Central Banks Think They're Still in 1956
It is not unusual to find US Treasury securities trading special in repo as the calendar winds closer to government auctions. Repo market participants prize on-the-run (OTR) collateral more than anything since it is the most liquid, the one aspect above all others that cash owners demand before engaging in secured lending. There is, then, a variable area of turbulence as the current OTR security heads toward off-the-run (OFR) closer and closer to being replaced in the unending line of auctioned bills, notes, and bonds. Liquidity at that fuzzy point, collateral liquidity, is less than ideal.
Since OTR securities are by far the most liquid and traded versions, it isn't surprising that there would be at times a significant premium for them as compared to those bonds already passed into OFR. There is not, however, any set rule or schedule by which this transition must take place. Some auctions are conducted smoothly with barely any notice; others are noteworthy events. The problem is that these kinds of inner workings in wholesale finance do not fashion well to the dominant views of money that passes for mainstream commentary.
Two months ago there was just that sort of unusual instability in repo. The 10-year maturity, in particular, had been trading not just special but special all the way up to and even above the 3% fail penalty. Trading special means that any piece of collateral is being bid in the repo market by cash owners willing to "lend" you their cash at a rate below the general collateral rate, so low at times it can be negative (they are paying you to lend you cash but in reality they are giving you cash and a negative rate in order to borrow your security - you just don't know if you'll get it back which is why you charge more special). Obviously that means in these times that, at the margins, collateral is more "valuable" than cash. Where that imbalance is greatest you find a rush or surge in "fails", meaning a great number of counterparties decide they will at the end of the repo just keep the security rather than take back the cash that was lent.
The worst of those was in early October 2008 and the repo stats from that time are unbelievable. The New York Fed estimates that in the last week of September 2008 total fails among primary dealer inventories were $3.5 trillion (combining both "to receive" and "to deliver"), up from $577 billion the week before when Lehman first failed. By the week of October 15, total fails hit $5.3 trillion and stayed at that level again the following week. The S&P 500 lost 32% during these weeks where repo fails surged.
To counteract this "hoarding" of US Treasury collateral, a 3% fail penalty was instituted meaning that there would be a high cost of failing to deliver collateral. For the most part, the penalty has worked in keeping fails far below those desperate days during the panic. The middle of March 2016 was not such a period.
Treasury collateral fails had soared to $380 billion the week of December 16 and then $392 billion for the last week of the year. Those were the highest total fails since June 2014. Both of those periods correspond to obvious acts of illiquidity that followed them; in the middle of 2014 it was the start of the "rising dollar"; just after December 2015 was the latest and as yet most intense global liquidation.
For the week of March 9, 2016, the New York Fed reports dealer fails of $889 billion, by far the highest since 2008 and greater than at any point during the 2011 crisis. This was not a surprise, however, as repo specials proliferated throughout several maturities with trading close to and even slightly below the calculated fails penalty (it's not a strictly 3% rate, it is a calculation related to federal funds). The following week, the actual week of the government funding auction, fails held high again at $873 billion. The middle of March is a seasonally strong part of the auction calendar, so clearly OTR to OFR was in that dreaded "dead spot" of transition but this was way, way out of those proportions.
Mainstream commentary trying to explain this unusual occurrence has been, to put it bluntly, absurd - when there is any mention of it at all. It just doesn't fit the "narrative" of a fully restored financial system pledging only great thanks to the wisdom of QE's and the sustained recovery that orthodox faith still believes. Since the media defers totally to orthodox economics, all explanations begin only with that premise, which accounts for the absurdity. It is the repo equivalent of the round fails peg being bashed into the "recovery is complete" square hole.
Essentially, we are told, primary dealers have been forced to fund greater inventories of UST's because foreign central banks had been selling far too many of them. Primary dealers are supposed to act as a liquidity buffer especially at times of stress to ensure that prices do not move too far too quickly. While that much may be true, it just doesn't apply here. UST's were never in danger of being sold off dramatically; quite the contrary, treasury yields have almost uniformly fallen. If dealers were acting as a liquidity buffer, they were doing so when the market was a much greater such buffer than they were.
