Long before too big to fail became a thing, Citigroup and the Federal Reserve often worked together. Traders at the bank would quite often talk and exchange action advice with their ostensibly regulatory overlords. Chummy, sure, but also a necessary working relationship, nothing nefarious, just standard practice.
When funding markets began to freeze up in 2007, Citi was there – at first. One of Wall Street’s biggest names was among the first to jump into the fray, and do so in the manner consistent with current monetary theory.
What used to be called the Discount Window, its name changed to Primary Credit way back around 2003, didn’t exactly work the way it is taught in schools. We are meant to believe banks shut out from short-term wholesale funding (or suffering depository issues, once upon a time) head down to their local Fed branch and apply for emergency redress.
Offering up the eligible collateral, the central bankers save the day. With the Primary Credit rate set above market benchmarks, they even do it in a manner perfectly consistent with Bagehot’s doctrine of lending freely at high rates on good collateral.
Elasticity.
Like everything else, in practice it gets far more complicated. To start with, stigma. The Discount/Primary Credit window isn’t anonymous, banks singling out the Fed’s H.4.1 release on that count. Show up on the list, you become the private market target the very next day (if not before).
Everyone is aware of this issue, yet before the events of 2007 no one expected much to come from it. Why? Dealers.
Money markets are not monoliths, each far more complex than the textbooks make them appear. Even something like the federal funds effective rate is, in some cases, an illusion. While it weighs by volume to come up with a median we all hope represents the vast majority of what takes place, the truth is such statistics can become wildly misleading.
Take, for instance, the case where a firm or especially a group of wholesale borrowers are perceived as riskier. It could be because of idiosyncratic reasons, or just general uncertainty which leads to more careful scrutiny. Whatever the cause, this unlucky cohort finds it more difficult, meaning more expensive, to borrow today when compared to yesterday.
Should the cash-rich lenders in the market as a whole charge too much of a risk premium on this group, what’s their recourse? Not Primary Credit, because that would just finish off what the marketplace had already started.
Sensing an opportunity, some dealer in good standing (literally and figuratively) would access Primary Credit instead if only to relend funds taking advantage of what should be an easy arbitrage – even after paying the penalty rate to the Fed.
So long as what the rate the afflicted borrower pays is high enough to compensate for the risks, then the Fed’s balance sheet is forwarded into the marketplace through the actions of these dealers whose own balance sheet implications should be minimal given funding originates at the Federal Reserve window whatever it might be called.
These commercial banks aren’t stupid, though, nor reckless. Just like Primary Credit is collateralized, so, too, will our heroes demand collateral. Maybe they will accept a wider variety than the Fed branch might, still the first rule is to protect yourself; arbitrage is meant to be as close to risk-free as possible.
You can, I trust, already see where this scheme begins to break down.
Going back to 2007, Citi was doing just this. Weaker firms – perceived by reputation as well as what collateral they might have had available – were facing much higher borrowing charges in the private market. Citi and several others not yet in the same boat would bid at the Primary Credit Window and then relend to the aggrieved.
In fact, the very first move the FOMC made in the Global Monetary (not financial) Crisis was to reduce the rate charged for Primary Credit. Understanding how this works (for once), officials expected by offering a better profit opportunity more dealers would get involved and at higher volumes.
According to the transcripts from the FOMC meeting held at the end of April 2008, Citi’s traders would even go so far as to call up their FRBNY counterparts to let them know just how much of a help the commercial bank was being.
“MR. HILTON. Coming back to one of the other questions that Don had about what we are hearing about stigma from some of the banks, one of our better contacts, Citibank, as Jim mentioned, used to do a lot of arbitraging and using the discount window, the primary credit facility. On occasion, after they borrowed to re-lend in the market at a higher rate last year or so ago, they would call us in the morning to let us know how it was that they were helping us out with the funds rate.”
Yet, they were talking about Citi “helping us out” in the past tense by then. The very next thing Spence Hilton said was, “That has pretty much stopped cold…”
Hilton thought the reason the Fed’s former friend began to refuse the arbitrage was worry over its own potential for stigmatization. No doubt that was, in part, true, yet it wouldn’t explain everything else that was going on.
This was, for April entering May 2008, a critical juncture. Ever the optimists, particularly where their own handiwork was to be assessed, those at the Fed believed they had admirably weathered the worst of the storm. Bear Stearns in mid-March, would, many expected, prove to have been the nadir, the point of maximum danger.
However, there were these worrisome signs still lingering more than a month and a half following Bear. What had triggered this April 2008 FOMC discussion was exceptionally wide variation across intraday money market trading.
The federal funds market would be “firm” early in each New York morning which was, alarmingly, consistent with rates in overseas dollar markets, those like LIBOR. And while, today, LIBOR is considered an unreliable measure, back then it was backed up by something ICAP had created called the New York Funding Rate (talk about misleading; New York?)
In other words, while overseas dollar markets were in the midst of their prime sessions, dollar funding was particularly difficult. That spilled over into New York, and then calmed down in New York once the overseas (dollar!) markets shut down for the day.
When they did, the fed funds rate wouldn’t just “soften”, it would drop considerably such that by the end of the session the whole fed funds market appeared very differently from where it started.
Which was the more appropriate description?
More to the point, why were these kinds of sessions still more frequent than not? As one FOMC official, Gary Stern, pointed out, “It sounds like an obvious arbitrage opportunity.” And this was, in exactly the same way as Primary Credit was supposed to be.
