Defining Liquidity

Europe’s revised Markets in Financial Instruments Directive will impose strict pre- and post-trade transparency requirements on all classes of derivatives deemed to be liquid. Establishing the definition for liquid instruments is therefore critical, but ISDA believes the thresholds currently proposed by ESMA are too low


❏ Under ESMA proposals, an interest rate derivatives instrument could be classed as liquid if it trades just once a day, with an average daily notional of as little as €10 million.

❏ ISDA believes the thresholds should be set at 15 trades a day with an average daily notional of €500 million.

❏ This is more in line with the MIFID level-one text requirement that a liquid instrument is one with “continuous” buying and selling activity.

❏ Liquid instruments are subject to pre- and post-trade transparency requirements, as well as potentially being required to trade on regulated venues.

❏ Applying pre- and post-trade liquidity too broadly could discourage market-makers from participating in less liquid markets.

❏ The end result will be less liquidity and higher costs for end users.

There would appear to be very few parallels between defining obscenity and classifying liquidity. But having been asked to rule on the threshold for obscenity in 1964, US Supreme Court Justice Potter Stewart used a phrase that could equally be applied when defining liquid markets: I know it when I see it. The European Securities and Markets Authority (ESMA) has been asked to go much further than that, however.

As part of the process to put meat on the bones of the revised Markets in Financial Instruments Directive (MIFID II) and Markets in Financial Instruments Regulation (MIFIR), ESMA is required to set quantitative thresholds for determining whether a financial instrument is liquid. It’s an important job: liquid instruments will be compelled to meet pre- and post-trade liquidity requirements, as well as potentially being subject to an obligation to trade on a regulated market, organised trading facility, multilateral trading facility or recognised third-party venue.

Transparency regime

The pre-trade transparency regime requires the publication of bid and offer prices before a trade takes place, while further information—including price and volume—is required to be reported soon after the transaction is executed. Waivers do exist in certain circumstances. The pre-trade reporting requirements won’t apply if the trade is large compared to ‘normal’ market size (a block trade in US parlance). The reporting of post-trade information would also be deferred by 48 hours. An additional ‘size-specific-to-the-instrument’ waiver also exists for request-for-quote and voice-trading systems, meant to ensure liquidity providers are not exposed to undue risk.

Despite these waivers, there are risks in setting the initial ‘is it liquid?’ hurdle too low. Dealers take on risk to facilitate client orders and then look to hedge their exposures. In less liquid markets, that can take time—and if details of the trade are published before the executing dealer can hedge, then other participants may attempt to take advantage of that information. This could deter dealers from facilitating client trades in less liquid products, causing a further decline in liquidity and ultimately leading to higher costs being passed on to end users.

How regulators determine what is and what isn’t liquid is therefore crucial. And ESMA’s December 19, 2014 consultation paper contains the blueprint of how it proposes to do this.

ESMA proposal

To some extent, its hands are tied by the level-one MIFID II/MIFIR text agreed by the European Parliament, Council of the European Union and European Commission. An overriding criterion is that a liquid market is one where there are “ready and willing buyers and sellers on a continuous basis”. The level-one text also sets out a list of variables that should be considered: the average frequency and size of transactions over a range of market conditions, the number and type of market participants, and the average size of spreads where available.

ESMA suggests focusing primarily on the first two—frequency and size—when assessing liquidity. The other variables will only be taken into account in specific cases or for certain asset classes. When considering frequency, ESMA proposes to set both a minimum number of transactions over a given period and a minimum number of days on which trading occurs over that time.

Taking the interest rate derivatives market, ESMA proposes to first identify which broad classes of derivatives are liquid at a high level—for instance, whether interest rate swaps as an entire product class are liquid. It then plans to drill down to the sub-class level, looking at tenor, underlying and/or currency. These sub-classes would then be assessed for liquidity based on asset-class-specific thresholds.

Table 1 summarises ESMA’s proposed thresholds for certain interest rate derivatives, along with the results of its analysis. For instance, ESMA suggests that a floating-to-floating interest rate swap only needs to trade once a day to be considered liquid. Combined with a notional threshold of €50 million, it has determined there are 48 liquid sub-classes, covering 72% of trades and 80% of the notional traded.

Table 1. ESMA analysis results—interest rate derivatives

Table 1. ESMA analysis results—interest rate derivatives
Source: ESMA

ISDA response

This analysis has triggered counter proposals from the industry. In its response to ESMA’s paper, published on March 2, 2015, ISDA argues that one trade a day does not satisfy the “continuous” buying and selling requirement set by the MIFID legislation, and proposes using a higher threshold of 15 trades a day with an average daily notional of €500 million. These levels better reflect the concept of continuous buying and selling activity, ISDA argues.

