Equity v credit derivatives showdown: Difference between revisions

From The Jolly Contrarian
Jump to navigation Jump to search
No edit summary
Tags: Mobile edit Mobile web edit
Replaced content with "{{essay|eqderiv|equity v credit derivatives showdown|{{image|ninja showdown|png|Ninja showdown, yesterday}}}}{{c|Credit Derivatives - Premium}}"
Tag: Replaced
 
(8 intermediate revisions by the same user not shown)
Line 1: Line 1:
{{aai|cdd|{{image|ninja showdown|png|Ninja showdown, yesterday}}}}{{c|Equity Derivatives}}After a long hiatus learning the ways of the {{eqdefs}}, the JC is back in the land of [[credit derivatives]], and has been undertaking remedial ninjery with the sacred texts of the {{cdd}}. You can learn about his travails from [[NiGEL]], in the panel.
{{essay|eqderiv|equity v credit derivatives showdown|{{image|ninja showdown|png|Ninja showdown, yesterday}}}}{{c|Credit Derivatives - Premium}}
 
Here we do the service of comparing, in broad strokes, [[equity derivatives]] with the [[credit derivatives]].
 
==Documentation==
Each has its own definitions booklet — {{eqdefs}} and {{cddefs}} respectively — though their genealogies are different.
 
The {{eqdefs}} were published in 2002 and, while not perfect, do a serviceable enough job at describing what is essentially, and usually, a fairly straightforward product, though they get a bit gummed up about dividends. The product traded is for the most part a [[delta-one]] exposure to shares, share baskets and indices and, while hedging can be fraught in times of dislocation, and hedging costs get passed through to end users, the basic notional value of an equity derivative is not: the market price if a listed share: you can see it printed in 6 point font in the Financial Times every day.
 
The {{cddefs}} are for the connoisseur [[ISDA ninja]]. Their original abstract intellectual purity has long since evaporated,  brutalised repeatedly by savage real-world market dislocation. They are now a fearful, paranoid, jabbering wreck. It is as if the winsome fever dream of some JP Morgan brainboxes, strained through the gusset of the [[First Men]] and wrung through some [[Potts Opinion|QC opinions]] has taken root, allowed to flourish, run wildly out of control, threatened life as we know it and then been mercilessly beaten, bent and twisted by a community of embittered banking regulators, themselves branded by the white-hot iron of civilisation-threatening financial disaster.
 
Indeed, that is pretty much what ''did'' happen.
 
The product emerged in the 1990s, the brainchild of JP Morgan boffins, became highly fashionable, by 2003 had earned its own definitions booklet, and as the [[CDO]] mania of the noughties reached fever pitch, it began to standardise. Legions of chancers, grifters and joiner-inners flooded the market and before you knew it there were all kinds of “exotic” structures, each more convoluted and less plausible than the last. Growth was periodically set back by actual credit events in the market, each it's own life lesson about the multifarious ways in which over-engineered, too-clever-by-half structured products can surprise their Inventors by finding unexpected ways to fail.  The real “come-to-Jesus” moment for credit derivatives was the [[credit crunch]] of 2007 and then 2008’s full blown [[global financial crisis]],  which between them revealed the degree to which nice ideas in theory don’t hold up in the sweaty throes of market panic.  There was a ''lot'' of litigation about misfiring — or allegedly misfiring — credit derivatives.
 
The {{cddefs}} were, consequently, monstrously overhauled in 2014, and at the same time the product standardised yet further, moving away from single name, bilateral, privately negotiated deals and towards cleared, standardised, broad-based index products. There are still some privately negotiated deals but, compared with equity swaps, which are the bedrock of hedge fund equity long/short strategies, not many. More than ten trades a week on a given Reference Entity rates special mention in ISDA’s credit market summary.
 