Why, then, would they hold so much federal paper that it would lead to irregular funding circumstances for the whole of US dollar repo, including their own liquidity condition? The "wisdom" of the mainstream explanation is that they had trouble offloading this inventory. Absurdity may be too kind of a qualification as we are supposed to believe that dealers who did not wish to hold UST's at that moment could not find enough buyers at the same time the whole world was piling into UST's. You don't have to know anything about repo to know what to make of that suggestion.
Unlike 2008, however, there was no obvious mark of illiquidity destroying asset prices. Though the repo market indicated some great disruption, a cursory examination does not find it ending up anywhere obvious. Stocks have (had) been rising since February 11 while oil and junk bond prices surged, too. However, as they say all great lies contain a kernel of truth so it is here. The key part is the foreign central banks "selling UST's."
While an American audience may not easily see the fruits of repo disorder, a more foreign focus finds it quite easily. As I acknowledged contemporarily on March 18, highly negative cross currency basis swaps in Japan, a country identified in TIC figures as one of those heavily involved in "selling UST's", indicated "something" highly unusual:
"The negative yen basis swap acts like leverage where even yields on the interim ‘investment' are negative. Any speculator or bank with spare ‘dollars' could lend them in a yen basis swap meaning an exchange into yen. Because you end up with yen you are forced into some really bad investment choices such as slightly negative 5-year government bonds, but that is just part of the cost of keeping risk on your yen side low. Instead, the real money is made in the basis swap itself since it now trades so highly negative. The very fact of that basis swap spread means a huge premium on spare dollars; which is another way of saying there is a ‘dollar' shortage."
It certainly fits the surge in repo fails far, far better than trying to claim primary dealers were having such great difficulty selling treasury inventory in a market with absolutely no shortage of buyers in size. It was such a robust bid that dealers could have made more money by not only selling what they had but also selling what they didn't (going from net long to net short). From the view of Japan and the further evidence provided by the New York Fed of dealer net holdings, it is abundantly clear that primary dealers were actually part of the cohort doing the collateral hoarding and paying the penalty to do so.
Confusion surrounding these kinds of problems arises not just from seeming complexity but more so the fact that economists and their hold on the mainstream convention views money and the financial system as if it were still capitalist and free market. Repo fails make absolutely no sense to a capitalist; how can one party unilaterally decide to not live up to its obligation to return a financial security at the agreed upon moment?
If the bedrock proposition of capitalism is private property and clear title to it, then repos have done little except make a mockery of it. First of all, the term "repurchase agreement" sounds abundantly clear and concise: one party sells something to another with an attached agreement to buy it back at some later date at some predetermined price. Legal title to the "something" should easily follow accordingly. But as with so many financial practices, the intrusion of "financial" changes everything.
On October 24, 1968, Union Planters National Bank of Memphis, TN, sued the federal government over taxes charged by IRS assessment for tax years 1961-1964. At issue were some municipal bonds that the bank contended it owned but the IRS assessed that it did not. While seemingly a tax case, the central issue was repurchase agreements. Union Planters claimed that while holding the bonds in questions, which were obtained in repos from various securities dealers in and around Memphis, it was the beneficial owner of those securities and thus any interest gained in redeeming bond coupons was municipal bond income and exempt from taxation.
The Tax Division of the Department of Justice argued that the transactions were collateralized loans, and that the securities never passed ownership and therefore subjected the bank to income from interest arising from a loan to the securities dealers.
The case was a tangled mess. At the start, there was no uniformity in how Union Planters' counterparties treated the transactions. Some dealers held the municipal bonds as continuously owned on their own books with instead an offsetting and corresponding financial liability to the Bank, while others recorded contingent liabilities to the Bank with, as the lawsuit pointed out, "one dealer advertised for sale to its customers those bonds which had been transferred to the Bank." The United States District Court eventually found in favor of the plaintiffs, Union Planters, that the Bank "exercised all elements of control and collection over the interest coupons" and ordered the Bank entitled to a judgement of recovery from the IRS; but not before the judge noted in his opinion that the transactions did hold several key elements of lending arrangements.
