Moving Beyond Generic 'Money Supply' Thinking
There are clear signs that monetary policy, indeed monetary thought, is evolving. That would seem to be a given with recent history of the last half-decade, but ideology is often loathe to change anything even with clear empiricism defining against it. But that is really the challenge here, as there is a very real operational standard to which monetarism has to meet. Failing to do so will be, in no uncertain terms, catastrophic once more.
Outwardly, there is very little to project such changing content. From a lay perspective, monetary policy seems about the same as it was the whole of the last forty years, particularly the last three decades (or thereabouts) under interest rate targeting. For all the "newspeak" about "forward guidance" and better managing investor expectations, front and center in all of it remains the federal funds target. Yet that is by now an antiquated notion that has very little relation to actual bank and financial function, especially as it relates to the global dollar standard, the eurodollar.
In recognition of that, we were presented something of a radical shift in the Fed's reverse repo program. It was described publicly as another tool to help the FOMC manage the "exit" of "emergency" monetary measures, but in effect it is instead a appreciation (very belated, about a decade too late) that the actual and practical nature of monetary policy has shifted out of the 20th century versions. The target on the federal funds rate is nothing more than a guide for projecting policy intentions, but at some point the financial rubber needs to meet the banking balance sheet road. The reverse repo was one method, it was thought, of actually keeping the public face of policy but welding to it a durable alteration toward repo and shadow finance.
Of course, shadow banking is no longer so shadowy in at least awareness of its existence. However, the actual manner in which it functions is still quite murky, even to those who have spent years and years peering through the fog of technical jargon, as well as the counterintuitive and often impenetrable entanglements that render interpretations and description as something out of science fiction.
My own history here began in the 1990's as I kept on running across these "unusual" indications that were supposedly elements of "money." As one does in these situations, I started with "base" money and worked outward, except by the time I got to M3 there were these other components that nobody could seem to explain, let alone define. Instead, my inquiry was far too often the first time anyone had even heard of such things (which, somehow, remains true today). As the broadest cross-section of the "money supply", M3 contained both eurodollar deposits (both overnight and term) as well as repurchase agreements (both overnight and term).
Anyone who observed or participated in the autopsy of the Savings and Loan crisis would easily note the role of the shadowy elements in the buildup of what was, until the 1970's, a sleepy little section of banking. Despite the outward appearance of the uninteresting local, neighborhood bank, some of the biggest S&L's were growing themselves with repurchase agreements, which really didn't make much sense until you began to see the contours of what they were and how they fit.
The idea of a repurchase agreement makes very little sense in what was transpiring, especially since the owner of a security remained the same and entitled to all interest payments on that security. There was no actual transfer of ownership at all, which belies the notion of a sale and repurchase, which is how it is legally crafted - possession mattered most, which makes collateral here almost "money-like." Those niceties relate only to the accounting treatment and tell you nothing about the incentives in doing the transaction. A repo is just a collateralized form of transferring cash back and forth, done so because counterparties on each side are looking for either the cheapest interest rate in funding or as little risk to cash as can be reasonably obtained.
Altogether, eurodollars and repos amounted to about 1% of M3 in January 1970. They were essentially negligible to the banking system, though their very existence at that time showed that the foundation had already begun to shift. By January 1980, there were (at least as measured by the Fed's statistical inferences) about $100 billion in eurodollars and repo outstanding - a significant increase to about 5.6% of total M3.
By the time the S&L crisis broke into the wide open, repos and eurodollars were up to 8.2% of all M3. Then, they went backward into hibernation as regulatory changes and the retreat of the S&L's really from history pushed banking into a very uncertain place. It was there that the early 1990's were so concerning to policymakers; they really feared a rerun of the Great Depression at that moment. And in some ways, though fleeting in broad economic recognition, you can appreciate that palpably, as any chart of total credit and broad money shows a sharp slowdown and nearly a full halt in the early 1990's.
That represented a great demarcation between what was dominant before and what has been dominant since. The traditional mode of banking has really been irrelevant in the credit system that rebuilt from the ashes of the S&L crisis. While the shadow elements, this wholesale funding via eurodollars and repo, were in use prior to the 1990's, they were to become central to the system as it attained dominance.