Practitioners will tell you part of their lack of popularity is the sheer complication of the {{cddefs}}. Unlike the {{eqdefs}}, the 2003 Credit Derivatives Definitions really didn’t work, the move away was propelled by regulator angst and infrastructural imperative, so there was not the option of flat-out ignoring them, as the market did to the ill-fated [[2011 Equity Derivatives Definitions]]. Even though now impenetrable, they are ''still'' finding snafus needing quick fix patches.
 
In any case visiting the credit derivatives now after a few years away, is like visiting a long lost friend now institutionalised for her own good, straight jacketed, and fed cold soup through a straw. Occasionally she still manages to shout something outrageous and upset the common room.
 
==Synthetic investment versus loss insurance==
{{Eqderiv}} are means of gaining exposure — positive or negative— to an instrument without owning it. The basic point of the contract is to replicate exactly the economic features of the [[Underlying|underlier]], minus the physical, reporting and funding aspects of being on the register. You buy or sell an {{eqderiv}} ''instead of'' buying or selling the [[Underlying|underlier]].
 
{{Cderiv}} assume you already own the [[Underlying|underlier]], but want to hedge away a specific embedded tail risk:  namely, that it blows up. While you needn’t own the underlier to buy or sell {{Cderiv}} — the Robin Potts opinion is at great pains to stress that a [[Credit derivative|CDS]] is not a contract of insurance — generally speaking you ''will''.
 
 
==Overview and capital structure==
{{cderiv}} address the market value of public, quoted, usually senior unsecured debt obligations — for this discussion, let’s call them “{{cddprov|Bonds}}” while [[equity derivatives]] address the value of public, quoted common shares.
 
The instruments have very different qualities: Bonds repay principal and return income, equities return capital. Unless something really catastrophic happens, the return on a {{cddprov|Bond}} is predictable — interest and principal — and values will be as affected by prevailing interest rates as by deterioration (or improvement) in the Issuer’s creditworthiness. That one really catastrophic thing is the Issuer’s failure: its ''{{cddprov|Bankruptcy}}''.
 
Because there are literally no expectations about what they will pay — not even a redemption date — equities pogo around, depending on the issuer’s quarterly performance, paranoia, market sentiment, geopolitical currents, internet memes, Reddit, whatever Elon tweets about and whatever the madness of collected action believes to be germane to the business of, well, business.  Equities could quite easily be up 200% or down 70% over a six-month period with neither making a tremendous statement as to the basic creditworthiness of the Issuer.
 
So oddly, while you would expect equities to be about corporate default and credit derivatives to be about relative performance in the absence of default, it is the other way round. Credit derivatives, like Bonds, tend to have a fixed term and a fixed premium and, while you can mutually terminate them at market price, there is no ''right'' to do so in the absence of a catastrophic failure — any exit is negotiated and dependent on prevailing liquidity. Equity swaps tend not to have a tenor (they may ''have'' one, but it doesn’t ''mean'' anything)  and to be terminable on the client side at any time.
 
==Buy side and sell side==
Like all swaps, CDS and equity swaps are optically bilateral and a customer can take a short or long position. But there is still almost always a “[[buy side]]” investor looking to the swap to ''take'' a position, or to hedge an existing exposure, and a “[[sell side]]” swap dealer looking for a commission or premium, who will therefore hedge away the market risk presented by the swap.
 
An investor buys a swap to put on a hedge; a dealer buys a hedge put on a swap. The investor seeks to change its market position with derivatives, the dealer seeks to keep its position flat.
 
This is a deep, profound market feature, and the [[JC]] thinks it gets misunderstood too often, and talked about not nearly enough. [[Regulatory margin|Bilateral margin regulations]], for example, were introduced in ignorance of, or disregard for, this distinction. [[When regulatory margin attacks|They have already contributed to catastrophic loss]].
 
==Vibe==
Bear in mind the characteristic distinctions between equity and debt. Equities have no term, no stated repayment, no defined yield or return. They represent what would be left were all such instruments satisfied in full. By contrast senior debt obligations have all those features: a defined repayment amount, due date, and deterministic yield (“deterministic” in that it can be observed objectively by reference to something other than the performance or business condition of the issuer).
 
