In Capitalism, Failure Isn't Rewarded and Repeated
It was one of those tricky situations that proved yet again the difficulties of dealing with the fallout from monetary crisis. AIG had been bailed out by the government the previous fall, and then to begin March 2009 it reported the largest corporate loss in history. Only a week after that stunning news, word leaked out that the company was planning to pay some $165 million in bonuses to executives.
The numbers just didn’t add up to the public. In six months, the insurance giant had lost $100 billion or so, give or take, had received $182 billion in government support through several different means, and during that particular month in economic history the US economy alone had lost 800,000 jobs. In just March 2009. Everyone knew by then millions of further layoffs were still coming.
To then pay out $165 million to mostly the same executives who created this disaster? It was righteous to ask why in the world was anyone willing to bail out Wall Street.
The brand-new Obama administration was flabbergasted, in public. Treasury Secretary Geithner was at the forefront of the government backlash. Formerly the head of FRBNY during the worst of the crisis, Geithner had also headed up that systemically important branch since 2003 during the very period within which AIG had drastically scaled up its activities.
In March 2009, however, as Treasury Secretary Geithner demanded CEO Edward Liddy to flat out cancel the bonuses or at least renegotiate. It was insulting to reward such failure. The irony was lost on the media.
Liddy, who had been brought in after AIG’s collapse to clean up the mess, could do nothing about the bonuses. It was a contract issue not a matter of profit. Most of the payments were about retaining key employees. The cratered insurance firm as hard as it was to believe needed top talent as much to wind down the worst parts as it did in winding them up to near ruin.
The idea of Too Big To Fail (TBTF), or Systemically Important Financial Institutions (sifi) as they are now called, was more complex than it was often made out to be. For Democrats, the issue was regulation - as in a stunning lack of it. But AIG wasn’t short of regulation. Quite the contrary, in fact, the company as basically an insurance firm was regulated quite onerously by both state and federal agencies (including the Office of Comptroller of the Currency).
For (many? some?) Republicans, this should have been a market issue. Bankruptcy and failures were options that had served to instill discipline in market players throughout the country’s often rough and messy history, a fact of life that maybe the financial system needed to be reminded of at that moment in time. But this element wasn’t so straight forward, either. A capitalist system is benefited by creative destruction, but was this capitalism?
So much crisis review focused on AIG’s role in it narrows in on only credit default swaps (CDS). This is not without good reason, of course, but even then their massive purpose is still misunderstood (regulatory capital relief provided for primarily European banks). As is the proximate cause for AIG’s September 2008 downfall, which wasn’t CDS. It was securities lending.
This highly regulated insurance company’s securities lending business was arranged for the benefit of the highly regulated insurance company. How do we know? The company said so numerous times, including its 2007 Annual Report which states securities lending activities were undertaken “primarily for the benefit of certain AIG insurance companies.” What good are regulations when regulators don’t have any idea what the regulated are up to? Markets, too, for that matter.
The reason securities lending originated with regulated insurance companies is the one part that’s simple; they had all the securities. As any insurance company throughout history, AIG’s insurance subsidiaries received cash premiums and invested the proceeds in financial instruments. Like any of them, that meant they had acquired substantial portfolios of otherwise idle securities.
The wholesale system that had developed globally from the 1970’s forward, what I call the eurodollar system for most of its attention on the dollar and in offshore formats, opened the door for qualitative as well as quantitative expansion. Of the former, securities lending evolved particularly in parallel to the US housing bubble.
It may sound complicated in theory, but in practice it wasn’t all that difficult to unpack once you bothered to understand what each counterparty was doing and why they were doing it. One of AIG’s insurance subs would lend securities to another special purpose subsidiary, AIG Global Securities Lending. Acting as agent, this latter firm would then relend the same securities to banks and broker dealers who had created a need for them in burgeoning repo, derivatives, and offshore FX markets.
These banks or brokers would deposit cash with AIG Global Securities Lending as collateral for the securities. Cash was collateral for the collateral they sought (see what I mean about capitalism).
AIG, in turn, would take this cash and invest the proceeds in other things. Because these securities lending transactions were callable at a moment’s notice, prudent practice across the insurance industry (this was not uncommon outside of AIG) was to keep those investments to highly liquid instruments. The last thing you wanted was to take on too much liquidity risk, adding more maturity transformation to a firm-wide portfolio already stocked with nothing but.