When the Fed ultimately stopped tracking M3 in 2006, repos and eurodollars were just over $1 trillion, 9.8% of all M3. But that was only what the Fed could see and measure, leaving off a huge proportion that went unseen and untracked. Thus, the impetus for dropping M3 was really an acknowledgment that the wholesale funding model had not only become dominant, it had gone to places unimagined by contemporary definitions and understanding. Worse, it had done so in a way that defied all attempts at actually trying to estimate just how much "leverage" had been applied, and in what manner it was used.
I have relayed many times before the BIS post-crisis attempt at measuring the "dollar short"; that they came up empty but not for lack of trying. Just from European banks, they figured somewhere between $2 trillion and $6.5 trillion by 2007- that was the total "funding gap" the European banks rolled with between short-term liabilities and long-term assets, both denominated in "dollars" but for the most part containing "eurodollars" on both secured (repo) and unsecured terms.
That meant that "money" supply had become more than just even M3 or its components in eurodollars and repos, as the global dollar short was really a concoction of all manner of bank balance sheet efforts - from currency swaps to interest rate swaps to mathematical expressions of risk and even bank "capital." Again, no wonder the Fed stopped trying to estimate all of it, as it was beyond impractical (and remains so, unfortunately).
That itself, however, should have caused great alarm inside the regulatory and policy complex. When the crisis erupted in August 2007, they were ill-prepared for what was to come, particularly as they debated largely 19th century policy prescriptions to fit a system whereby nobody knew how to track or even define "money."
As I said at the outset, that is changing somewhat, with the reverse repo as a very belated attempt at recognizing the shift toward at least repo (eurodollars will remain another matter, though, like reverse repos, I have no doubt that both the Fed remains confident its dollar swaps with other central banks have bridged that gap and that it will ultimately be proved wrong on that account once more - also like the reverse repo). Even the academic character of the Fed has been altered toward the shadow reality, as some of the best work at gaining a basic level of understanding has been offered by a few participants inside the system (at the Fed itself, at the Treasury Dept. and in ultra-national institutions like the BIS and IMF).
There is at least some urgency on the part of pieces of these bodies to come to grips with all of this. One such attempt, a worthy paper in several key aspects, was published in November 2013 that linked very much the anachronistic policy framing with the modern conventions of collateral and wholesale "money." Titled, Bagehot Was A Shadow Banker, authors Perry Mehrling (Columbia Economics Professor), Zoltan Pozsar (US Dept. of the Treasury), James Sweeney (Credit Suisse) and Daniel Neilson (Institute for New Economic Thinking) argue that modern banking is not all that different from the 19th century.
Bagehot's main thesis was the central banks should be the liquidity backstop to asset prices and banking in times of distress, "lend freely at high rates on good collateral." In that ancient "money" system of actual money, lending currency at high rates meant exactly that, though doing so in practice is not often so easy (Great Depression). What Mehrling et al do is transfer that idea to shadow banking, recognizing that "currency" is a fairly sturdy concept that survives even shifted to some other bank property and function.
In this framing, currency and elasticity properties can be applied equally to collateral in repo markets. In other words, a bank run in the modern sense does not take the form of depositors exchanging fractional claims on currency, but rather a shortage of usable collateral to allow free-flow of "liquidity" from one perception to the next. In that strict sense, a central bank aware of this derivative distinction can act not just as a "lender" of last resort, but more likely as a "dealer" of last resort.
"The modern shadow banking system, at its core, bears a surprising resemblance to the 19th century world that Walter Bagehot helped us to understand in his magisterial book Lombard Street: A Description of the [London] Money Market (1873). At the heart of both worlds is the wholesale money market, and operating as crucial liquidity backstop in both worlds is the central bank. At the time Bagehot was writing, this backstop function was not yet fully understood, much less accepted; much the same could be said of the central bank's backstop of the shadow banking system today (Capie 2012). We are living today in a Bagehot moment, when the outlines of the new are just emerging from the ashes of the old."