It kind of follows that you can’t “protect” or “insure” the return of common equity. Protection is measured as loss against an expected return: equities do not have one. Equities have just a prevailing market value, which can pogo around unpredictably — literally unpredictable — propelled by macroeconomic and geopolitical conditions and, always, the madness of crowds.
 
You can set an ''arbitrary'' return — a hypothetical limit — and measure equity performance against ''that'' but this is not ''protection'' but ''speculation''. It is to buy or sell an {{eqprov|option}}: a true derivative. Equity swaps are in this sense genuinely derivatives, and no one hit up Robin Potts Q.C. for an opinion just to check they weren’t disguised insurance contracts.
 
Because they have all those deterministic features — due principal, due income, due date — the debt obligations ''are'' intrinsically insurable. They are also far less prone to market fluctuation, and their volatility tends to zero as maturity approaches. There is a lot less call for speculation, therefore (not ''none'', to be clear, just a lot less, and it tends to be highly levered) and a lot more call for protection against those deterministic features.
 
Credit events are not concerned with capricious market value fluctuations, therefore, but only portents of utter destruction: indicators that a {{cddprov|Reference Obligation}} really will not pay the ''whole amount'' due on the ''day'' it is due. This is binary:  market indicators that a {{cddprov|Reference Entity}} is merely ''less likely'' to be able pay in full on time — ratings downgrades, [[common equity]] breaching price triggers and so on — should ''not'' trigger a {{cddprov|Credit Event}}. To be sure, they may cause “CDS spreads” to spike — the premium one must pay to buy credit protection on the Reference Entity for new contracts will increase, so the replacement cost of existing trades will have a value — and may thus lead to gains or losses in the mark-to-market value of a [[credit default swap]], but these gains will reflect the present value of that spread differential against the originally traded price, and not anything more profound than that. As long as the {{cddprov|Reference Entity}} does in time repay its debt (or at least not indicate finally that it ''won’t'' before the credit protection expires) then the credit derivative will expire unexercised.
 
Thus {{cddprov|Credit Events}} cross over with, but are different to {{isdaprov|Events of Default}}, and are more relentlessly focussed on non-payment in full and on time: {{cddprov|Bankruptcy}},  {{cddprov|Restructuring}},  {{cddprov|Repudiation/Moratorium}},  {{cddprov|Obligation Acceleration}},  {{cddprov|Governmental Intervention}}.
 
==Subordinated debt==
The {{cddefs}} also deal with subordinated debt which only makes things more complicated, not less. Firstly, subordinated debt has many of the equity characteristics that make it much less debt like and insurable. It pays a lot more of its total return out in interest; over time the principal amount invested becomes of less and less significance. [[Credit Suisse]] [[AT1]]s, paying over 7% p.a., returned something like 45% of their original investment over 5 years, even though they were wiped entirely in April 2023.
 
And speaking of [[AT1s]], since the [[global financial crisis]] bank capital structures have, by regulatory fiat, become a lot more complicated. Most G20 nations have enacted Bank recovery and resolution regimes, and while they’re broadly similar, outside the European Union, they are subtly different. And banks have reacted to them in idiosyncratic ways, too: there are multiple tiers common equity tier 1, alternative tier 1, alternative tier 2, and some old fashioned perpetual subordinated instruments which were crafted with no such fine distinctions in mind. And banks have opted different ways of bailing in: some convert to equity by design; some are written off. So generically providing for subordinated debt in a conmoditised, way is not straight, in the way plain old common equity, or senior unsecured, debt is.
 
This makes determining credit events on these new instruments fraught, and litigationey. Dear old [[Lucky]] is, we dare say, going to be filling the coffers of our learned friends for some years yet.
 
The {{cddprov|Designated Requirement}} — credit derivatives ’ answer to the {{isdaprov|Cross Default}} threshold — is typically a lot lower: if not specified, USD10m equivalent, where under Section 5(a)(iv) you might expect ten or a hundred times that, referenced to a percentage of shareholder equity.
 
