There Is No Money In Monetary Policy at the Moment
When last we left Citibank, it was in the middle of 2014 very briefly overtaking JP Morgan for the top spot in derivatives dealing. For the longest time, ages even, JPM ruled that space due to the nature of its business. As the primary triparty custodian in repo as well as the many other money dealing activities the bank sat on top of, to find them so far out in advance of all the others simply made sense.
According to bank call reports aggregated and published by the Office of the Comptroller of the Currency (OCC), at the apex of this eurodollar world in Q3 2007, JP Morgan showed $91.7 trillion in gross notionals, primarily interest rate swaps but including also a (un)healthy dose of credit derivatives. That was up from $48.3 trillion just two years earlier. To be king, you had to expand by all means – parabolically even.
JPM’s nearest competitor, Citi, was doing much the same just not as fast. By Q3 2007, that bank’s derivative book had reached $34.0 trillion, up from $20 trillion in the same two years.
For JP Morgan, that was it. There was going no further than that level of what is true leverage, if unaccounted for and off in the shadows. It wasn’t and isn’t strictly leverage for itself, rather systemic capacity offered to the rest of the global monetary system to achieve monetary ends. By the middle of 2014, it had cut back in that offered capacity to $68.1 trillion, allowing Citigroup for one quarter to get ahead.
Citi took a different track. Though (because?) it had received the largest rescue (bailout) in US history, by the time the panic was over, starting in Q4 2009, the bank’s derivative book began growing once more. It jumped sharply in 2010 under QE1 and QE2. We see the same effects time and again at various banks, even at JP Morgan. In the latter case the bank didn’t expand its book so much as pause in its shrinking.
The correlation with QE might propose some kind of direct monetary linkage. After all, Citi and JPM were primary dealers who in their capacity as investment banks, loosely, were at the receiving end of those QE transactions in both UST as well as MBS. In other words, on-balance sheet on the liability side there was this filling up on bank reserves created as the byproduct of Permanent Open Market Operations.
But their very different trajectories in these further reaches of the shadows prove it wasn’t so. Instead, the overall opposite QE effects in general were left up to the interpretation of each bank rather than any monetary contributions. It seems very clear that Citi believed not that QE was money but that if enough people did so it would work anyway (inflation expectations). JP Morgan was far more circumspect about its potential, and perhaps of no surprise given its central role above all others in creating the modern, 21st century dealer mode of operation.
Those two conflicting views would finally meet in absolute terms in the middle of 2014. Noting that Citi had taken the top spot from JP Morgan, a bank spokesman in September that year attributed it “as a result of client demand.” They were betting on QE, now on its third and fourth iterations, that it would finally produce the recovery everyone had all along sought.
What does it say when a bank goes so far in that one hopeful direction, to where its derivative book more than doubles from 2007 to $70 trillion by the middle of 2014, and then over the next two and a half years cuts it all back by 40%? To me the answer is obvious, and I think over those three years now a lot of other people are starting to see it, too.
For many, however, that was just prudent regulation, the thundering adoption of stricter bank rules attempting to well after-the-fact address the excesses of the eurodollar upswing. The people who believe this never account for the differences in behavior between banks as well as both before and after the inflection of this “rising dollar.” JP Morgan clearly needed no Dodd-Frank or Basel III. It has been withdrawing if unevenly going back to August 2007.
Whether or not the other American dealers followed JPM or Citi seems to be a function of their perceived probability that things will work out favorably in the end. The same is true for overseas eurodollar banks, though their operations and books don’t show up in the OCC statistics. Oftentimes, you will find from individual bank filings a similar dichotomy between banks that might otherwise seem very similar; Swiss banks UBS and Credit Suisse, for example, took very different paths after the 2011 re-crisis, or at least until, like Citi, the middle of 2014.
In the aggregate, the pessimists outmuscled the optimists so that overall the trend in derivatives dealing has been almost steadily lower after 2011. The timing itself proposes exactly this idea, that between the full panic and its later, smaller cousin, there was a great deal more indecision about how it would all work out. The balance was more optimistic than after 2011, where the number of banks behaving like Citi was more plentiful if also less enthusiastic. In derivatives as well as other eurodollar spaces, 2011 was for a great many the last straw.
It remains to be seen whether 2014 and the “rising dollar” was the same for these remaining holdouts.
As a measure of offered balance sheet capacity in the systemic construction of leverage, it wasn’t prudence so much as lack of promise. JP Morgan, after all, bet in its own limited way on the recovery very early on, only to suffer in its CIO office the indignity of the London Whale fiasco. That episode might further propose that maybe it wasn’t JPM’s internal desire to reduce its money dealing capacity, but in reference to Citi’s spokesman instead the product of JPM’s customers demanding it; meaning not demanding more of this (dark) leverage.
I know that this idea and these processes are unclear and still elicit a great deal of confusion. Derivatives are by their very nature difficult to understand even for professionals on the inside of the financial services business. A few paragraphs offered in this context are woefully inadequate, but minimally required nonetheless.
The whole paradigm revolves around bank balance sheets. Almost everyone can conceive of the traditional money multiplier, the basic, ancient idea of fractional reserve banking. A bank takes in deposits of money and from that offers claims on it to several people at once. It can do this because people don’t all want to hold money at the same time, and when they do there are more than enough people willing to leave it in the vault for those wanting it back from the bank.
Deposits and therefore loans were created by this process, governed by the liability side concerned mostly over the ability of the bank to meet any sustained withdrawal of whatever (over time, and as economies evolved, it was far more than just gold) resided in its vault.
That system is largely no more. Regulation has moved mostly to the asset side. The Basel Accords were the final step in that direction, an idea that bank managers had agitated for going back into the 1800’s. But how does a multiplier work on the asset side?
The basis for this approach was twofold; the first being that definitions of money especially with the rise of eurodollars (meaning both offshore as well as wholesale funding) were no longer so easily determined. The second rationale was the growing emphasis on quantification and mathematics. By the 1980’s, it was believed that banks with government approval could describe themselves numerically; meaning that through capital ratios they could present to the public an objective, unambiguous snapshot of their operations. To do so would allow them to be easily categorized as a “good” bank, or a “bad” one. Liquidity wouldn’t be so hard if that could be accomplished.
It had been the history of bank panics where “good” banks were often destroyed the same as “bad” banks. One of the founding fathers of the Federal Reserve, Senator Robert Owen, was in his prior occupation a banker in Oklahoma who in the Panic of 1893 nearly lost everything – and in his view his bank was both solvent and strong. It didn’t matter because the public in its emotional state withdrew money indiscriminately. A central bank providing currency elasticity was supposed to prevent it from happening again.
But it did in the Great Crash of the early thirties, the waves of bank failures plunging the US and the world into the Great Depression. Bank examiners would often judge a firm in good standing by all prudent rules and exercise, and it would end up suspended or closed anyway. The Fed’s introduction hadn’t been near enough to through theoretical currency elasticity keep money multipliers from imploding.
Capital ratios had been around through all of that, and had been included in the bank examiner’s handbook ever since there were bank examiner’s handbooks. They had never before played so prominent a role because it was extremely difficult to make good use of them. A bank that invested primarily in government bonds was not at all the same as one that invested primarily in low quality mortgages or commercial loans. Yet, assuming that raw dollar amounts were the same or similar at each of those institutions, simple capital ratios would treat them equally.
The concept of risk-weighted assets (RWA) intended to overcome this limitation and achieve useful discrimination. Holdings of government bonds would be given a low risk-weighting, whereas holdings of unqualified mortgages or loans to commercial entities unbacked by sufficient collateral would be treated with higher risk-weightings. The net effect, from the regulatory perspective, was that the first bank buying only government bonds would need to hold far less capital than the one buying/lending in only risky assets, a seemingly sound and legitimate proposal.
The view from the banking system was the other way around; meaning that as they saw it you could hold that much more assets if they were to be somehow classified by a low-risk weighting. Though it may not seem it at first, this is a money multiplier. Changing bank assets by RWA is the same as achieving leverage without seriously disturbing capital ratios.
The trick was in being able to change risk-weightings. It isn’t immediately apparent how an unqualified mortgage that initially would be assigned a 100% weight could then appropriately and legally be given one of 50%, or less. The answer is simply hedging, or derivatives.
If in very general terms, you lend to a company in what at the start is a relatively risky position but then purchase a single name credit default swap from a qualified counterparty in good standing (highly rated) written on a more liquid but different corporate name with much the same characteristics, that kind of “insurance policy” against default would greatly reduce the risk of the original loan (theoretically). So long as it met the myriad accounting standards and tests, it would even qualify for an often much reduced risk-weighting.
Again, from the perspective of regulators, it would (properly) allow the bank to hold less capital against what was at the start a risky loan made outwardly less so. From the bank’s point of view, a shift from the 100% RWA “bucket” to, let’s say, the 50% bucket would mean the firm could do two such loans for the same amount of capital.
From this simple example you can begin to appreciate what from the outside appears to be nothing more than an insurance policy or raw hedging instrument, the CDS, when it is actually a primary element in bank leverage or money multiplication under this modern format. When Bear Stearns nearly failed (it really did, but technically, I suppose, it was bought out) it was “well-capitalized” by every regulatory standard while at the same time people marveled how that was possible given clearly ridiculous leverage on the asset side of its balance sheet. It wasn’t ever the intent of regulators for this kind of disparity, but that is always the downside of regulation – especially bank regulation.
In March 2001, a Senior Financial Economist at the Office of Thrift Supervision in Washington, Mark D. Flood, wrote what was a pretty standard paper for the time attempting to analyze the effects of Basel on US banks (as was that agency’s directive). Almost all of it was focused on whether or not the regulatory rules had inhibited US banks and bank lending, especially in its early adoption period which happened to be the dramatic credit crunch of the early 1990’s (that was the full impact instead of the S&L crisis).
There were only two mentions of “regulatory capital arbitrage.”
“Of particular interest for us is the research into the effect on bank loan portfolios, which falls roughly into three categories: (a) whether the accord caused reductions in overall bank lending; (b) whether the accord caused banks to reallocate within their portfolios away from (toward) assets with high (low) risk weightings; and (c) whether banks have accommodated the accord through cosmetic accounting gimmicks known collectively as regulatory capital arbitrage.”
Flood’s paper further fixated upon securitizations as a means to achieve “capital arbitrage”, though that was to be expected given securitization was at that time exploding throughout the global banking system. And it also played further into the role of derivatives (pricing, liquidity, and correlation) in achieving this dark form of leverage that regulators were struggling to just minimally comprehend.
The 2008 crisis would eventually expose these practices for that leverage formation, as well as the often mind-blowing extent and excesses of it. The most prominent example was AIGFP’s super senior credit default swap portfolio that in the majority ($379 billion gross notional) was written for primarily European banks to, in AIG’s 2007 words, achieve just this sort of “regulatory capital relief.”
It wasn’t then nor is it now about just credit derivatives, or even really derivatives. They are the means to the end, and that end is the mathematics of balance sheet management. A bank balance sheet is governed by much more than capital ratios, all sorts of things that in the modern era now quantify risks and present value calculations that in each and every one open a door to using techniques to arrange the math in the most charitable of ways.
An interest rate swap, for example, can be used to reduce the downside probability of an MBS position, so that the net present value of holding that security therefore goes up (lower volatility) even after accounting for paying whatever premium on the swap. A higher NPV and lower modeled volatility makes for a much (outwardly) less risky security. And then you can even hedge the hedge, or barbell it, or whatever else. There are even ways to reduce “tail risk” probabilities, too, if not the actual “tail risks.”
Why did JP Morgan offer in Q3 2007 some $90 trillion in gross notional derivatives? Most people can never get past the number “90 trillion”, and simply see it as greater than the whole US economy six times over. The scale of notionals, as well as the use of derivatives, is just that misunderstood. I’m not saying there is nothing to the concern over the size of derivative books, only that the size actually tells us much more important things about the demand/supply dynamics of this hidden, shadow leverage.
When the eurodollar system, and that is a term that I use often too loosely, but in my analysis encompasses all this stuff (witness AIG’s CDS going to Europe to manipulate balance sheet mechanics for mostly dollar assets), was growing exponentially, so, too, were derivative books for the simple fact that one could not happen without the other. Those derivatives were needed to achieve the favorable mathematics that allowed leverage to increase so dramatically and worldwide (the US housing bubble wasn’t the only credit explosion that needed to be financed by leverage, there were all those EM “miracles”, too).
And it was offered cheaply because before 2007 money participants were lulled into a false sense of security. Everything was considered low risk because of both recency bias (nothing systemically bad had happened, so it was assumed nothing bad could happen) and the mathematics that was at the time treated as something like scientific fact. If you bought XYZ subprime tranche and paired it with a hedging derivative from highly-rated AIG, why wouldn’t it be a great deal for everyone involved, directly and indirectly?
You get a higher return on something that would otherwise be a huge risk, but with the security that if something goes wrong AIG will stand behind it and compensate, and with a substantially lower capital charge. AIG gets to pocket premiums on a CDS its math as well as its offsetting hedges declare almost no risk whatsoever to be paid out. Regulators are happy that all this happens with only well-capitalized banks freely lending and buying securities. Policymakers are ecstatic, too, because all that lending and buying has achieved this Great “Moderation” for the global economy, making their quarter-point adjustments in federal funds or Euribor seem like true genius personified.
It did all seem, on the surface, just so stable and moderate.
It doesn’t, however, anymore to JP Morgan, Bank of America, Goldman Sachs, HSBC, Deutsche Bank, etc., etc. It may not, as well, to each bank’s customers in this arena. Maybe demand is down simply because money dealers can no longer price this leverage so flexibly and low. Having seen risks on both ends, no regulation is necessary to do what common sense seems to have done. Credit default swaps, for example, aren’t even offered anymore at any price.
What was only exponential growth prior to August 2007 has been since revealed as, frankly, idiotic. All that math did was to hide leverage and lull everyone, even those providing it, into a false sense that risk could be so easily set aside and managed. You can’t destroy risk, but you can, for a while, get enough people and banks to believe it possible. A “no risk/all return” paradigm is, in fact, too good to be true.
This is not to say that leverage and math have disappeared. The CDS market is well on its way to just that fate, but there is a whole lot left amongst the rest. The difference is one of degree, not yet a binary choice. That’s why the global economy is stuck where it is in low-growth hell. The monetary eurodollar system has shifted to in the aggregate appreciate the inherently unstable nature of this condition, and therefore persists in knowing the contradictions but without a plausible exit (either economy, meaning actual recovery, or the full replacement with some other money system) from them.
Derivative books in Q1 2017, according to the OCC, rebounded. Clearly leading the way was, perhaps not surprisingly, Citi (along with Wells Fargo on the second tier of dealers). It for the second time overtook JPM for the top spot, only this time the bank did so at a much reduced level; just $50 trillion compared to $70 trillion not yet three years ago. Nothing ever goes in a straight line, even a now decade-long decay/awakening process. The relative increase in offered capacity (leverage potential) in Q1 would certainly help the Fed to explain its latest conundrum, why it supposedly “tightened” but little seemed to do that.
That is the final point for the whole ongoing affair. In everything going on as I described above (which is still just the tip of the ice berg), where does the Fed fit in? It doesn’t, maybe apart from psychology. There is no money in monetary policy, and there isn’t even money in this modern money multiplication. There is, however, a bunch of derivatives, mathematical liquidity processes, and balance sheet techniques that don’t obey accounting conventions or geographic borders. It’s easier to just call it all eurodollars.