ratios, liquidity-taking algorithms would be under less scrutiny, given they don’t quote and/or cancel many orders.

This type of regime, depending on tick sizes, would benefit traders who have good intrinsic value models and – given the wider tick and the increasing need for the buy side to provide liquidity (as Basel III will push banks away from providing liquidity in scale via capital provision and market making) – attack liquidity when it’s mispriced, which will occur more often. This would be especially true when stocks trade outside of their mandated tick regime.

So buy-side traders and MiFID II/anti-HFT proponents, get ready for the rollercoaster ride of your life. These changes will push Europe into trading in greater size, with less effective tools, searching for liquidity that is increasingly less accessible, while Basel III increases the cost of providing capital in size. This forces buy-side traders to more aggressively quote, and in addition to leaking information, these less sophisticated algos will be picked apart by very smart and fast liquidity-taking strategies that will most likely fall under less scrutiny than current HFT posting strategies that exhibit greater cancellation rates.

Rather than vanilla algos, this market will require the buy-side to have more-sophisticated algos employing effective order cancelation and repricing tactics to manage displayed liquidity. But unfortunately, depending on cancelation frequencies, what the buy side actually needs may trigger order-to-trade caps.

If these rules are implemented the buy side, which is angling for a safe and more effective market, will wind up with neither. Larger orders will be picked apart, displaying liquidity will be harder, and the capital that once was offered to the buy side or displayed in the market will disappear.

So what if you created a market to trade in size, and size just doesn’t show up?