In “How The ECB Is Distorting Euro Money Markets [19]” we summarized Barclays take on the effects of ECB QE as follows: “short-end core paper will trade below -0.20%, extreme supply/demand imbalances will cause general collateral rates to trade through the depo rate, money market fund yields will turn decisively negative testing investor patience, and central banks had better make good on promises to make some of their inventory available for lending or risk impairing the functioning of the repo market (never a good idea).” A little over two weeks into the PSPP and sure enough, signs are already beginning to show that the ECB is effectively breaking the market. Last week, we got this via Reuters [20]:
The soaring cost of borrowing government bonds in secured lending markets highlights the distortions caused by the ECB's asset-purchase scheme, which analysts say could clog up Europe's financial system.
Uncertainty over how the European Central Bank will counter the scarcity of top-rated debt could further shrink repo markets -- a source of funding that is essential to the smooth running of bond markets...
One broker said every German government bond eligible for ECB purchase was now trading 'special', meaning exceptional demand had made it more expensive to borrow for three months than general collateral.
Then today, this from Mizuho’s Peter Chatwell via Bloomberg:
Some bonds in German market are trading special in repo, Peter Chatwell, strategist at Mizuho, writes in client note.
Picture for relative-value trades has deteriorated, with the number of bonds that trade rich vs fitted curve becoming even richer.
As list of DBR specials grows, relative value in Germany may become dysfunctional until Eurosystem lends out bond holdings under QE.
Recall that we've seen a similar dynamic [21] in the US of late with the two-year trading negative in repo. To demonstrate the dramatic effect PSPP purchases are having on the market (and by extension, how important it is for the ECB to get the securities lending operation right), consider the following from JPM (this is from one week into the program):
The first issue of collateral shortage can be seen in the collapse of GC repo rates to negative territory since the beginning of last week for terms of greater than 3 months. 1yr Germany has been trading at close to -30bp; i.e. one can currently fund purchases of Bunds via the term repo market and achieve positive carry by even buying Bunds with yields between -20bp to -30bp. We note that this expensiveness in term repos shows how unwilling Bund holders are to depart from their collateral for more than a few days or weeks.
And in terms of liquidity — and remember here that liquidity means the degree to which you can trade without impacting prices too much, or as Howard Marks recently put it, “the key criterion isn’t “can you sell it?”, it’s “can you sell it at a price equal or close to the last price?” — the ECB pretty clearly had a rather outsized negative effect very early on. Here’s JPM again:
...by the sharp decrease in Bund liquidity as our market depth metric; i.e. the ability to transact in size without impacting market prices too much, collapsed this week . We measure market depth by averaging the size of the three best bids and offers each day for key markets. Figure 2 shows two such measures, for 10-year cash Treasuries (market depth measured in $mn) and German Bund futures (market depth measured in number of contracts). While both UST and Bund market depth have been trending lower in recent months and while the former moved recently below the Oct 15th low, what was striking this week was the divergence between a modest increase in UST market depth vs. an abrupt decline in Bund market depth. We note this development effectively challenges the market neutrality condition of the ECB from the first week of purchases already!
[22]
JPM goes on to explain — as they have before — that QE really just replaces one form of collateral with another and “shortage” isn’t really the appropriate term but rather “scarcity,” as “scarcity” implies that one form of collateral (in this case EGBs) has simply been made more expensive vis-à-vis another form of collateral (in this case cash). However, because cash isn’t as efficient as a form of collateral as the bonds it's replacing, the ECB has in fact engineered a shortage:
However, this assessment is complicated by reduced usage and efficiency of cash collateral in recent years. In particular, usage of government bond collateral has increased at the expense of cash collateral by both banks and investors…
Banks are responding positively to reduced appetite for cash collateral by the buy side as this also helps banks to increase the efficiency of their own collateral management processes via re -hypothecation of security collateral and via consolidating and optimizing collateral across OTC derivatives, securities lending and repos to meet more onerous regulatory requirements. Re -hypothecation or re -use rate of security collateral has decreased post the Lehman crisis , but at around x2 currently it makes bond collateral more efficient than cash collateral…
...and so…
In a way an argument can be made … that the ECB not only creates scarcity of one form of collateral vs . another but that it also creates shortage of collateral by replacing high efficiency collateral with low efficiency collateral.
Here’s Barclays summing it all up:
Importantly, the shortage of government bonds would reduce the liquidity of the repo as well as cash markets. In the legal act of its public sector purchase programme (PSPP) the ECB stated that securities purchased under the PSPP are eligible for securities lending activity, including repos. This will be very important, in our view, to mitigate any negative implications of QE purchases on the repo market’s functioning.
