5.II.2.1.3 Background

CESR noted that post-MiFID I reference price systems gained in popularity, in particular as deployed by MTFs.1 CESR noted that the drivers behind using the reference price systems changed over time. Non-disclosure based on reference prices was no longer primarily due to the concern that the publication of orders, especially in illiquid shares for which the systems were most often used, would increase the incentive to manipulate the continuous (reference market) before the reference price was fixed.2 Under MiFID I the reference price waiver was also used to get potential price improvements, including for smaller orders.3 Most reference price systems executed orders at the midpoint, that is – the average between the best bid and offer, of the reference market. Execution at the midpoint permitted investors trading under the reference price waiver to obtain potential price improvements.4 Reference price systems could also use the best bid or best offer as a reference price,5 thereby permitting to distribute the price improvements more to the seller (best bid) or buyer (best offer).

During the MiFID I-review concerns were raised that the reference price waiver was also used to get price improvements for smaller orders. Several respondents to the CESR consultation argued this was inconsistent with the general intention of the MiFID I waivers to protect against market impact. In their view, a minimum threshold for large orders eligible for the reference price waiver needed to be set. This was to protect price formation and to stay close to the ‘original intention of the waiver’.6 It was argued that the reference price needed to be limited to the midpoint, as only that point of the spread would be justifiable for those orders not contributing to the price formation process (i.e. both the buyer and seller would get a price improvement).7

Contrarily, others market participants claimed that the reference price waiver was designed to promote competition on trading costs between dark and pre-trade transparent trading platforms. These market participants argued for more flexibility of the waiver, allowing more scope for execution within the visible market spread (instead of only at the midpoint, best bid or best offer).8 It was argued that price improvements also needed to be available for smaller orders, as well as the protection from information leakage and exposure risks. It was noted that this was of particular importance for large orders that were ‘sliced’ into smaller orders with the aim to reduce market impact.9

The controversy among market participants was also evident within EU. Political controversy caused the debate on the detailed aspects of the reference price waiver to shift from level 2 to level 1. The Council proposed a strikingly specific Level 1 text. This reflected the political tensions in drafting the reference price waiver.10 The Council proposed that the reference price could only be derived from: (i) a trading venue where ‘the average daily trading volume is (would have) a sufficient proportion of the average overall trading volume of that financial instrument’ (reference market); (ii) where that reference price would be widely published and regarded by market participants as a reliable reference price. The Council also wanted to end flexibility in setting the reference price. The Council suggested that the reference price needed to be established by obtaining (iii) the midpoint within the current bid and offer prices; or (iv) the open and/or closing price of the relevant trading session (only for prices outside the continuous trading phase of the relevant trading session).11 Finally, the Council wanted to reduce the application of the reference price waiver with the aim of preserving price formation. In effect, the Council proposed to apply the double volume cap mechanism to the reference price waiver.12

The intervention of the Council would restrict the use of the reference price waiver compared to MiFID I, that is – stricter rules would apply to (a) the reference market (‘average daily trading volume’); (b) the reference price could only be set at the midpoint (save for non-continuous trading); and (c) the double volume cap mechanism would apply. The subsequent legislative process in the trilogue only altered the rules of the reference market (point a). A broader set of reference markets was proposed, being the primary market (i.e. where the financial instrument was first admitted to trading) or the most relevant market in terms of liquidity.13

The position of Council, as slightly altered during the trilogue, is evident in the final MiFID II text. MiFID II requires the reference price to be based on: (i) the midpoint; or (ii) the opening and/or closing price of the relevant trading session (only for prices outside the continuous trading phase of the relevant trading session).14 The reference market can be the primary market or the most relevant market in terms of liquidity. The double volume cap mechanism applies to the reference price waiver. Last, but not least, the reference price needs to be widely published and regarded as a reliable reference price.15 The main requirements of the reference price waiver are laid down in the level 1 text (MiFIR).16