Though they never come out and say it, the growing body of evidence shows that what’s called quantitative easing (QE) is more akin to involuntary or inadvertent tightening. The implications are enormous, though still largely unexplored if not entirely misunderstood. We may now be seeing the global consequences of it currently at dollar auctions in Switzerland.
Our problem may end in dollars while transiting via the Swiss National bank, yet it begins in Italy where the ECB intends to help the most. Just a few months ago, policymakers were jarred by the rising spread between Italian government bonds and their respective German counterparts.
While officials called it (IT) an anti-fragmentation policy to make it sound, like QE, more than it is, in truth the idea was haphazardly thrown together at the last minute. Buying Italian BTPs while selling German bobls or bunds is a program any kindergartner might have come up with.
And it hasn’t worked, no surprise.
Fragmentation has proceeded despite (in part because of?) these childish plans, spreads continuing to rise to alarming proportions. The primary problem for everyone else in the monetary system is actually scarcity. This is something QE does only too well. Study after study has (finally!) shown just how badly central bank bond purchases screw with the vital repo system’s collateral sufficiency.
One of those published by the BIS back in 2019 appropriately titled Euro repo market functioning: collateral is king, it noted the clear demarcation in certain repo spreads right when the ECB’s QE (called the Public Sector Purchase Program, or PSPP) got started in March 2015.
“After 2015, as the ECB purchased bonds, the German SC segment exhibited the highest and most volatile premium, consistent with the scarcity of collateral…The introduction of the PSPP, however, pushed the scarcity premium into negative territory for all GC collateral segments.”
The paper defines this “scarcity premium” quite simply as the difference between GC repo rates and the ECB’s deposit rate floor. In any normal environment where collateral is sufficiently supplied and redistributed, there would be no reason why any bank might lend at a GC repo rate less than what it might obtain simply parking funds with ECB and getting paid its deposit rate.
Just like US Treasury bill yields dive below the Fed’s deposit rate equivalent, the RRP (reverse repo program), whenever euro GC rates do go below the ECB’s floor – and persist that way despite negative rates – it can only be for the one reason. Thus, the “scarcity premium.”
This spread is most obvious and evident for the better issues, those offered by Germany as well as in French OATs. Italian or Spanish bonds attract some premium, but nowhere near as much as those others.
When this does happen, however, there should be dealer banks standing at the ready to arbitrage spreads. If you could borrow cash at such obscenely low rates in GC repo because you have access to the collateral the market is paying such a premium for, then “[lend] cash against any collateral that attracts the higher rate and [borrow] cash by offering the collateral that attracts the lower rate.”
Yet, the study’s authors find that given such an arbitrage opportunity it mainly doesn’t happen; or, while it does, not anywhere near enough to reduce let alone eradicate these spreads and premiums and close collateral gaps.
“This means that less than one billion of collateral swaps per country pair are driven by repo spreads. This is a small amount compared with the overall market turnover of €200–300 billion. Importantly, given our estimates of the price impact of trades, these positions are too small to eliminate rate differentials.”
Where are the dealers?
The primary explanation put forward is, ahem, fragmentation, meaning nationalism. Not of the political sort, rather convenience and home market. For example, repo for each type of sovereign collateral, at least what is cleared on exchange, is dominated by a single issue. Nearly all cleared repo using German securities takes place at BrokerTec; MTS Repo for Italians.
Furthermore, these individual platforms tend to be further dominated by national specialists. Traders who regularly trade BTPs in MTS Repo don’t often have trading experience or capabilities with BrokerTec.
It’s an interesting theory, sure, though hardly compelling and certainly not enough to explain persistent collateral issues which have hampered repo markets thereby the financial system broadly.
To arb any repo irregularities requires first and foremost the balance sheet capacities to undertake both ends of what is really a collateral for collateral swap. Borrowing cash where collateral is scarcest, when GC repo rates for that sovereign issue are incredibly low, to then lend against collateral where rates are higher means transforming one kind of collateral into another; moving collateral from where it isn’t so scarce to where it is.
The very sort of function the global system did well (maybe too well) before around August 2007.
So, if there isn’t as much demand or premium on, say, Italian issues over at MTS but there is for the Germans on BrokerTec, find a partner to borrow the cash in BrokerTec using bunds, transfer the funds interbank (correspondent) so that you can lend in MTS Repo on Italian collateral.
Italian collateral goes in and comes out as German.
It’s the incoming collateral that really might be the issue rather than any differences in clearing platforms, national markets, or geographical boundaries. As spreads rise for BTPs against bunds/bobls/schätz, maybe no one wants to take the risk of holding the BTPs particularly when liquidity for it grows suspect.
This is, after all, the only real characteristic any repo counterparty cares for. If the cash borrower who posted the Italian debt defaults on the borrowing, the lender has the right to seize the instrument and sell it to be made whole (including any interest). But if the market for that instrument isn’t predictable because liquidity in it dries up, you just can’t take the chance of losing money on a repo trade.
Such fears aren’t theoretical, either. Back in November 2011, the FOMC was drawn into a discussion about Italian and Spanish bonds as it debated a response to what was (wrongly) called the European Sovereign Debt Crisis.
