“Too much information” is an (aging) pop culture reference to an uncomfortable situation.
It’s also a good way to describe the prevailing macro and geopolitical environment.
Each and every day, traders (and market participants more generally) are bombarded with so many ostensibly consequential headlines that paralysis ensues. Eventually, apathy kicks in – the longer-term outlook becomes so impossibly uncertain that everyone submits to a kind of fatalism.
That’s the broad context for a characteristically brilliant year-ahead piece from Deutsche Bank’s Aleksandar Kocic, whose outlook for the vol. market describes low volatility levels, compressed vol risk premia, and flat vol forwards as “a consequence of our reaction to the underlying informational overload that eclipses our capacity for statistical approximation and affects the temporal regime of our decision making”.
Anyone familiar with Kocic’s work knows his notes are beyond compare, both in terms of the depth of analysis and especially the craftsmanship. His 2020 outlook is no exception.
A half-dozen pages in, he documents two decades of vol flows, noting that in order to understand “the current predicament of rates vol” (i.e., levels and the shape of the vol surface), and to determine what the future holds, one needs to appreciate history.
He breaks things up into two regimes, illustrating the dynamics with a pair of schematic representations in which “each ‘orbit’ corresponds to a different size relative to the largest bubble”. Here’s his quick explainer:
For example Pension fund flows accounted to about 20% of the spec flows, which were about the same magnitude as the GSE flows etc. Buyers reside in the upper hemisphere and sellers in the lower. They are separated by the equator. Dealers are in the center of all flows
Prior to the crisis, the vol market was demand driven, he notes, describing it as “a period of unprecedented systematization of mortgage hedging practices and concentration of negative mortgage convexity in a relatively small number of portfolios, as well as an aggressive growth of the hedge funds community”.
During that regime, negative convexity hedging transmitted the optionality of homeowners to the market. “The market was negatively convex with convexity of mortgages distributed across different strikes over a wide range of rates”, Kocic writes, adding that “there was always some hedging to do, which kept realized vol elevated and volatility risk premia well supported”.
In the second regime (i.e., post-crisis) mortgage-related demand disappears and the Fed comes off the sidelines, taking an active role as convexity supplier/manager.
The new era is, Kocic says, “defined by the post-2008 regulatory changes, nationalization of mortgage negative convexity, reduction of risk appetite, and recentering of supply and demand”.
The “topology” is the same, he writes, but the “players and their role has changed”.
Thanks in part to regulatory changes, the market now functions in “insurance mode”, something Kocic has talked at length about before, notably back in March. You can read more on that in “The Gamma See-Saw: Deutsche’s Kocic Explains This Week’s Biggest Market Story“.
Post-2008, demand has been re-centered. As shown in the second visual, money managers and hedge funds are the biggest participants. Here is Kocic’s bullet-point breakdown of how the market evolved:
- 2000 – 2003 Mortgage convexity hedging dominates; high consolidation of hedging practices and high concentration of negative convexity.
- 2004 – 2007 Convexity hedgers encounter regulatory difficulties; most of mortgage convexity resides on bank portfolios and is not hedged (transmission mechanisms are severed); unprecedented growth of hedge funds as core sellers of volatility; compression of risk premia, yield grabbing and carry trade.
- 2008 – 2011 Financial crisis is in full swing with massive short covering of convexity amidst unprecedented spread widening and dry out in vol supply
- 2012 – 2015 Return of supply; emergence of the Fed as the main supplier of convexity; compression of rates range and, low realized volatility; dominance of the carry trade; beginning of Formosa supply.
- > 2015 Complacency reaches new highs; Fed exit and management of convexity through their communication with the market; Formosa supply in full swing and yield enhancement continue to push vol lower
Underscoring and otherwise summarizing the setup at the heart of many of the dynamics we touch on in these pages each and every week, Kocic writes that “money managers [have] emerged as major players — buyers of vega and sellers of gamma as yield enhancement catalyzed by implicit convexity supply of Central Banks” while hedge funds have “shifted from structural to cautious sellers”.
Some specs, Kocic notes, have stayed “true to their pre-2008 character” but a “growing number are now competing with the dealers” whose capacity and willingness to hold assets is circumscribed by a more onerous regulatory environment. “Dealers are pushed into a long vol position which further reinforces low volatility and encourages gamma selling”, Kocic goes on to say.
