In this piece, we offer additional thoughts on the subject of bond market liquidity (see also Fixed Income Perspective “Fractured Liquidity”, 29 May 2015) which has come under sharp focus over recent months. In recent days, a prominent bond asset manager has issued a view in which risks associated with potential liquidity accidents were significantly downplayed.
However, we think it is extremely dangerous to present the significance of current concerns on bond market liquidity as non-material, especially given the trajectory of regulation and the unprecedented importance of monetary policy (deployed via direct participation in bond markets) in affecting both market and economic outcomes in the current cycle.
Assessing damage to bond market liquidity
In our view, it is important to assess some key metrics of bond market liquidity in order to give a more concrete sense of the issues investors are currently facing. The sharp fall in inventories held by the banking sector in corporate bonds in recent years (in response to tighter regulations) alongside a sharp rise in the assets under management is well understood by the market. In addition to this structural development, it is important to look at specific examples of damage visible in the indicators for both market depth and breadth during the latest episode of sharp moves in bond markets witnessed in April/May.
Recent reports by JP Morgan record a sharp decline in both market breadth and depth witnessed in 30yr bunds during this period (i.e., the market thought to be at the centre of the latest liquidity-induced accident). For example, using JP Morgan’s market depth proxy, it appears one could have traded around 100, 30yr bund contracts at the beginning of 2014 without any real market movement , but this number fell to below 20 during the most intense days of this recent episode. Of course, one of the key reasons behind this deterioration is the sustained decline in German repo market activity, which has contracted by 30% over the past five years (a trend which is set to continue as ECB’s QE policy runs its course).
Turning to corporate bond markets (again using numbers compiled by JP Morgan using FINRA Trace data) there is a clear decline in both US IG and HY market turnover between 2007 and 2011. However, since 2012, there has been a divergence, with some improvement showing in HY turnover, while rebound in IG has lagged considerably. All in all, current turnover is 30% below its pre-crisis peak for HY versus 55% for IG. Similarly, looking at average trade size (trading volume in USD divided by the number of trades), these figures are around 30% and 40% below their pre-crisis peak for IG and HY respectively.
Does fractured liquidity pose a systemic risk?
In a world where monetary policy is doing the heavy lifting of supporting output growth and inflation expectations via extensive use of QE programmes, the importance of both the level and the volatility of interest rates in determining economic outcomes is critical. For instance, take the example of the Eurozone, where, given German-led political opposition towards using fiscal policy, monetary policy is the only policy tool available to deal with what is essentially a combination of cyclical and structural issues. Within this framework, it doesn’t take a big leap of faith to see how micro-liquidity induced moves in interest rates can inflict damage to what is already a very weak starting point in terms of both economic strength and inflation.
This transmission channel is not just relevant for the Eurozone but also applies to the US. For instance, recall how the Fed changed its policy stance back in 2013 when the taper-tantrum ended up forcing a shift in policy. Tapering was delayed and communication was made outwardly dovish as economic damage inflicted as a result of sharp tightening in financial conditions started to become visible in data.
Turning to systemic issues stemming from the inter-linkages involving the financial sector (the key source of systemic risk using 2008/9 and 2011/12 as templates) despite a move towards a banking union in the Eurozone, the nexus between the sovereign and banking sectors risk remains as tight as ever (a point raised several times by Bundesbank Chief Jens Weidman in his recent public speeches). Here, it is easy to see how an episode of sustained sharp volatility in rates market (either triggered or exacerbated by the dismal liquidity situation) can put the spotlight on the issue of sovereign debt holdings on bank balance-sheets (especially, in Southern European countries). Indeed, if anything the recent Greek saga (where sovereign risk quickly morphed into banking sector risk) shows that cracks in the union are still there (and exposed to political risks), even if the backstops available now are much stronger-than the dark days of 2011/12.
Role of LCR Regulation – A case study in safety leading to “over safety”
Studying the case of the Liquidity Coverage Ratio (LCR) regulation (a key plank of Basel III regulations) can help us better understand the far reaching implications of the current regulatory landscape on the functioning of government bond markets.
As highlighted by Gary Gorton and Tyler Muir of Yale university (14th Annual BIS 2015 conference proceedings), the LCR regulation essentially requires all repo activities be backed dollar by dollar with government bonds – a kind of narrow banking, where one kind of money backs another kind of money. In fact, according to the authors, the LCR regulation is an attempt to make collateral “immobile” once again. This shift is a reversal from the pre-crisis years, which was effectively an age of mobile collateral (for example, packaging of bank loans in the form of MBS/ABS, which could be then be traded, posted as collateral etc). The authors conclude that such a system-based immobile collateral has been tried before (during the US National Banking Era) and it failed miserably.
Specifically , Gorton and Muir link repeated instances of Primary Dealer repo fails in treasury markets in recent months, which they argue is driven by increase in convenience yield of US treasuries as a result of regulatory changes. The authors also raise the point that during the National Banking System era, the size of shadow banking increased (surprise, surprise) and those years saw five major banking sector panics.
Interestingly, at the same BIS conference, Randall Kroszner of University of Chicago commenting on “Mobile versus Immobile Collateral” invoked the Titanic tragedy (1912) after which the regulatory response was “lifeboats for all”, which had the unintended consequence of playing an important role in the Eastland disaster in Chicago in 1915, which led to the deaths of more than 800 people.
Implications for global fixed income investing in a highly indebted world ravaged by fractured liquidity
First and foremost, given our thinking above, we believe that the current fractured liquidity situation is here to stay. The “life boat syndrome” currently afflicting policy makers is playing a strong role in damaging market liquidity and therefore is an important third dimension to understand when it comes to fixed income investing (in addition to risk and return).
Of course, risk premium focused investors are no longer the only player in global fixed income markets. The usage of quantitative easing policies by central banks in the aftermath of the 2008/09 crisis means they have become a dominant player in fixed income markets with very different incentives to traditional investors.
Given this broader environment, we think investors no longer have any influence on the direction of regulation (which is likely to tighten further) or the consistency of central bank policy making going forward. However, investors do have an important degree of freedom in the form of portfolio construction. Here, we think deep assessment of the current market liquidity landscape in a highly leveraged world where debt burdens have continued to rise despite the 2008/9 balance-sheet crisis is critical. For instance, according to BIS, G20 countries now have USD 40 trillion more debt than at the beginning of the crisis.
Indeed, liquidity assessment is not only a second level risk-management exercise for us, but it enters directly as an important factor influencing our fixed income portfolio construction process (both for corporates and sovereigns ). In fact, we think that the current environment involving direct central bank interventions in fixed income markets and tightening regulatory landscape (with the unintended consequence of fracturing liquidity) calls for a longer-term mind set; the best anchor of which is a focus on underlying fundamentals when building fixed income portfolios which are designed precisely to mitigate credit risk.
All in all, in our view, damaged liquidity-induced accidents are likely to become a regular occurrence going forward and we believe that a fundamentally sound portfolio (that is built transparently) is likely to weather the upcoming storms better than market-cap based allocations which continue to be used by vast majority of investors to access fixed income risk-premia.
Salman Ahmed
Lombard Odier Investment Managers
