CDS and CDS Index long or short

[Johnny Firpo]

Hi,
I just got confused concerning "to be long or short a CDS".

Is this correct?
"Long CDS" means to be the protection buyer (which is equivalent to be short the bond)
"Short CDS" means to be the protection seller (which is equivalent to be long the bond)

Is the same true for a CDS Index?
"Long CDS Index" means to be the protection buyer (which is equivalent to be short the bonds)
"Short CDS Index" means to be the protection seller (which is equivalent to be long the bonds)

Thanks in advance!

 

[David Harper CFA FRM]

Good question. I do think it is easier to use credit protection "buyer" and credit protection "seller" to avoid confusion. That said:

• Clearly, you have CDS exactly correct (or several FRM authors, including some notable names, have been incorrect for years!); e.g., the long CDS is "buying credit protection" and this long CDS is effectively short the credit risk of the bond (and, notice, effectively selling credit risk)
  
• As I understand, CDS Indices are the opposite: in an index, the long is buying credit risk (i.e., selling credit protection) and the short CDS Index position is selling credit risk (i.e., buying protection) and this index short is equivalent to being short the bonds

Now, current issue R74 (JP Morgan Chase Whale) did confuse me, fwiw. I do think they (the subcommittee writing about the whale trade) has it correct for CDS Index (annotation mine):

Investing in a credit index, whose value reflects multiple credit instruments, can be analogized to investing in a portfolio of bonds or loans. The short buyer [dh: better is "short" or "seller." Delete buyer here] of a credit index, as with a credit default swap [dh: disagree], typically makes an upfront payment reflecting the value of the index and then makes fixed periodic payments to the long party over a specified timeframe. Those periodic payments are, again, typically referred to as premiums, coupon payments, or credit spreads. When the instrument matures or expires, or a trade otherwise closes, the short party may be required to make a final payment reflecting the change in the value of the instrument. On the other hand, if a credit event takes place during the covered time period, it triggers a typically substantial payout by the long party to the short party. After the credit event, the defaulting credit instrument is effectively eliminated from the index.."

note: I think the term "short buyer" is awkward and should be "seller of a credit index" or "short the credit index"

However, I think they are, at best confusing, or just wrong about the CDS:

Traders often analogize credit default swaps to insurance contracts. The long party is essentially selling insurance, or “credit protection,” against the occurrence of a negative credit event, while the short party is essentially buying that insurance or credit protection. To buy the credit protection, the short party typically makes a payment upfront and then additional periodic payments to the long party, analogous to insurance premiums. Those periodic payments are sometimes referred to as “premiums,” “coupon” payments, or the “credit spread.” In exchange for receiving those payments, the seller, that is, the long party, is obligated, if a credit event like a default takes place during the covered period, to make the buyer, that is, the short party, whole.

The CDS terminology is pretty well established, at least in the FRM (Meissner, Choudhry). When the subcommittee report here awkwardly says "long party" they are referring to the short CDS who is synthetically long the underyling reference (bond). I admit I'm less certain about the CDS Index, but I'll come back with a reference later if nobody else supports me on the index. Thanks!

 

[Johnny Firpo]

Thank you for the prompt answer David. The reading you are referring to was the issue that confused me.
Looking forward to your reference concerning the CDS index!

 

[David Harper CFA FRM]

Hi @Johnny Firpo

I think we are correct, here are two reliable sources, in reference to CDS Index (I am so confident about single name CDS that I am not going to bother referencing it):
1. From http://www.amazon.com/Modelling-Single-name-Multi-name-Credit-Derivatives/dp/0470519282

2. From Markit, a company that would know, at https://www.markit.com/news/Credit Indices Primer.pdf (probably not the most recent ....)

"Section 3--Trading Credit Indices. How do I trade? Buying and selling the indices can be compared to buying and selling portfolios of loans or bonds. A buyer takes on the credit exposure to the loans or bonds, and is exposed to defaults, similar to buying a cash portfolio (buying the index is equivalent to selling protection). By selling the index, the exposure is passed on to another party. Exposure is similar in both cases.

