I wrote a piece recently on the practical problems that ESMA are coming up against implementing MIFID 2, like the absence of a pan-European trade tape, and how this was not necessarily a bad thing given what they were trying to do. On the basis of feedback (always gratefully received) I thought Id elaborate on why I really dont want them to succeed.
What ESMA are struggling to do is define what is and what is not a liquid covered bond. They argue that the broad brush criteria that they use size, 750mn in our market is decently correlated to empirical measures of liquidity number of trades. But they are only correct in one direction. Liquid bonds are large but not all large bonds are liquid. Their criteria throw up a huge number of so called false positives.
Given what they are going to do with this definition I am far more concerned about false positives illiquid bonds that they deem to be liquid - than false negatives the opposite.
I would guess, on the basis of nothing more than instinct, that the vast majority of covered bond trades in practice are in bonds that meet ESMAs size criteria (750mn or above) but that dont meet the empirical measures of liquidity that ESMA have set themselves. In other words, false positives.
The reason that false positives are a problem is clear when you look at why they are measuring them. If a bond is not liquid the market maker can gain some protection from their obligation to provide price transparency (pre- and post- trade). That allows them to make sensible bids. If they are asked to show a price in an illiquid bond to the whole market, and if they then trade on the basis of that price and the whole market knows about it, the trade is clearly far less attractive for them. Which means they wont do it. So investment banks will allocate less balance sheet and eventually traders to providing liquidity in these bonds.
The fact that the size criteria by which they measure liquidity differs in different markets is also a problem for me. Sovereign bonds must exceed 2bn to be liquid whereas bank sub notes only need to be 500mn. Why? This might be a similar size relative to the average issue in each market but that only makes sense if you assume that each individual submarket only allows a certain proportion of all bonds to be liquid, on-the-run if you prefer. A 500mn sovereign bond is probably more liquid than a 500mn sub bond even if it is far smaller than average in its own market.
All of which sums to one big regulatory contradiction. Most of the covered bonds actually traded are eligible for bank liquidity buffers in tier 1 because of their liquidity. But most of them are false positives, therefore in most of them market makers are vastly discouraged from showing a meaningful price.
Which doesnt feel like a very sensible policy outcome to me.
Event, Articles and Videos that might interest you
No related events !