This highly regulated insurer, however, didn’t do that. Bucking all manner of prudence and common sense, they had invested this incoming cash collateral heavily in subprime RMBS (about 65% of the almost $90 billion at the height of its securities lending business). This was not just the fault of AIG alone, it was the common refrain of the pre-crisis eurodollar paradigm. There was no risk too big that couldn’t be easily managed by this monetary arrangement growing exponentially in all directions.
In pure repo terms, what had happened was nothing more than collateral transformation; a three-legged approach that essentially used existing insurance subsidiary securities (good quality largely corporate bonds) transforming otherwise subprime RMBS to be used in a repo-like funding. The good collateral was pledged in one leg, the cash collateral coming back was then used to fund RMBS, which then served as collateral on the last leg.
From AIG’s perspective, it was able to repo the worst of the worst toxic waste on the highest quality corporate bond terms. Hidden leverage, this collateral transformation. Tim Geithner never saw it coming.
One of the reasons for this systemic blindness was that the system operating in this way truly did appear to be robust, backed by any number of seemingly solid redundancies. If you had trouble repo-ing your toxic waste, there were backups in bank standby’s, ABS commercial paper, and a rich menu of risk absorbing derivatives freely offered by the world’s biggest and most complex financial institutions.
And the “Greenspan put” behind everything.
These were proven starting August 9, 2007, to be something other than redundancies: bottlenecks. To begin with, as AIG was doing this stuff so were all the other firms (if not quite to the same repugnant degree) who AIG would be depending upon when things started to go south. The Greenspan put was even more worthless than those since the promise of central banks is given by blinded central bankers.
Collateral transformation, like repo markets themselves, isn’t something we can easily or readily observe. It happens, we know it happens, but finding it or measuring it with some level of precision is impossible. Like so many ultimately fragile systems we only really start to see it when it has become a major problem. By then, serious systemic damage is already being done.
One of the few purely collateral statistics we have available is compiled by Tim Geithner’s old outfit. FRBNY reports weekly upon information gathered from primary dealer reports. They detail repo fails, a fact of life in repo markets that is pure anathema to capitalism, as well as net UST positions in dealer inventories.
If there is an overall cloud of misinterpretation over the whole system, it also permeates down to the most basic layers. On first perception, the level of UST securities reported by dealers seems nothing to do with repo or collateral. If they are net long +$34.1 billion in UST coupon bonds with maturities greater than 11 years, as FRBNY shows in the latest weekly data, compared to just +$8 billion in the 2-3 year maturity range, this all seems like market processing; the direction of nominal interest rates, investor appetite, etc.
But the history of dealer positions does not ebb and flow with interest rate risk, except whenever that one risk is being defined by liquidity risk emanating from, or in concert with, collateral issues.
Take, for example, the middle of July eleven years ago in 2007. The word subprime had been blasted into the public consciousness and housing concerns abounded, but at that moment it was still more of a nebulous tail risk kind of scenario than upfront and immediate fear. FRBNY’s primary dealer report shows that for the week of July 18, dealers were claiming a net short -$178 billion in UST’s, -$199 billion including bills.
They were short, hugely short, not because they were betting on interest rates to start rising again (at the long end, interest rates hadn’t really budged since 2000) but because systemically they were all lending UST collateral all over the place. The huge minuses denoted the free flow of collateral into, and then out of, primary dealers in repo as well as derivatives.
By the time the Federal Reserve met in September 2007, with the events of August 9 in between, dealer UST positions (including bills) had changed dramatically to -$70 billion. Dealers started hoardingcollateral (hoarding = a positive change in reported positions; securities lending = a negative change in reported positions), adding another ugly dimension to growing issues in other parts of money markets. Since these were bottlenecks, they proved to be self-reinforcing and therefore procyclical (balance sheet capacity).
This is not meant for merely historical review, the further remembrance of ten years ago for the sake of trivia. Those long-ago events have come to still define our current day. Dealer hoarding has been an intermittent problem throughout the last decade, punctuated by extreme changes like those of August and September 2007 in these intermittent monetary dysfunctions that have kept the global economy from recovering.
Just when you think we’ve shaken off the last of 2008, boom, the eurodollar again.
The latest, sure enough, shows up in collateral the week of April 11, 2018. The week before, April 4, dealers reported UST coupon positions of +$54.2 billion, down (less hoarding) from +$59.3 billion the week before for that. For April 11, however, the total shot up to +$75.8 billion (a lot more hoarding). Including bills, the change was even more dramatic; from +$70.6 billion to +$104.2 billion (a whole lot more hoarding).
And this is just what we can see.
It would grow worse thereafter. By the week of June 13, dealers were reporting +$108.6 billion in coupons and +$131.9 billion including bills. These were the highest net long positions since the week of December 19, 2012 – the week after QE4 was announced.
Between the weeks of April 11 and June 13, this was the (latest) EM currency crisis that saw the “dollar” rise pretty much against everyone and everything; the renewed short squeeze of the global dollar short.
This collateral squeeze was spotted in other Federal Reserve statistics, too. In one of the Fed’s memo items, the central bank reports the balance of UST securities it holds on behalf of foreign official and commercial institutions. Between the week of April 18 and May 23, the level declined sharply, nearly -$74 billion. The contraction would continue until the week of…June 13.
This was one of the biggest drops on record. The last time we find one of similar magnitude was October 2014, -$74.3 billion. That specific month truly stands out in terms of collateral concern given the collateral-driven buying panic in UST markets that took place on its 15th day. A big one before that? Between the weeks of July 18, 2007, and that first FOMC meeting in September 2007, -$56.2 billion.
Where are these foreign-held UST’s disappearing to? No one knows for sure and the figures themselves tell us nothing about how or why they change. We are inferring collateral issues, an offshore collateral call, by the corroboration between them and dealer holdings as well as this “rising dollar.” All that plus historical experience tells us there is something serious going on out there in the somehow still hidden eurodollar world.
Maybe Tim Geithner should have been far more worried about other things.
And despite more positive surface conditions of late, dealers are still hoarding (as of the latest data) just as intensely and UST’s are still disappearing from Fed memos off into the offshore collateral netherworld.
This shouldn’t be surprising (but it will be written, again, as if it will be) because the history of collateral over the last eleven years easily explains why funding market participants would remain so on edge. To begin with, there was AIG proving the fallibility and inherent flaws of what was supposed to be as risk-free as anything ever invented in the monetary realm.
Beyond that, though, there has been an ongoing and chronic shortage for what may count as the basis of the collateral environment; so-called pristine securities like the corporates AIG Global Securities Lending was borrowing from its internal insurance brethren as well as the UST’s that still define the system.
At one time, almost any highly-rated prime MBS was treated in repo as the same as UST’s. Agency debt, too. Those have largely disappeared from the list of “pristine.”
Before 2011, any OECD sovereign debt security was taken this way, as well, and then PIIGS redefined the government debt sector, subjecting it to never-before-seen gradation. They were crossed off the register.
With a shortening catalog, the global system has been since 2011 busily and creatively transforming collateral again, with an updated twist. Insurance companies are lending their securities to brokers and banks who exchange them for junk corporates, US and EM, to secure the derivatives transactions of still others.
That didn’t work out so well in 2014, thus October 15.
We have every reason to suspect collateral transformation has continued to take place – up to around the middle of April 2018. We’ve even witnessed a near buying panic, though not so panicky, on May 29, 2018. On that day, UST yields fell sharply (huge waves of buying) as did those for German bunds (even bigger waves). No doubt a global collateral call right smack in the middle of this “dollar” squeeze.
In mainstream media convention, a narrative that starts with the Tim Geithner’s of the world, everything is great, booming even. GDP this quarter might even be 5% in the US, how could it be anything but booming?
GDP in Q3 2014 was also 5%. October 15, 2014, was just two weeks after that quarter had ended. The collateral pressures that were building up to that date had been evident all throughout those surely robust three months. Experience in 2015 and 2016 proved it was a boom, alright, just not of the economic variety.
History repeats because Tim Geithner had the public on his side complaining about AIG’s bonuses. People should have been up in arms instead about Tim Geithner’s role in missing so badly on AIG. In capitalism, failure isn’t rewarded and repeated.