Unfortunately, that last sentence is taken literally; which is more than a shame since as a neophyte in the 1990's I saw that content shift all too clearly, particularly as those around me were unable to explain either eurodollars or repo, and thus simply ignored them, like economists and policymakers, as holding little appreciable distinction and thus as irrelevant. My own journey down the rabbit hole began at that point; if no one else seemed interested in eurodollars I might as well find out for myself.
Tugging at that thread meant pouring over not just government statistics of "money supply" and trying to relate definitions of pieces that seemed ill-suited, it directed me to actual participants in the global system, the first of which was, for me, Enron. As if that didn't open my eyes quite wide enough, from there I followed to AIG and other "insurers" that seemed more interested in securities lending than property and casualty policies. Yet, there was such a casual dismissal of all this taking place, as if it weren't really real, that was very unsettling.
It was then that I made up my mind that there was nothing much of the former banking system to be found in the 21st century operation. So in reading Mehrling et al, and the attempt to tie backward to the "father of central banks" is a bit unnerving for several reasons. Again, I think there is much to admire in their attempt, and I also think they have "it" right in that there are very heavy similarities, particularly in the concept of elasticity, between then and now.
Those fuzzy echoes, however, are not enough, for me, to allow safe passage intellectually. Almost from the start, the foundation is set awry by what seems like a very conscious attempt at downplaying that last part of Bagehot's creed, "lend freely at high rates and on good collateral." In the 19th century version to which everyone was all new, the means of "wholesale" funding were discounted trade bills; thus collateral was not just some nondescript financial instrument but a personal guarantee of property and reputation. Good collateral was as much literal in that respect.
Undoubtedly the modern orthodoxy is colored by experience (or revisions thereof) in the 1930's. The idea of sound collateral has been lost in the noise of politics - the Fed was founded on great mistrust of ordinary people, and thus markets. Too much focus has been determined on the concept of "good" banks and "bad" banks, whereby, in the early 1930's, orthodox logic posits that "irrational" investors unsuitably drove prices "too low" that "too many" good banks were destroyed alongside bad ones. That is a lot of subjectivity about the character of markets, but it forms the basis of not just opinion but central bank operation in the current era.
Once you start down the road of sifting, via government, which banks are "bad" and which banks are to discard market prices there is only one logical destination - that all prices are inherently "wrong" whenever they are deemed to be contrary to the "public good." That totally upends Bagehot's creed, as what he said was almost mechanical by process, that prices have already determined which institutions are "good" by the amount of solid "collateral" they can put up to a central bank.
Certainly, complications thus arise from comparing the collateral of discounted trade bills from England in the 1800's versus mezzanine tranches of some private label mid-2000's securitized mortgage pool. However, the main point stands - central banks have a vested interest in determining not just what institutions are "good", but what forms of collateral take that distinction as well.
This is not just some academic parsing on my account, either. We have practical experience in just this fashion, as the ECB has amended its collateral rules more times than anyone can count these past few years (and is doing so yet again in the name of covered bond "buying" and targeting its "purchases" of asset-backed loans for small and medium businesses). They are not alone, given that the Fed itself was taking all sorts of financial securities as collateral in 2008, including equities and absolute trash credits (with "appropriate" haircuts). The expansion of "eligible" collateral is the new definition of elasticity, exactly as Mehrling et al have it.
Part of this new determination toward upgrading policy understanding (which is itself a sort of delayed admission of failure) is to almost stylize the role of financial components. In liquidity, at the very center, are the dealers. In every instance I have seen, and this counts beyond just the paper I cited above, dealers are believed to be some benign functionary of mechanics - almost like the Bagehot imagining of discounting bills. A dealer runs a "matched book" for this framing, meaning that it is nothing more than a conduit through which "flow" occurs.
This is a false narrative, but placing it as such essentially sterilizes the function of dealers to be inside of the "good" collateral category regardless of merit. In reality, dealer banks do not run matched books in anything they do, though they often might seem that way. We know all-too-well of the history of proprietary trading, which belies any notion of matched book philosophy. In other words, dealers take positions and thus express "market" ideas of their own. However, under the cover of this Bagehot as a shadow banker, those "market" expressions are peremptorily deemed to be "more valid" than anyone else's; because they are given that "good" collateral benefit of the doubt right at the outset.