Part of the problem was due to the fact that banking regulations themselves were not clear upon how securities in repo transactions were actually structured. In September 1956, the Office of Comptroller of the Currency ordered the agency's bank examiners to begin treating repurchase agreements as loans. That was further clarified (clouded?) in the summer of 1957 by an order from OCC in the treatment of repos. Prior to that point, repo agreements under Section 5136 had not been subject to any limitations when involving United States obligations as the securities at issue since they were treated as investment transactions. Starting August 15, 1957, OCC with approval and authority of the Secretary of the Treasury issued new regulations under Paragraph 8 Section 5200, which effectively placed capital restraints upon repo transactions to a single counterparty.
Now treated as a loan rather than distinct purchase and sale trades, repos would be subject to capital limitations (up to 100% of capital and surplus). For its part, Union Planters complied with the updated regulations and booked each repo as if they were loans as required; recording them on its financials as "Loans in Repurchase Agreements on Securities." If that had been the end of the regulatory intrusion, there never would have been a court case. Instead, OCC reversed itself in 1964 using still other statutes as the foundation for apparently renewed respect for investments. Citing Paragraph 1131 of the Comptroller's Manual not subjecting repurchase or resell agreements to limitations provided under 12 U.S.C 82 and 12 U.S.C. 84, OCC reneged and declared them once more "neither borrowings nor lendings but rather are purchase and sales."
The justification for the contradictory nature of the OCC's flip flop was supposedly the balancing of very clear Congressional intent upon imposing lending restraints on banks and what the OCC had itself ruled for the prior seven years, as that was surely a glimpse into the future of non-tradition (wholesale) banking. Without explicitly stating it as such, in the reversal of 1964 we can reasonably infer the rationale for turning repos into lending in 1957; OCC declared then that the "only possible justification" for treating repos as loans was the possibility of some transactions taking place via banks wishing to evade lending restrictions under 12 U.S.C 82 and 12 U.S.C. 84. Thus, the reversal in 1964 taking repos back to investment transactions was, so it was declared, a desire to avoid punishing all banks for what a few might be doing nefariously.
By the agency's own words and declaration, it seems that as early as 1957 it became aware that "some" banks were engaged in a form of regulatory arbitrage to avoid strict counterparty limitations in these types of financial, funding transactions and arrangements. We have to assume that the practice was at least conspicuous and widespread enough so as to force the OCC into what was really a dramatic alteration in the banking paradigm. After all, even by 1957 repos were becoming a significant part of the increasingly modern American banking landscape. To suddenly amend what was standing practice going back decades indicated then, and even now, shifting behavior.
Of course, this very gray area in the legality and format of repurchases is something of an original sin. The Federal Reserve itself holds no statutory authority with which to conduct reserve operations via open market repos as collateralized loans. However, by construing them as purchases and sales, as Union Planters was arguing half a century later, the Fed covers its actions under its legally authorized power to buy and sell government and agency securities.
Many cases that followed Union Planters referred to it as a landmark case in the area, but almost as much, it seems, to distinguish the instant case as to confirm the legal status of repos in systemic fashion. In other words, Union Planters did not serve as clarification for the financial system so much as to signify rather the opposite; repos were a mess. Just a year and a half later, another court case was filed that stood as an almost exact replica of the Union Planters case. In American National Bank of Austin v. United States, the Bank argued that it, too, had engaged in municipal bond repos also in the years 1962 through 1964 - during OCC's repos-are-loans-phase - which the Bank now petitioned for a refund on taxes that were assessed by treating the transactions as loans.