This reflects the different intentions of the provisions: one is to preserve one’s existing position in an ongoing trading arrangement should the counterparty be unable to pay what it owes, by allowing one to terminate and ''avoid'' loss; the other is to ''compensate'' for losses already incurred under a reference instrument (yes, yes, it isn’t ''actually'' insurance and there is no requirement for an insurable loss, but still).
 
The former position requires the company itself to be done for; if it isn’t, you still have the prospect of full recovery; the latter position is concerned only with the market value of that specific instrument. If it is canned, rescheduled, or defaulted upon such that the original bargain it promised is no longer to be had, then the CDS pays out even if the Reference Entity remains a going concern and chunters happily along back towards insolvency. (If its CDS is triggered that becomes vanishingly unlikely of course, but if there is one thing the market tells everyone, it is YOU NEVER KNOW.)

Latest revision as of 02:21, 10 August 2023

Equity Derivatives Anatomy™
Ninja showdown, yesterday
Tell me more
Sign up for our newsletter — or just get in touch: for ½ a weekly 🍺 you get to consult JC. Ask about it here.

After a long hiatus learning the ways of the 2002 ISDA Equity Derivatives Definitions, the JC is back in the land of credit derivatives, and has been undertaking remedial ninjery with the sacred texts of the 2014 ISDA Credit Derivatives Definitions. You can learn about his travails from NiGEL, in the panel.

Here we do the service of comparing, in broad strokes, equity derivatives with the credit derivatives.

Nominally just derivatives attaching to a different part of the capital structure, but don't be fooled. Almost everything about them is different. So, with feeling:

Each has its own definitions booklet — 2002 ISDA Equity Derivatives Definitions and 2014 ISDA Credit Derivatives Definitions respectively — though their genealogies are different.

The equity derivative is an age-old family of products that ISDA simply codified: puts, calls, collars — options, basically — and contracts for difference on shares and baskets of shares. The 2002 ISDA Equity Derivatives Definitions, published in 2002, pulled in share indices, but these are really just complicated, large baskets.

While not perfect — they get a bit gummed up about dividends, and tax can get complicated — the equity definitions do a serviceable enough job of describing what is essentially a straightforward product. It is mainly traded as delta-one exposures and, while hedging can be fraught in times of market dislocation (and hedging costs and losses get passed through to end users, the basic notional value of an equity derivative is not: the market price of a listed share: you can see it printed in 6 point font in the Financial Times every morning, and printed on your Bloomie every second of the day.

Credit derivatives specialists who come to the equity derivatives, as droves did after the collapse of the structured credit market in 2008, tend to be surprised and alarmed at how straightforward and unfunky equity derivatives are. They are just not that fiddly. This is not for want of trying — hello hypothetical broker dealer — but you sense much of that is wanton complicationeering from our risk controller friends: in the decade following the financial crisis, the basic synthetic equity product became markedly more fiddly. No small part of that was fugitive structured credit lawyers overengineering through force of habit and ignorance. The ISDA ninja’s refrain: never hesitate to complicate.

Now, to be sure: while the equity swap product is structurally simple, it is no less risky for it, and you can just as easily lose your shirt, as Archegosprime brokers would tell you.

But unlike the dysfunctional credit derivatives definitions, it still mostly worked. As the saying goes, “if it ain’t really broke, don’t fix it”: the market has remained loyal to the 2002 ISDA Equity Derivatives Definitions despite a noble attempt by a squadron of misguided Linklaters complexity bombers to overhaul them in 2011, which failed rather epically. We have quite a bit of fun at the expense of “Flight 19” but we mean well as, undoubtedly, they did too.

But as far as we know, not a single trade has ever been documented under the 2011 Equity Derivatives Definitions. If you happen to know of one, please don’t write in to tell us: that would spoil an impression of outright calamity we are quite happy labouring under.

