A MIFID brainteaser: define liquidity

What do we mean when we say a financial instrument is liquid? That it trades 100 times a day? Ten times a day? Less than that? It’s a question the European Securities and Markets Authority (ESMA) currently has the unenviable task of trying to answer for all non-equity instruments as part of its efforts to draw up detailed rules for the implementation of the revised Markets in Financial Instruments Directive (MIFID) in Europe. Unenviable because a lot is riding on getting the definition spot on – continued market liquidity at current levels for one.

The definition is crucial to much of what is in MIFID and the associated Markets in Financial Instruments Regulation (MIFIR). Liquid instruments will be subject to strict pre- and post-trade transparency requirements, as well as potentially having to comply with an obligation to trade on a regulated market, multilateral trading facility, organised trading facility or equivalent third-country venue.

Under ESMA’s May 22 proposals for post-trade transparency, for instance, information on price, volume and time of trade would have to be made public within five minutes of a transaction in a liquid instrument taking place (although deferrals exist for trades that meet yet-to-be-decided size-specific and large-in-scale thresholds). If an instrument is deemed to be illiquid, however, publication of the most sensitive information is deferred until the end of the following day.

ESMA does have some guidance on how to approach the liquidity definition within MIFIR. A liquid instrument is one where there are “ready and willing buyers and sellers on a continuous basis”. MIFIR also sets a variety of criteria that should be used to determine this: average frequency and size of transactions over a range of market conditions; the number and type of participants; and the average size of spreads, when available.

On a hunch, it seems likely the most standardised parts of the over-the-counter derivatives market may meet these liquid instrument parameters. A study conducted by analysts at the Federal Reserve Bank of New York, based on three months of interest rate derivatives transaction data from 14 large global dealers in 2010, found the most popular interest rate swaps traded up to 150 times each day – a frequency that would presumably meet the continuous buying and selling requirement under MIFIR. However, plenty of instruments barely traded: the New York Fed reported more than 10,500 combinations of product, currency, tenor and forward tenor over the three-month sample, but found approximately 4,300 combinations traded only once during that period.

ESMA’s proposals are meant to weed out these kinds of infrequently traded instruments from the post-trade transparency requirements, with regulators recognising that an overly broad regime could deter dealers from facilitating client trades in less liquid products, causing a further decline in liquidity. ESMA suggests looking at a minimum number of transactions and the number of trading days on which at least one transaction occurred within a certain time frame, alongside the other criteria on average transaction size, number of market participants and spread size.

The critical element in all of this is where the thresholds will be set. ESMA hasn’t yet started its analysis of derivatives data, but it took an initial stab at setting some possible numbers for bonds within its proposal, suggesting six possible combinations. For the minimum number of days on which at least one trade takes place over a year, it suggests a threshold of 120 or 240 days – in other words, the instrument must trade at least once every other day or once a day. That’s combined with a minimum-number-of-trades-per-year threshold of 240 or 480 in all but one of the six scenarios, equating to an average of just one or two trades a day.

It’s clear the choice and combination of thresholds can dramatically alter the proportion of trades classed as liquid. In two of the ESMA scenarios, where the minimum number of trades per year and average daily volume are fixed at 480 and €100,000, respectively, increasing the minimum number of trading days from 120 to 240 reduces the universe of bonds classed as liquid from 4.71% to 1.61% and lowers the percentage of volume captured from 86.67% to 62.90%.

This same kind of analysis will also be performed for OTC derivatives once ESMA has collected the relevant data – and ISDA is working to help pull that information together. In setting the thresholds, it’s likely ESMA has a figure in mind with regards to the percentage of traded volume in each instrument it wants to apply the liquidity requirements – in fact, it says as much in its proposal, noting that it will combine expert judgement with a coverage-ratio-type approach when setting thresholds. Given US regulators set a coverage ratio of 67% for the purposes of the Dodd-Frank rules, it’s not beyond the realms of possibility that ESMA will target a similar level.

The question then is whether the thresholds required to reach this percentage coverage level truly reflect continuous buying and selling activity, as per MIFIR’s requirement. The challenge for ESMA will be in balancing a desire to apply the new transparency rules to a large enough portion of trading volume in each market with thresholds that reflect real liquidity, based on an objective analysis of the data.

All told, a lot of work for both market participants and regulators – and a matter of months to do it in.

Liquidity Is King

While it is summertime, the livin’ is not (always) easy. The EU regulatory process continues unabated and with the publication of a first complete Compromise Text on MiFID 2/MiFIR earlier this summer, the journey to trading obligations and platforms in the EU has well and truly begun. The process will take another step forward with the ECON Committee vote in the European Parliament, scheduled (after a postponement) for late September.

One of the big questions that remains unanswered in the MiFID/MiFIR debate – and that explicitly needs to be addressed before its resolution – relates to liquidity.

There are two aspects to this issue. One has to do with the use of liquidity as a benchmark or trigger for determining whether a rule applies to a particular obligation. The second has to do with the impact of the rules on the liquidity of traded financial instruments.

First things first: On the subject of a liquidity trigger, at what point should a financial instrument or market be considered liquid enough to support particular regulatory obligations? Such obligations might include those related to mandatory central clearing of OTC derivatives; mandatory trading of derivatives on regulated venues; or pre- and post-trade public transparency. That question is at least in focus in the MiFIR debate. Much less attention is being paid to an equally important question: when does an instrument or market stop being liquid enough to support those obligations; and how do you suspend them?

This is not just an academic discussion. Various provisions in the MiFID/MiFIR proposals reference liquidity as a trigger for particular obligations (notably the trading obligation and pre-trade transparency requirements).

Moving now to the second issue, the impact of the proposed rules on liquidity: To what extent could additional regulatory obligations increase or impair the existing liquidity of a particular financial instrument or a market as a whole? To be fair, some policymakers show awareness and sensitivity to this issue.

In the continuing dialogue over the proposed rules (particularly over the trigger issue), it is important that policymakers consider the potential for liquidity to vary over time. To put it bluntly, there is no market where permanent liquidity can be guaranteed (or, by the same token, mandated). This might reflect changes in market conditions in general, or simply the nature of the individual product.

Changes in liquidity levels could be inherent to the product in question, with liquidity falling in the period after the conclusion of the contract. For example, an ‘off-the-run’ credit index contract is likely to be significantly less liquid than the ‘on-the-run’ equivalent. And, even for an on-the-run index, the five-year maturity will be liquid where, say, the four-year is not. Changes in liquidity levels could also reflect external factors, such as weakening in the supply of credit. For this reason, it is important that the market infrastructure is sufficiently flexible to accommodate any change in liquidity.

In addition, given the potential for liquidity to change, liquidity triggers should also typically be two-way, i.e. the trigger should not only define when a particular regulatory provision applies, but should also define when the provision ceases to apply. Ideally, it should be possible to suspend particular obligations immediately, to ensure that markets continue to function (albeit with less turnover) during times when liquidity is stressed.

It has long been our view that legislative reform should support liquidity in the interest of systemic resilience, should protect the funding requirements of corporates and sovereigns and should advance the principle of strong risk management. And to that end, we support linking obligations to liquidity in appropriate circumstances. However, liquidity triggers:

• Must consider the various factors that collectively constitute liquidity;
• Must be forward-looking;
• Must be two-way so that obligations can be disapplied when necessary.

There are many welcome and positive developments in the Compromise Text. However, where the European Securities and Markets Authority (ESMA) is given the task of determining or applying a liquidity trigger, then its mandate should be such as to allow it to consider all relevant factors.