The indices trade at a fixed coupon, which is paid quarterly (except for EM which is semi-annual) by the buyer of protection on the index, i.e. a short index position, and upfront payments are made at initiation and close of the trade to reflect the change in price. Correspondingly, the protection seller, or buyer of the index, receives the coupon. The indices are quoted on a clean basis."
-- Markit Credit Indices, A Primer

 

[Johnny Firpo]

Thank you David! Very much appreciated!

That's all kind of weird...all these guys talking about being long or short anything without knowing what this means and trading or investigating billions of dollars!? Scary!!

 

[emilioalzamora1]

Hi David,

I would like to raise this topic once more for two reasons.

1.) I just got through Fabozzi's book 'Collateralised Debt Obligations' and perhaps unsurprisingly (to your above comment about how many authors got it wrong) he says (page 221):

'the protection seller under a CDS is LONG the credit and LONG the CDS.....while the protection buyer is SHORT the credit and SHORT the CDS' (with credit he means the 'reference asset' to be in line with Culp and your slides).

Usually we should/can trust what Fabozzi says, right?

2.) In the view of credit protection buyer and credit protection seller in a CDS:

The CAIA gives an example (not sure if this 100% correct) in a sense that the credit protection seller protects against credit risk in the following way - is this the right explanation?

Protection seller: 1.) LONG PUT on the bond price 2.) LONG CALL on the spread

- The LONG CALL makes sense, as the protection seller is hoping that the spread he receives will increase over the time of the CDS contract
- However, I don't understand the LONG PUT: usually the protection seller is hoping that the reference assets 'do well' in order to keep his 'contingent payoff' (using Culp's words) low. In the case of a LONG PUT he would hope that the reference assets will decline (default).

(And perhaps important as well: is the mechanics with CALL/PUT on the reference asset the same for a 1.) CDS and 2. Equity Default Swap?)

Can you please give some input here?

Thank you!

 

[Arnaudc]

Good morning Emilio,
I am not David but while we wait for him to reply, I will try to shed some light (it's only my 2 pence)
The protection seller (who sold a CDS) is exposed to the risk that:
i) the the reference entity default. This is covered by the put on the debt. If the default occurs, the option will greatly appreciate
ii) The spread on CDS for this entity increase which means a paper loss on its existing CDS contract (Spread is fixed and agreed at contract initiation from my understanding). So the Call on spread makes sure that the paper loss on sold CDS is covered by appreciation of your call on the spead.

I hope it is clear?... (and correct, if someone could confirm that would be great )

 

[David Harper CFA FRM]

Hi @emilioalzamora1

I found the Fabozzi text to which you refer. But I don't think i've ever seen it described this way ...

Collateralized Debt Obligations Structures and Analysis, Second Edition, Chapter 11 (Introduction to Credit Default Swaps and Synthetic CDOs) [I don't agree with "seller of CDS" here]

"The protection seller under a CDS is said to be the seller of the CDS, but he is also long the CDS and long the underlying credit. The logic behind selling protection and being long the CDS is that the protection seller is in the same credit position as someone who owns, or is long, a bond. Both investors root against default. Of course, there is a major difference between being long a bond and long a CDS. To be long a bond, one must pay its market price. To be long a CDS, one must simply promise to pay future credit event losses."

I've always used and assumed the terminology explained, for example, by Choudhry ...
Here is from The Credit Default Swap Basis By Moorad Choudhry, Chapter 1 (emphasis mine):

"The credit default swap enables one party to transfer its credit risk exposure to another party. Banks may use credit default swaps to trade sovereign and corporate credit spreads without trading the actual assets themselves; for example, someone who has gone long a default swap (the protection buyer) will gain if the reference asset obligor suffers a rating downgrade or defaults, and can sell the default swap at a profit if he can find a buyer counterparty. [see footnote 5] This is because the cost of protection on the reference asset will have increased as a result of the credit event. The original buyer of the default swap need never have owned a bond issued by the reference asset obligor.