In the initial trial case, a nonjury trial, the District Court agreed with the Bank that the repos were purchases and sales, not loans, and ordered the IRS to issue the claimed refund. The Appeals Court ruled against the District Court, essentially confirming that repos were to be treated under individual circumstances and further that the track of established trial precedence, including Union Planters, which was then being appealed, was all over the place.
Because of the inability to centrally and generically clarify what a repo was to all peoples and parties, they remained individual executory contracts to be determined idiosyncratically. From the perspective of money and banking, then, you might understand the difficulty in accepting a repo as some new kind of money or bank funding arrangement. It was, for the longest time, treated as only one form of remote investment from the perspective of both sides of a repo transaction.
For non-banks, in particular, this flexibility was one of the central elements driving tremendous growth. Local governments, for example, were required to only invest in US government obligations, but they could do so in short-term money markets where the repo was structured as an actual purchase and sale rather than collateralized loan. But the inherent structure of repos hid perhaps their own monetary character. Because repo transactions were ultra-short term maturities and because the closing end of the transaction was often in the form of federal funds, counterparties could write checks upon their repo balances because the checks would clear after the federal funds were delivered. Repos were, in essence, a deposit-like substitute.
Large companies in order to get around deposit interest restrictions could enter into repos with commercial banks to gain money market interest rates. From the perspective of the commercial bank, the company's deposit balance declined to be replaced instead by an interest-bearing asset; exactly like the investment properties that were assumed. Because of that transformation, the interest-bearing asset would not be counted in monetary aggregates. This was true even though the maturity was ultra short-term and the company could still treat the repo as if it were a demand deposit balance and write checks against it.
The effects upon the bank were much more substantial than that, however. Repos offered what was essentially a non-reservable funding mechanism. The law and banking examinations treated repos as investments and therefore had only a limited treatment of them within the context of what they viewed as money and banking. Banks, of course, were never limited exclusively by regulation but rather had always sought to circumvent them as much as humanly possible. As noted before, the Office of Comptroller of the Currency had by 1956 seen enough of this in practice of repos to completely reinvent their statutory treatment - to see them as collateralized money market lending as surely they were becoming. The fact that OCC reversed its ruling eight years later only suggests that banks had complained effectively enough in political terms to allow them to turn repos into what is now standard global practice.
As with eurodollars, the operation of domestic and global banking in the 1950's and 1960's served to sever the relationships of economy and money in all these different and undefined ways. That was, in many cases, entirely the point. We are conditioned to think of inflation in a strictly quantitative sense of "more money" chasing fewer goods. What we find in the years just prior to the Great Inflation, relevant still to our own more dour circumstances, is that money was altered qualitatively. True money operates as property under property law; eurodollars, repos, and other wholesale monetary deliveries operate exclusively under financial law far more malleable and flexible.
Those terms favor banking and the "money" sector over the real economy; they also demand the one-way of expansion. The wholesale system that spread from there to every corner of the on the credit-based "dollar" as reserve currency was as much a qualitative expansion as quantitative. In fact, without this transformation in the monetary basis massive, rapid volume expansion especially in the 1990's (post-1995 in particular) never would have been possible. Eurodollar banking operated "best" where fuzziness was greatest (even though, it should be pointed out, the terms of repo were standardized in the 1980's including who actually held title to collateral, but by then it had already accomplished pushing the financial limits as we see in fails where title often means nothing). It was (seemingly) its source of greatest strength and now it is its greatest weakness.
Because a eurodollar is not a thing like a dollar is a thing, nobody knows what to do about its decay. The eurodollar standard is a system, not money (strictly speaking). Thus, the Federal Reserve expanding its balance sheet to more than $4.5 trillion was entirely inappropriate and not surprisingly, as even economists are starting to concede, ineffective. One form of additional bank liability, even in great quantity, does not offset malfunction in all the various forms of the eurodollar system. Central banks are not at all equipped to handle qualitative disruption - they just think they are because they are still in 1956. The rest of us have to muddle through the continuing aftermath of financialism.