“MR. SACK. Yields on Italian and Spanish debt have increased over 100 basis points over the past three weeks, with the 10-year yield approaching 7 percent in Spain and surpassing that level in Italy. Liquidity in these securities has deteriorated significantly as the perceived risk has increased.”
The “scarcity premium” for German and French issues exploded during the second half of that year, while, not surprisingly, being absent for Italian and Spanish securities. There was also an enormous premium (this isn’t included in the paper) for US Treasury bills, too.
Between August 17, 2011, when the crisis really got going, and January 11, 2012, the yield for the 3-month T-bill was never higher than 2 bps, and was more often zero. Rates for the 4-week were quite often nil, and even the 6-month bill was frequently entertaining the same low range.
Collateral doesn’t get swapped only between same denomination markets, such as euro-denominated Germans for euro-denominated Italians. How many times and in how many billions (hundreds?) might a euro-denominated Italian be used to secure a dollar-denominated Treasury?
We have no idea.
In fact, this is one of the primary weaknesses not just from this particular BIS paper but for our overall knowledge base. If it isn’t exchange cleared – and most repo is not – then whatever takes place takes place without anyone but the parties directly involved (borrower, lending, and often dealer) knowing about it.
Should someone be borrowing US dollars starting from Italian BTP collateral, transforming it along the way into USTs, what might happen to them if the BTP issue like in 2011 (or 2018) becomes, well, an issue? Under the ideal scenario, dealers would arbitrage and there would be alternatives if slightly more costly to the Italian collateral user (the dollar cash borrower).
But, as we’ve seen, arbitrage has become a tricky business and that’s just across Europe. Crossing oceans, spanning continents, getting wrapped up in the eurodollar dealers, the collateral deficit gets complicated real quick.
Doubly so during periods when central banks have made it all worse by removing some of the best quality issues mindlessly purchasing instruments so as to raise the systemic quantity of otherwise useless bank reserves. Like a funnel or really a bottleneck, there was a reason the “scarcity premium” in Europe sharpened so much following the ECB’s PSPP.
And again during 2020’s much larger PSPP2, not to mention the Fed’s turbo-charged QE6. Each may have flooded the world with bank reserves, but what did they really do?
Our beleaguered, increasingly flustered (by lack of alternatives) Italian bondholder at some point it might even make sense to shop among central bank offerings as a last resort. If the market won’t provide useful levels of funding for borrowing dollars on Italian collateral, then, say, the Swiss National Bank would (drawn from the Fed’s overseas swaps).
As of this week’s dollar repo auction in Switzerland, now seventeen banks have bid for $11.01 billion; up from fifteen taking $6.27 billion last week and nine securing $3.1 billion the first week in October. There’s at least a short run progression here.
No, $11 billion is not much in the grand scheme; a drop in the bucket. Back during 2011, overseas funding at these central bank auctions (mainly ECB) topped $100 billion total – not that it did much good. Just like the $600 billion or so in the thick of the 2008 crisis.
That’s the thing. Use at any of these swaps in any significant amount (defined as anything more than test operations) corresponds with only the worst circumstances of the last decade and a half.
The first time began all the way back in December 2007, active right on through 2008 into 2009 followed by 2011 and 2012, before again in 2020. None of those periods can be associated with anything but gross financial instability causing or at least contributing much to serious economic recessions and not just for Europe.
Why? As the BIS paper’s title displays, collateral is king rather than bank reserves. Not just repo, either, even more derivatives where the complications are greater and even less trackable (since those transactions are mere footnotes to bank reports). Unless there are dealers willing and able to intermediate, some lucky few hit up a central bank while the rest are left stranded in the shadows which devilishly disrupt markets and sharply curtail commerce.
And of those we can see stranded at something like the SNB auctions, it tells us something very important about those global money dealers and what they might not be so willing to do – exactly what functioning markets and economies need them to.
Despite rising yields pushed upward by central bank rate hikes and expectations for more, bond market participants continue to bet on rate cuts sooner than later. It’s not just this Fed “pivot”, rather why the Fed might be forced to pivot.
Since officials there have made it absolutely plain how they’re sole focus is glued to the CPI, any involuntary turnaround would have to come from either a rapid drop in it (say, inventory liquidations due to recession), or some monetary disruption so substantial it takes that focus away.
Curves and markets, particularly collateral premiums have indicated rising stress and therefore the exact probability which is now breaking out in something like SNB dollar swap use. If it was only $11 billion in Switzerland, this would be interesting yet largely not concerning.
But it is everything, from the screaming dollar exchange value to the swaps market right down to spreading inversions (including an unprecedented one in, of course, Germany).
In other words, this Swiss stuff merely the latest in a series of escalating shadow issues still following the same course set last December (and even before) when eurodollar futures inverted for the first time. Rather than move in a different direction, at each juncture the world and its reserve currency system (eurodollar, not dollar) converge closer to the downside scenarios which had initially alarmed, who?
Our missing in action money dealers who also know the difference between what QE really is and how the world actually works, or doesn’t.