And here’s Soc Gen citing liquidity as a possible reason for EU EGB relative underperformance (sans-bunds) vis-a-vis SSAs :
Our conclusion is that, once again, the market is responding positively to aggressive monetary policy but remains somewhat distorted due to the lack of structural adjustments. The underperformance of EU bonds – unchanged since the start of PSPP, while other issuers’ curves have flattened – may be partly explained by the lack of liquidity.
Soc Gen sees large scale asset purchases effectively limiting euro issuance in the SSA space and squeezing investors into higher-yielding issues:
Some issuers have advanced their programmes to well above 25%... However, what is very significant is the much smaller share of EUR issuance...The slowdown in the pace of new EUR issuance could therefore be the result of a pause after the strong start to the year, in combination with the shift to foreign currencies. However, the lack of interest from EUR-denominated accounts is no doubt linked to the implementation of the PSPP. By squeezing spreads so much, the ECB is crowding investors out of the sector.
Moving now to corporate credit and going a bit further in an effort to put the pieces together and paint a comprehensive picture, consider the effects all of the above are having outside of the market for EGBs and SSAs. From Barclays:
The ECB QE has caused a dramatic flattening of government bond curves and caused Bunds to trade with negative yields past the 7y maturity. There is now €1.9trn of negative-yielding government debt in Europe (almost 20% of the outstanding bonds) and this has a profound impact on investor behaviour. In fact, there is strong evidence that EGB investors are already heavily involved in short-end, highly rated corporate bond markets. From June 2014 to February 2015, short-dated, highly rated paper outperformed. We believe this reflects the investment constraints of bank treasury desks, which have responded to a lack of positive yielding collateral by taking more credit risk and more rates duration risk, but are unlikely to hold long-dated credit. This will lead to a persistent bid for this area of credit, capping shorter-dated, higher-rated credit in Europe.
The implications of lower government bond yields generally and the influx of displaced EGB investors are already being felt in credit markets, with 95% of the IG-rated market (excluding subordinated debt) trading below 1.5% yield.
[23]
Given that, we now need to consider everything we’ve said about illiquidity in the secondary market for corporate credit lately. Recall that reduced dealer inventories (as a result of new regulations) combined with high issuance (due to corporates looking to take advantage of record low borrowing costs) and in conjunction with investors’ hunt for yield (due to misguided monetary policies), have the potential to coalesce into a nightmare scenario, or, as we put it recently [24]:
Thanks to new regulations ostensibly designed to, among other things, bolster capital cushions and keep the market safe from the perceived perils of prop trading, banks are more reluctant to facilitate trading. This comes at the absolute worst possible time. Borrowing costs are so low that the Fed is basically daring companies not to take advantage, so while issuance is high, secondary market liquidity is non-existent meaning, effectively, that the door to the theatre is getting smaller and smaller and if someone yells “fire,” getting out is going to prove decisively difficult.
Here’s Barclays again, with their riff on the same narrative:
What is unique to Europe is the influx of non-traditional credit investors directly into €IG (not via ETF or the like) and their behaviour. Where the motivation is primarily to ‘hide away’ from negative bund yields, anecdotal evidence suggests the focus is on buying bonds of ’large, stable’ companies, with little discrimination between issuers of that kind. The risk is that we have what could be called a ‘Tesco moment’ where one of these ‘large, stable’ companies runs into negative headlines. Significant selling from non-traditional credit investors could ensue, which could spread into other credits owned by this segment. With reduced dealer balance sheet, there could be few buyers of such paper (in particular given tight valuations) and the spread widening could be disproportionate.
...and Howard Marks simplifying things:
Usually, just as a holder’s desire to sell an asset increases (because he has become afraid to hold it), his ability to sell it decreases (because everyone else has also become afraid to hold it). Thus (a) things tend to be liquid when you don’t need liquidity, and (b) just when you need liquidity most, it tends not to be there.
[25]
* * *
Coming full circle, we can see that some of the strain here could be alleviated if the ECB is able to implement an effective securities lending program so that €1 trillion of in-demand collateral isn’t locked away where the repo market can’t access it. JPM’s suggestion is for euro area NCBs to utilize existing relationships with dealers to facilitate this, and for dealers to then relax collateral standards “to ease the pressure on one country’s GC repo levels,” and to allow cash for collateral, in effect reversing the effect of the ECB’s low efficiency for high efficiency swap.