Next, he recaps a broader discussion around the redistribution of leverage. In short, households and the financial sector have deleveraged, while the public sector and the non-financial corporate sector took the baton. You can read more on that in “There Are No More High-Speed Collisions, Only ‘Parking Accidents’“, but the bottom line is that whereas previously, depressed volatility would have been viewed as an ominous sign, portending (potentially imminent) disaster, now, it’s assumed to be the sustainable status quo.
“Volatility and curve are now causally trapped by each other”, Kocic remarked on Tuesday evening. “For volatility to return in a meaningful way, the right side of the (rates) distribution needs to open up and, for (long) rates to be liberated, risk (and volatility) has to be brought back”.
There are a couple of different avenues by which the curve could bear steepen, opening the door to higher vol. One is simply for the regulatory environment to become less restrictive, liberating banks’ balance sheets.
The other is fiscal stimulus. So far, supply-side measures (e.g., the tax cuts) haven’t produced the desired results, thanks at least in part to the Ricardian-equivalent nature of the Trump administration’s deficit-ballooning stimulus push (See: Trump May Have A Ricardian Equivalence Problem).
As for the types of policies which could bring about a real bear-steepening, suffice to say political opposition to a truly redistributive agenda remains entrenched, both inside the Beltway and among the electorate, large swaths of which have been conditioned to believe that “socialism” everywhere and always equals “Venezuela”.
And yet, through it all, long-term risks certainly haven’t disappeared. Indeed, the future looks more indeterminate than ever. Geopolitical risks have proliferated and it’s hard enough to get a read on what’s coming next week, let alone next year or five years hence.
But as we exit the “second decade” (a reference to the visuals above) and enter the third, volatility levels are low and risk premia compressed, with seemingly no path to escape velocity barring deregulation, the adoption of a progressive agenda in D.C. (which could aggressively bear steepen the curve) or central banks all at once abandoning efforts to keep volatility suppressed in the face of intractable political entropy.
Although the persistence of low vol can appear strange, Kocic writes that “such state of affairs reflects less the distribution of long-term risks than indifference to what lies ahead”. He continues:
Given the underlying political entropy and reality contortions, which it continues to produce, we are forced to abandon as inadequate the tools and frameworks that used to provide insight and access beyond immediate future and, in the absence of their replacement, confine our efforts to short-term horizons.
And then, prior to outlining trade recommendations, Kocic delivers two paragraphs that serve as a reminder of why we continually describe him as peerless. We’re going to present them without further editorializing and without attempting to paraphrase, because, as you’ll see, they are perfect on their own (we should note that the excerpts presented in this post, including those below, represent but a fraction of Kocic’s note, which, in its entirety, spans more than a dozen pages and delves into myriad technical discussions and tradable dynamics that are reserved for the bank’s clients). To wit, from Aleks:
Perspective is not a static concept, it is a function of condition under which it is formed. In skydiving, eyeballing consists in visually assessing our distance from the ground during the fall. We evaluate our altitude and work out the exact moment we need to open our parachute based on a dynamic visual impression. At 2000 meters, immediately after the jump, we cannot see the ground approaching. But when we reach the 800 to 600 meter mark, we start to see it “coming”. The sensation is very different than when observing the ground from a “static” perspective, e.g. from the plane at the same (600m) altitude. Speed affects our perspective: As the ground “rushes” towards us, the apparent diameter of objects increases faster and faster than our distance from the ground shrinks, and we suddenly have the feeling we are not seeing them closer but seeing them move apart suddenly, as though the ground were splitting open.
Perspective is dependent on speed — it is a function of acceleration. When extreme events begin to saturate the info-sphere on daily basis (sometimes even intraday), reality unfolds too fast – we no longer remember (or don’t care about) the headlines from two or three weeks ago. Our perspective and assessment of the horizon (“cognitive eyeballing”) is distorted. We are blinkered by intensity of information we have to process – overwhelmed by both its quantity and speed of its arrival – and no longer seem to be unable to properly assess the risks ahead of us, and have become insensitive to them. The distorted perspective downplays the risk of “hitting the ground”. The informational intensity transcends our capacity for statistical approximation and this rarefication of control affects the temporal regime of our decision making. As a consequence, our horizons flatten and everything that resides beyond immediate future is bundled as “long-term” — we appear to be indifferent to its temporal distance — it is all equally remote and equally out of grasp and we capitulate on our efforts to forecast beyond short term horizons. In this environment, the term structure of volatility becomes congruent with flatness of horizons and the shape of the surface becomes an expression of the disorienting condition based on a confusion of the actual and the virtual.