In fact, dealers had several options for making “easy money” here. Not only Primary Credit, also accessing overseas dollar funding made available by what’s now called Central Bank Liquidity Swaps (you’ll see them referred to in the transcripts as foreign exchange swaps). I’ve always noted them as overseas dollar swaps, yet the terminology doesn’t matter because in the end they work as well as the Discount Window.
At this time, the balance offered in those liquidity, foreign exchange, overseas swaps was $36 billion and was about to be raised to $52 billion before ultimately $62 billion by the end of May 2008. What might have seemed plenty of funds readily available, accessible via either the Swiss National Bank or ECB. Borrow dollars from them, relend at a higher rate, profit.
In response to Mr. Stern, Mr. Bill Dudley (who else?) responded, “Well, we’re not seeing much arbitrage” before Mr. Hilton chimed back in with, “We are finding a great reluctance to do intraday arbitrage. We’re hearing this from the banks that in the past would do that from time to time.” This then led into that discussion of Citi noted above.
But, unlike Primary Credit, there is no stigma when borrowing dollars via the ECB, SNB, or any of the other later central bank counterparties. So, why no relending?
There was, on the other hand, borrowing. If not for arbitrage, someone was bidding for that eventually $62 billion and all in the months before Lehman/AIG. It was, as things turned out, the very European banks who were suspected to be “weak” firms all along. They couldn’t get dollar funding from US markets, US banks, nor elsewhere in the eurodollar system.
So, they bid for the proceeds from overseas swaps at the same time procuring the majority of the funds offered at the Fed’s TAF auctions. As Mr. Dudley realized by the end of June 2008:
“MR. DUDLEY. In contrast to the U.S. auctions, the bid-to-cover ratios in the ECB and SNB auctions have risen sharply over the last three auctions. This likely reflects several factors including (1) a reduction in the willingness of U.S. banks to lend at term to European banks—due mostly to balance sheet constraints…” [emphasis added]
And there it is. Stigma, maybe. Balance sheet difficulties, now that’s the whole ballgame.
The Fed can and does offer some liquidity and funding to some firms who are sufficiently “good” as to directly borrow at Primary Credit, TAF, or from the foreign central bank counterparties on the opposite end of these foreign exchange, central bank liquidity swaps.
For everyone else, they were supposed to rely upon dealer arbitrage only dealers can’t and won’t arbitrate when their own house is in gross disarray.
But why were dealers balance sheet-constrained in the first place? Because of all the reasons these convoluted schemes don’t work. The monetary system becomes inelastic which produces widespread volatility, the death knell for dealer balance sheet plasticity.
Collateral chains shorten, collateral availability threatened, further reducing the ability of the entire global monetary system to remain necessarily elastic. The entire hope of the monetary policy regime is on dealers who become entangled in the mess anyway; make it worse because their constraint means problems for everyone else leading to more volatility and less systemic liquidity.
This is generally why none of these Fed programs produce the desired effect; the Fed isn’t rescuing the system as is widely believed and still claimed. Rather, it offers some reprieve to small parts of it leaving far too many stranded with no options. Dealers who helped out a little in 2007 were conspicuously gone in the wake of 2008’s Bear dealing.
Within months, despite $62 billion in dollar swaps drawn, Lehman, AIG, Wachovia, etc., the worst monetary crisis since that which caused the Great Depression.
During the worst of that crisis, these dollar swaps would swell to a maximum of $583.1 billion. During the worst part. This didn’t tell us how good a rescue the Fed was providing, it gave us just a small window into how restricted (including collateral) dealers had become. That is how you can have nearly $600 billion end up at the Fed and still the most devastating worldwide monetary panic anyone had seen in four generations.
Despite all its fluffy media sensationalism, the Federal Reserve’s reach into the actual dollar system is very limited. Dealers not the Fed do the dirty work of elasticity. Arbitrage is a central component of keeping the wheels necessarily greased, and it takes balance sheet capacities to get this done and keep it getting done.
All the swaps and TAFs won’t make much or enough difference when money dealers are constrained. In fact, we know when this happens by how much the “lucky” parts of the system show up for something like TAF or at the foreign end of foreign exchange swap auctions.
When swaps get used, nothing good is happening nor coming. Dependable deflationary indication, not the systemic rescue we’re meant to picture.
So, last week, the Swiss National Bank reported it had accepted $3.1 billion in bids at its regular weekly US dollar auctions tied back to the Fed’s Central Bank Liquidity Swaps. Normally, there are zero bids for zero dollars. Then, you might have noticed, things aren’t going so well lately, particularly around Europe.
Many were quick to finger Credit Suisse, a certainly troubled bank in the news of late for all the wrong reasons. Yet, the SNB helpfully reported there had been nine banks accessing this auction.
Then at this week’s auction, SNB reports double the bids ($6.27 billion) from now fifteen banks. Given what’s been going on around the eurodollar system this year and particularly since the middle of September, more obvious dealer constraints have ruled out relending here.
Swaps go up, it’s because the monetary system is going down. Even though $6.27 billion is a small amount, the truth is, like 2008, if anything was going according to the textbook including the usefulness of bank reserves sitting idle and fat on the balance sheets of dealers there’d be zero at swap auctions alongside little to no volatility and unmistakable signs of inelasticity.
That’s the thing about these days; dealers have tons and tons of bank reserves. But what they obviously lack, still, the capacities to do anything. Ask yourself why. Better yet, ask the man presiding over the above conversations, Mr. Ben Bernanke. They just gave him a (shared) Nobel Prize precisely because he still can’t answer this question.
The rest of us, however, we do have to answer for it.
