Last week, the FCA published an interesting academic paper on corporate bond liquidity in the years following the financial crisis, stating: 
“…although the inventory of dealers has declined in this period, there is no evidence that liquidity outcomes have deteriorated in the market. If anything, the market has become more liquid in recent years.”
Unfortunately, as fixed income portfolio managers, our day-to-day trading experience has been the polar opposite of this academic theory. For me, the most telling section of the paper was this quote:
“…the regulatory interventions that have been introduced since the financial crisis did not result in less liquidity in normal times and did not result in liquidity being more ‘flighty’ when shocks of a mild nature hit the system.”
It is hard not to be cynical about this paper and I think the last quote hints at the real motivational impetus behind this work.
Some of the regulatory changes since the crisis have certainly been painful for bond investors, though we would argue that they have been necessary improvements. Specifically, the resolution regime and especially the requirements for banks to hold higher levels of capital are good things. Crucially the specific ability to write down creditors moves the burden for bailouts away from taxpayers to those private sector creditors.
All good stuff. But to claim they have had no impact on bond market liquidity is patently at odds with our experience.
Trading volumes in credit markets
Source: RBS Credit Strategy, SIFMA, MarketAxess
Reading through the paper in detail, I was struck by the fact that these conclusions were driven by transactional data of which I have big question marks over the quality and relevance of.
For example, the authors admit they had to remove transactions with weekend dates, amongst other irregularities. More worryingly, the dataset used over the key financial crises periods from 2007 through to August 2011 was for a universe of only 409 bonds, 92 per cent of which were financials.
92 per cent financials? Most corporate bond indices have around 40 per cent in financials.
Additionally, given many of these financials were rated ‘AAA’, where the bid/offer spreads were unsurprisingly low, I can only conclude a significant portion of these transactions were in covered bonds. Truthfully, how relevant is such a narrow universe?
But this is not the biggest problem with this analysis. Far bigger is that transactional data misses the key point that we observe: what about the transactions that never happened because of illiquidity?
Trades that never happened because either the quantity was too large, or the quoted prices for the actual trade were so far away from the screen quotes that no fund manager could evidence best execution being achieved?
The authors do attempt to sense check their data using quoted bid/offer spreads from corporate bond index data later in the analysis, and to be fair they do conclude that using this data the costs of trading have increased since 2008. 
But even this misses our earlier point – the reality is that screen prices are simply not that reflective of where you can actually trade in an over the counter market these days.
Some days they are – but many days they are not. Just last week, I had a broker offer me 509,000 bonds of an insurance company we like, at a price 1 whole point higher than the best screen offers. I contacted three other brokers who had lower offers on their screens, to ask if they could sell me just 500,000 bonds. All replied saying “flat – can’t offer” meaning they do not own that bond so cannot sell me any, as to do so would take them short.
So as a portfolio manager, do you buy from the only bank who has inventory, but then report an immediate 1.5pts loss compared to the mid-priced mark that will used on the portfolio valuation? Or do you pass on the trade?
This is the reality of life at the coalface, and this is not even considered in the FCA’s Occasional Paper 14. This is where real world experience counts over academic certainties, and is the driving reason the portfolio managers do their own trading at TwentyFour.
We do not sub-contract the execution of trades to a central dealing desk because the point of execution on any trade is such a key factor in fund performance.
Fixing the liquidity issues that we see every day has no simple solution, unfortunately. As Mark Holman (chief investment officer at TwentyFouyr) recently said, we think Bloomberg can be a key part of this solution by allowing buy-side and sell-side firms alike to post inventory through their system – but that is for the future.
For now, blog readers may remember the credibility disaster of the massive 80’s US soap “Dallas”, where the character Bobby Ewing was killed off in the last episode of the 1984-1985 season.
The actor Patrick Duffy left the series for a year, only to return from the grave at the end of the next series. The ridiculous plot twist this required was to have viewers believe the following whole year season had only been a dream in the mind of Bobby’s wife, and the very much alive Bobby stepped out of the shower into the wide open eyes of viewers, completely destroying the believability of the whole show.
It never recovered its credibility from that point, nor its viewer numbers.
Similarly, re-writing history by selecting an unrepresentative tiny sample of actual trades is akin to telling Dallas viewers it was just a dream. I can understand the paper’s conclusions from the specific, narrow data that they used – but that data set does not tell the real story. And to claim that in fact liquidity has gone up is incredibly unhelpful and could arguably be dangerous.
Dangerous to the extent that if it implies a level of liquidity for investors that leads to asset allocation shifts, and that liquidity inference does not match the reality, then investors could end up paying a higher price than they expected.
Lastly, we extend an invitation to the paper’s authors to come and spend a day with us trading credit, where they can witness first-hand the difference between theory and practice.
Chris Bowie is manager of the TwentyFour Corporate Bond fund. All the views expressed above are his own and shouldn’t be taken as investment advice.