ISDA argues that one trade a day does not satisfy the “continuous” buying and selling requirement set by the MIFID legislation

The response also draws attention to concerns about how ESMA calculates tenor in the consultation paper. Swaps that don’t have whole year tenors appear to have been mis-classified in some circumstances because ESMA didn’t take leap years into account and failed to recognise that some swaps have effective dates of T+2. This means that, in some cases, liquid 10-year swaps are classified by ESMA as 11-year instruments, skewing the results for the less liquid 11-year tenor.

ISDA corrected this in the analysis conducted as part of its March 2 response. Changing the methodology for tenor alone doesn’t alter the coverage ratio of the trades captured, but it does reduce the number of liquid sub-classes. For instance, it means some six-year, 11-year or 31-year sub-classes are no longer classed as liquid under the new methodology.

Table 2. Combining ISDA thresholds with corrected tenors: Fixed-to-floating interest rate swaps

Table 2. Combining ISDA thresholds with corrected tenors: Fixed-to-floating interest rate swaps
Source: ISDA, DTCC

Table 3. Combining ISDA thresholds, corrected tenors and more granularity of sub-class: Swaptions

Table 3. Combining ISDA thresholds, corrected tenors and more granularity of sub-class: Swaptions
Source: ISDA, DTCC

In fact, using the corrected approach for tenor combined with the higher thresholds recommended by ISDA doesn’t dramatically affect the percentage coverage in terms of notional captured for some instruments. Using the ISDA year fraction for fixed-to-floating interest rate swaps and applying the higher threshold of 15 trades a day and €500 million in notional reduces the number of liquid sub-classes from 247 under the ESMA approach to just 27 (see Table 2). However, the coverage ratio does not fall by an equivalent amount—notional captured, for instance, falls from
97% to 72%.

In some cases, there is a more significant impact. For multi-currency floating-to-floating swaps, for example, the number of liquid sub-classes falls from 39 to 0, meaning the coverage ratio falls from approximately 65% of notional to zero. ISDA argues this is appropriate: multi-currency floating-to-floating swaps are less liquid than single-currency swaps.

There are some caveats to these results, however. Importantly, ESMA based its analysis on three months of European trade repository data, while ISDA used publicly available information from the Depository Trust & Clearing Corporation’s US swap data repository. Using different underlying data likely leads to anomalies in the number of liquid sub-classes and coverage ratios between the ESMA and ISDA analysis. But rather than focus on the absolute numbers in its response, ISDA recommends ESMA re-run its analysis of European data using the revised approach for tenor and higher thresholds.


The granularity of ESMA’s sub-class definitions is another issue highlighted by ISDA. This is a particular problem for swaptions, where ESMA has used only currency to determine the sub-classes. As a result, it has declared that all swaptions in US dollar, euro, sterling, yen and Australian dollar are liquid. This leads to an unrealistic situation where an option on a 50-year euro swap would be classed as liquid, but the underlying 50-year euro-denominated interest rate swap would be classed as illiquid. ISDA strongly recommends that ESMA further breaks down the sub-classes of swaption, using currency, underlying index, tenor of underlying and tenor of option.

When using a more granular sub-class determination and higher thresholds of 15 trades a day and €500 million in notional, the number of liquid classes falls from five to one, and the coverage ratio drops from approximately 97% of notional to 8% (see Table 3).

Aside from interest rate derivatives, ISDA also raises concerns with the approach taken for equity and commodity derivatives. In equity derivatives, no attempt has been made to distinguish between exchange-traded and over-the-counter (OTC) equity derivatives. The ESMA analysis is based purely on exchange data, but an exchange-traded option is very different to an OTC equity option—they are not fungible nor economically equivalent in many cases. Unless an attempt is made to distinguish between the two, OTC products could be caught by inappropriate transparency requirements, ISDA argues.

ESMA’s own data also shows that most exchange-traded equity derivatives are illiquid. Yet it proposes to mandate pre- and post-trade transparency requirements to all equity derivatives traded on a trading venue.

imilar points can be made about commodity derivatives. ISDA argues that ESMA’s proposed sub-classes should be subject to a more granular determination, alongside higher thresholds, preferably based on open interest rather than notional. Thresholds should also be set in US dollars rather than euro. The vast majority of commodities are dollar-denominated, and using euro would mean contracts could become liquid or illiquid based on the movement of exchange rates.


MIFID defines a liquid instrument as one with continuous buying and selling activity. What does ‘continuous’ mean? It’s clearly up for interpretation, but ISDA argues that it’s more than one trade a day on average. In fact, ISDA believes that a more appropriate threshold for interest rate derivatives is 15 trades a day, alongside a notional threshold of €500 million. A higher threshold would reduce the coverage ratio slightly for some instrument classes, and more significantly for others. But less-liquid sub-classes would be removed from the liquidity determination.



❏ Read ISDA’s response to the MIFID II/MIFIR consultation paper at:

❏ Read ESMA’s December 19, 2014 consultation paper at: