The Real Economy Is Far More Than Banks and Regressions
In May of 2014, Deutsche Bank raised some eyebrows when it suddenly announced it had both privately floated 60 million shares while obtaining necessary underwriting commitments for an additional 300 million more. The anchor investor in the first was Paramount Holding Services, the main investment fund vehicle of a Qatari Sheik. The €8 billion in new "capital" was somewhat of a shock, but not entirely surprising since Deutsche had been underweight in its "capital" ratios dating back to the panic. The new subscriptions were anticipated to bring the bank's Tier 1 (under Basel III) to 11.8% from 9.5%.
Deutsche Bank itself had made plain for years prior to that that the bank would rebuild its "capital" base through profitability. Indeed, that was the plan at a good many global banks, if not all of them. Everything had been proceeding to that plan for the German behemoth until the middle of 2013. Under its Strategy 2015+ program, the bank was anticipating €8.5 billion in adjusted IBIT for FY2013 but more irregularity for 2014 forward. Whereas the trajectory for Tier 1 had been nowhere but up, striking 10% in Q2 2013 from only 6% in Q2 2012, from then on it was flat to slightly lower.
Given that trend, the stock sale in May wasn't again unexpected, but the size was in what it represented about bank operations moving forward. Deutsche Bank was not giving up on the profitability trajectory; it was only reloading resources to move in a slightly different direction to do it.
Last March, the Federal Reserve under its yearly Comprehensive Capital Analysis and Review, colloquially known as the stress tests, failed Deutsche Bank (as well as Santander, and only conditionally passed Bank of America) not in terms of capital absorption potential under some "doomsday" scenario (which usually don't amount to much of any doomsday) but in the qualitative systems and procedures that would allow a bank to actually identify and then quantify with some manner of predictability changes in credit and financial positions. The relevance of Deutsche Bank outside the Fed's approved macroprudential procedures is its derivatives book above all else.
In response to the Fed's rebuke, Deutsche announced it "had" hired 1,800 employees, "dedicated to ensuring that its systems and controls are best in class." Absent from that claim is when these were hired, as it was likely that the bank had been hiring regularly in the area of systems management since 2008. But according to Reuters, a former "risk chief" at the firm, William Broeksmit had been warning in December 2013 and January 2014 that its stress test scenarios were "way too optimistic." It may have just been a fluke, but in a December 6, 2013, email to the bank's risk committee, Broeksmit claimed the bank's "severely adverse scenario losses in 2014 and 2015 ... look way too small compared to history."
Just recently, of course, Deutsche Bank reported the largest annual loss in the company's history for CY2015 - "besting" even 2008. While legal fees and past wrongdoing are often attributed as the profit damage, the bank's internal revenue sources particularly its investment bank are the real problems. In Q4 alone, revenue there fell by 30% at the unit, which is based, not coincidentally, in London. All Deutsche would claim about those results was that Q4 was a "challenging environment."
While that was certainly an understatement, Wall Street and all its London eurodollar peers had undertaken just the sort of qualitative and systems assignments that were supposed to mitigate against any such surprises. Indeed, all of Wall Street had been on somewhat of an increased engagement with employees of ratings agencies, as turnover in 2014 and early 2015 was somewhat unexpected. As Bloomberg reported in February last year, some 300 analysts had left major ratings agencies since the crisis for work at the banks or other credit shops. In 2014 alone, 80 analysts had switched sides out of 4,000 total.
The purpose of Bloomberg's exploration on the topic was conflict of interest, in that analysts might not be impartial upon ratings for fear of disturbing future job prospects (a form of regulatory capture), but there were other considerations relevant to now. What were all these analysts to do at these banks once in their employ? You have to consider both the regulatory environment and the manner of bank "investment" which are more complimentary than they might seem. Ratings analysts are good at complex statistical models; just the kind that might cost a bank its "qualitative" standing with the Federal Reserve.
It's exceedingly odd, but that is the world as it is today where statistics dominates. You are quantitatively deficient if the calculations are negative, but qualitatively deficient where the ability to calculate is calculated as negative. To beef up qualitative control systems was to approach new and more "modern" versions of mathematical models especially banking's ages old nemesis - correlation.
Straight away, there should be some puzzlement about this kind of arrangement, particularly since ratings agencies did not transit the crisis in 2008 with sterling reputations. In fact, in both official and convention, ratings agencies shouldered a lot of blame for the worst financial debacle since the Great Depression. But even that overall sentiment, as true as it might be, misses the point somewhat. Ratings agencies themselves were not the problem so much as it was (is) how ratings were determined. There are no psychics working through divining bank vaults or rows of teller stations, nor are there bank examiners in the sense that prevailed before the 1970's; those funny little old guys with white hair and glasses who would read through reams of bank accounting statements, conduct numerous lengthy interviews and then apply their considerable experience to form judgments as to "soundness."
Ratings had become, as all finance has, an issue of statistical models. As any programmer will tell you, the model is only as good as the modeler; you put in garbage and you get out garbage. During the housing bubble, the garbage that went into these ratings models was obvious, with "way too optimistic" assumptions that "look way too small compared to history." Statistical models can be very powerful tools, but they can never be the only tool.
So Wall Street and London decided that the lesson they needed to learn was to build better models. It is, not ironically, the same lesson that the Federal Reserve took in its own very limited and very narrow attempt at cursory accountability. Monetary policy, too, is run on statistical models almost exclusively (I write "almost" out of minor deference to some that work there in a less robotic state, but in reality policy decisions are determined regressions).
One of the first major Federal Reserve papers after the panic was a San Francisco Fed effort that I have referred to repeatedly for very different reasons than why it was written. The piece was an actual honest look at the failure of all the various monetary policy models, but then took the tact that such failure was good in that it had given the Fed the opportunity to incorporate new data into the same models and thus making them more robust. That was true but only in the smallest of "qualitative" gains; the models went from wholly unrealistic predictions suggesting only trivial risks (overly optimistic about everything, of course) to wholly unrealistic predictions suggesting slightly less trivial risks.
The primary problem with all of these models, policy setting versions at the Fed and those guiding and calculating investments and risk in the banks, are that they all universally assume monetary policy effective - and not just effective, but effective in a predictable manner. That was the entire point of the "Q" in QE; suggesting to the public that there was mathematical "science" to the whole thing. Monetary experts wielding enormous technical knowledge could precisely recite and then harness forces to defeat and overcome hysteresis - the idea that an economic factor or even the whole economy requires a "push" in order to get it moving. Thus, QE was proclaimed as the precise calculation that would both figure exactly the resting force on the economy keeping it down and then the means and quantity of force (monetary) required to overcome it.
By that description alone, you can appreciate why anywhere a second QE might have been offered completely invalidated all of it - including and especially the models. If you need a second, it is axiomatic that there was no such precision in the first. And if the second is more like a shot in the dark, what might be a third or fourth? How about, as Japan, an eleventh? Just like Deutsche Bank, the Fed, the ECB and Bank of Japan all failed their own qualitative stress tests, but "tests" that were really real world economic experiments where the global economy now lives with the pain of such hubristic nonsense and failure. It is more than a touch hypocritical for the Fed to suggest any some bank might be weak in that regard when that same bank uses the same sorts of assumptions as what causes persistent monetary failure (including 2008).
The same monetary policy that was wholly ineffective in mitigating, let alone forestalling as the "Greenspan put" would have it, the Panic of 2008 and the Great Recession along with it was somehow assumed to be absolved of that affair. That remains the message of Ben Bernanke, who claims heroism in doing a great deal under the cover of burning wreckage that he had himself claimed either impossible or easily controllable via his immense power - "power" of, again, statistical models to predict conditions and then responses to conditions. Crashes always lead to rebounds, yet here was the Fed claiming monetary policy had nothing to do with the former and everything to do with the latter even though the math was written to cover all of it. The banks, or many of them, believed it - and modeled it.
If banks were enamored with these new "emergency" programs like ZIRP and QE, the public was at first somewhat apathetic if suspicious but far more so over time increasingly restive. There are consequences for both, but perhaps more so in the belief than the doubt. What was Deutsche Bank doing in May 2014 with all its new "capital" with which to direct bank resources? As this presentation leaves no doubt, the bank saw great promise in Emerging Market Debt and US High Yield particularly leveraged loans. May 2014.
Investors are starting to worry about more than just profit declines at the biggest banks. They're increasingly concerned about their ability to repay their debt.
A prime example can be found in Deutsche Bank, which isstruggling in the face of falling debt-trading profits and an unclear strategy for increasing income. And investors have yet one more thing to worry about: Souring energy prices pose a risk for substantial losses.
Deutsche Bank may have "significant" energy exposure "that is not investment grade and is not well secured," Amit Goel and Jag Yogarajah, analysts at Exane BNP Paribas, wrote in a note on Wednesday.
That was written just yesterday by Bloomberg's Lisa Abramowicz. What we can reasonably assume in addition is that the bank itself has little idea of potential losses or credit strains, either, having been "qualitatively" weak in figuring any of this out. The stock price was trading at about $50 a share in March 2014, having dropped to about $40 a share on capital and dilution concerns through that summer. Then the stock stabilized around $30-$35 for most of 2015 until August. But for a brief rebound, it has been almost straight down since, now around $16.
Worse, however, the bank's credit default swaps have surged just this year. The general CDS price had spiked starting in April 2015 from about 70 bps to over 100 bps, just as general "dollar" liquidity truly started to wane, warning about what was to come. After trading between 80 bps and 100 bps for the balance of 2015, Deutsche Bank CDS then hit the elevator in January to now trade around 160 bps. The costs of following Ben Bernanke's and Janet Yellen's models and interpretations of those models are just being calculated.
That is true for far more than just Deutsche Bank's self-inflicted mathematics, the lesson is increasingly being realized for the whole global economy (again). Despite all the mouthing of "macroprudential" policy and regulation, the promises to never repeat mistakes of the past, Deutsche Bank is already showing the same kinds of self-reinforcing, procyclicality that monetary policy always seems to produce as a means of expressing just those modeled projections. As I wrote in October when Deutsche first announced its growing unease and darkening projections, those past mistakes were increasingly locking the bank into only desperate struggle:
Annual revenues in CB&S were once €21 billion, but only €13 billion last year and who knows how low it might get by the time this year has mercifully ended. But the bank has other problems related to that CB&S book (especially the derivatives portion) leverage aside from the potentially nasty return environment. The way it is structured (heavy "dollar" presence) actually makes a rising dollar work against the leverage ratio and thus offsets some of the downsizing intent as it is taken.
The more the bank struggles against capital, leverage and all the rest of the math, the more it has to downsize, and the more that downsizing takes place the less efficient and effective it becomes and thus requires still more of all of it - increasingly that will turn to getting out at any cost. I get asked all the time how swap spreads could be negative and negative for so long; this is it. It's not about individual interest rate swap positions or even interest rate swaps themselves; this is the reductionism that is required to reckon having made plain bad assumptions at the worst possible time.
The net result of this toxic stew of bastardized banking is a highly negative return (revenue) environment for CB&S in 2015 beyond all scale of 2013, while its efforts to reduce assets (especially risk weighted calculations) continue to fly against its "dollar" activities. The firm managed to take the worst course possible by thinking the shrinking eurodollar system particularly post-May 2013 was an opportunity to replay lost pre-2007 financialism glory. To do that, the bank kept up its leverage and then went after the junk and EM bond bubbles with enthusiasm.
But as that one bank struggles, it leads to weakness in the chains of liabilities that then cause feedbacks and only further systemic resistances; once it accelerates past a certain point, wholesale revaluations and re-evaluations must be conducted in widespread fashion - commonly known as a cycle, credit or otherwise. To Deutsche, there was opportunity in those bubbles because their math synced well with Bernanke's, then Yellen's; that they still believed, even after having to do repeated episodes, there was indeed "Q" to that QE. Janet Yellen even went so far as to determine something of a counterintuitive ideological nugget from just those circumstances, what I have written as the "Yellen Doctrine." In essence, she recognized that there was "reach for yield" and maybe poor lending standards and covenants in leveraged loans and junk bonds but that such "bubbles" would be harmless if they worked. In other words, if it took financial bubbles as the method for monetary policy to actually create the recovery, the recovery invalidates the bubble and leaves it as just the means to the fruitful end.
If that sounds like rationalization, that is because it is. It can be nothing else as monetary policy is not just supposed to have just been effective but again predictably so. If she was then veering into bubbles as the only way to maintain her belief in that effectiveness, monetary policy was no longer predictable which could only mean rationalization of conditions to belief. Acting out in just such a manner will cause any kind of supposedly rational agent to keep acting irrationally long past the rational point of doing so; like investing massive resources in bubbles at almost exactly their top and all the while thinking shrewdly of it.
Having now done it, how does one undo it? It's no easy task for one large bank, perhaps impossible, but Deutsche is not alone; it can't be. There are trillions in EM debt and leveraged loans beside any number of financial "investments" and resources dedicated to a recovery that never existed in anything but rationalization. The problem now is as common as 2008 - when prices start to reveal those flaws. Credit Suisse also yesterday announced a much larger loss than anticipated, 5.83 billion francs. Of that, 3.8 billion francs were due to writedowns (unspecified) in, you guessed it, its investment bank unit. The bank's CEO even cited "material deterioration" during the fourth quarter, the same time as the FOMC declared the recovery ended in great and wonderful success.
The common flaws of Deutsche Bank and the FOMC are not limited to Deutsche Bank and the FOMC, which is entirely the problem in the latter half of 2015 and the first weeks of 2016. They are just, so far, the most visible. Forget "transitory"; that is long gone now even though ferbus and all the rest of the GARCH and VDO's were so sure that oil prices were temporary, as the math left no doubt. This is the same financial disease that has already infected, chronically, the Chinese banking system as it connects to the eurodollar system. All of it rests on one bad assumption, that these models contained no assumptions.
Math became money increasingly after 1995, but even after all the problems of 2008 it still seems as if math became decision-making, too. The statistics all decided that monetary policy just works, and that would be enough. It wasn't; not even close. The banks bought it, but the real economy is far more than banks and regressions. Having suffered that deficiency these eight long years, now we get to project the same kind of unwinding all over again. Being so qualitatively deficient, though, nobody even knows how to calculate the "unexpected" procyclicality, which explains the palpable increase in nervousness and sudden lack of cockiness from even the most mathematically assured.