This is the operational extension of "too big to fail", in that dealers are put upon a pedestal and thus given carte blanche on all manner of behavior. My intent here is not to say that dealers should be restricted or that they have behaved inappropriately, only that such judgment should be reserved to market forces for retribution. In the bigger picture, however, it supersedes any notions of central banks as "dealers of last resort" because it is an element of liquidity risk that transcends even collateral.
In other words, "dealer of last resort" under this framework is not even enough since dealers, given the benefit of the doubt as all "good" collateral, that express "risk" positions even in the course of hedging (over-hedging into speculation) dealer inventories essentially line up in the same direction over time due to central bank interruptions and interventions. That means that collateral is not the only consideration for liquidity, as dealer capacity in derivatives and the ability to absorb additional balance sheet "risk" is severely compromised under certain conditions. If every dealer is non-neutral in the same direction at the same time, there is nothing to absorb a monumental shift in the other direction - exactly what happened starting in August 2007.
A central bank that is committed to providing collateral in that case would be as ineffective as a central bank trying to truck Federal Reserve Notes to global banks during the 2008 panic. Indeed, the Fed itself on numerous occasions opened its holdings of collateral to the "markets" to little avail - even going so far as to work with the Treasury Dept. to essentially "create" collateral in the form of Supplementary Financing Bills. It was not nearly enough.
The reason was far more rooted than those could alleviate, owing to that basic premise that in the modern system dealers are neutral, and thus present nothing but good collateral. From that position, dealers essentially follow each other into "risky" tendencies all at the same time (chasing yield, performance, whatever in contrast to interest rate targeting's repression), and therefore essentially make even the collateral-shaped "rescue" impossible. In other words, the behavior of central banks does not compel a "dealer of last resort" but a full "market of last resort" whereby central bank activities must become an override of nearly all pertinent market prices - if dealer holdings are prejudged for non-market reasons to be "good collateral" a central bank is thus beholden to enforce all prices of dealer holdings to be as such! That is so far beyond Bagehot's conception as to be almost wholly unrelated.
Thus what started out in September 2007 as just lowering interest rates morphed to providing "dollar" swaps by December 2007; to serious collateral eligibility expansion by May 2008; to further collateral degradation and even collateral creation by September 2008; to simply "buying" as much MBS as possible through QE1 to hopefully overturn all market determinations of price. If the "market" would not accept dealer holdings as "good" collateral, the Fed would instead change the entire "market."
In that empirical sense, the desire of examination under the orthodox framework to see dealers as neutral is actually more basic and more corrupted - they have to assume it in order for any of it to survive a rebuttal; again, too big to fail in its gory details.
There are other faults with this line of understanding, mostly as it relates to the persistent desire to separate even dealer functions. Not only are dealers believed benign and functionary, the new orthodoxy seems to want to displace derivatives dealings from traditional collateral flow. I cannot figure out why that would ever occur, except that it allows simplicity, that old element of econometrics whereby reality must bend to conceptual compartmentalization and generalizations. In fact, everything I have tried to describe here is inseparable from each other - the very reason central banks have backed themselves into becoming the "market of last resort" is entirely due to the fact that dealer activities are actually complementary among various classes, especially derivatives. The proclivity of a dealer to engage in a liquidity providing transaction is as much about risk control, through hedging and derivatives, as it is supply of collateral. Thus, liquidity itself is not divisible into components, but very comprehensive across all functions tied together by very imperfect expressions (made so by, again, central bank intrusions).
All criticisms here aside, there is much to be hopeful about with these new academic searches toward catching up policy to actual banking beyond generic "money supply" thinking. These are very smart people who we can only hope have enough influence to begin to shape actual programs. The reverse repo program suggests that is bearing out in at least modest FOMC evolution, though its comically inept operation to this point leaves more than a slightly pessimistic interpretation about their ability to turn this new understanding toward anything actually useful.