By contrast the credit derivatives booklet has had quite the Odyssey since the “long-hand” credit default swap product evolved in the late 1990s. In any case the latest version, the 2014 ISDA Credit Derivatives Definitions, is for the true connoisseur of iatrogenic ninjery.

Their original abstract intellectual purity has long since evaporated, brutalised by repeated, savage, real-world market dislocations. They are now a fearful, paranoid, jabbering wreck of gabbling cabbage. It is as if the winsome fever dream of the JP Morgan brainboxes from whose brow they sprang, strained through the gusset of the First Men and then wrung out with some QC opinions has taken root, allowed to flourish, run wildly out of control, threatened life as we know it and then been mercilessly beaten, bent, twisted and hacked at by a community of embittered banking regulators, themselves branded by the white-hot iron of civilisation-threatening financial disaster.

Indeed, that is pretty much what did happen. It is not a pretty sight. Don’t look directly at the definition of Event Determination Date without peril-sensitive sunglasses. It may blind you.

Credit default swaps emerged in the 1990s, the brainchild of JP Morgan boffins. They became highly fashionable and by 2003 had earned their own definitions booklet. As the CDO mania of the noughties crested, the market felt a real need to develop standard terms for a CDS.

Just as it was doing this a cavalcade of chancers, grifters and joiner-inners flooded the market and before you knew it there were all kinds of “exotic” structures, each more convoluted and less plausible than the last. Growth was periodically set back by actual credit events in the market, each teaching its own life lesson about the multifarious ways in which over-engineered, too-clever-by-half structured products can surprise their Inventors and find unexpected ways to fail.

The real “come-to-Jesus” moment for credit derivatives was the credit crunch of 2007 and then 2008’s full blown global financial crisis, which between them revealed the horrifying degree to which nice ideas in theory don’t hold up in the sweaty throes of market panic. There was a lot of litigation about misfiring — or allegedly misfiring — credit derivatives. That has never really changed. The road to hell is, as they say, wallpapered with CDS Confirmations. The credit derivatives have always been a bit of a moving feast.

As a result of their categorical failure to accommodate the various threnodies of the global financial crisis the 2014 ISDA Credit Derivatives Definitions were monstrously overhauled in 2014, while at the same time the product standardised yet further, moving away from single name, bilateral, privately negotiated transactions and towards cleared, standardised, broad-based index products.

There are still some privately negotiated deals but, compared with equity swaps, which are the bedrock of hedge fund equity long/short strategies, not many. More than ten trades a week on a given Reference Entity rates special mention in ISDA’s credit market summary. Equity traders night trade the same name ten times a second.

Practitioners will tell you part of their lack of popularity is the sheer complication of the 2014 ISDA Credit Derivatives Definitions. Unlike the 2002 ISDA Equity Derivatives Definitions, the 2003 Credit Derivatives Definitions really didn’t work, the move away was propelled by regulator angst and infrastructural imperative, so there was not the option of flat-out ignoring them, as the market did to the ill-fated 2011 Equity Derivatives Definitions. Even though now impenetrable, they are still finding snafus needing quick fix patches for the new credit derivatives definitions.

In any case, visiting credit derivatives now after a few years away, is like visiting a long lost friend now institutionalised for her own good, straight jacketed, and fed cold soup through a straw. Occasionally she still manages to shout something outrageous and upset the common room.

Premium content
Here the free bit runs out. Subscribers click 👉 here. New readers sign up 👉 here and, for ½ a weekly 🍺 go full ninja about all these juicy topics👇
  • A long essay taking in:
    • The general “vibe” of CDS versus CFDs
    • Their respective places in an Issuer’s capital structure: the curious fact that credit derivatives, though “senior” are more concerned with bankruptcy
    • As a means to synthetic exposure versus insurance versus loss
    • As a buyside service and sellside hedging tool
    • How they handle — or in the case of CDS, really don’t handle — subordinated instruments
    • How thresholds and notionals relate to each
    • As a special treat, an AI-generated essay about what it takes to learn the ways of the credit ninja.

See also

References