----------
[footnote] 5. Be careful with terminology here. To “go long” of an instrument generally is to purchase it. In the cash market, going long the bond means one is buying the bond and so receiving coupon; the buyer has therefore taken on credit risk exposure to the issuer. In a credit default swap, going long is to buy the swap, but the buyer is purchasing protection and therefore paying premium; the buyer has no credit exposure on the name, and has in effect “gone short” on the reference name (the equivalent of shorting a bond in the cash market and paying coupon). So buying a credit default swap is frequently referred to in the market as “shorting” the reference entity"

Re the "protection seller protecting against credit risk"

• As @Arnaudc writes (and I agree), a CDS protection seller (i.e., the counterparty who writes a CDS, collects CDS premium and is exposed to a contingent payoff in the case of a credit event) might hedge their (synthetic) long exposure with a long put on the same bond; a loss on the CDS will be offset by a gain on the put. In brief, with respect to the underlying credit, the written CDS is long the credit such that a hedge would include a long put (or to lesser extent a short call) on the credit's price.
• Re credit spread option: To similarly hedge their (synthetic) long exposure due to a written CDS, the counterparty would purchase (go long) a credit spread put, not a credit spread call. The credit spread put payoff is given by: duration * notional * max(spread - strike,0) such that that the spread put profits on a spread widening (ie, credit deterioration). In short, directionally, a credit spread put is similar to a long put on the bond's price as both are short the credit (a credit spread call is directionally similar to a long call on the bond's price).

I hope that's helpful!

 

[berrymucho]

Hello,

I was reading some older posts to gain clarity on this question, in particular in regard to "CLN investors being short the CDS" (and vice versa, "CLN issuer being long the CDS"), reference:
https://forum.bionicturtle.com/threads/cds-vs-trs-vs-cln.4796/

Now I found this question 608.3, derived from the Crouhy reading (GARP Chapter 18), bottom:

https://forum.bionicturtle.com/thre...tives-1td-cds-trs-cln-crouhy.9312/#post-40084

where I believe option (c) is the answer (max downside is the initial $20MM investment by the investors). However, option (d) "investors are effectively long a credit default swap written by the bank", assumed correct, provides the opposite view in comparison to the above. The investors would be long the credit (risk), or credit default risk, but short the CDS since they sold (and funded) the protection.
Note that option (d) is a direct quote from Crouhy, who might be using the "wrong" definition.

Am I correct with the disconnect or am I missing something? Thanks!

 

[David Harper CFA FRM]

Hi @berrymucho That's such a great observation/comparison, thank you! My thoughts:

• From an FRM perspective, I do want to defend the longstanding convention--consistent with my Choudhry quote above--that long the CDS is equivalent to buying credit protection (or selling the credit risk). At the same time, in most cases, this language can be easily avoided in favor of "purchase (or sell) credit protection" or, less useful in my opinion "sell (or purchase) credit risk." Notice Crouhy himself mostly avoids the long/short CDS terminology in Essentials of Risk Management (2nd Ed, Chapter 12, emphasis mine):
  

  "Credit Default Swaps: Credit default swaps can be thought of as insurance against the default of some underlying instrument or as a put option on the underlying instrument. In a typical CDS, as shown in Figure 12-3, the party selling the credit risk (or the “protection buyer”) makes periodic payments to the “protection seller” of a negotiated number of basis points times the notional amount of the underlying bond or loan. The party buying the credit risk (or the protection seller) makes no payment unless the issuer of the underlying bond or loan defaults or there is an equivalent credit event. Under these circumstances, the protection seller pays the protection buyer a default payment equal to the notional amount minus a prespecified recovery factor."

• With respect to the generic credit linked note (CLN), your link to my Christopher Culp reference also remains consistently true (from an FRM persepetive). We've always treated the investors in a CLN as economically similar to the credit protection sellers (aka, short CDS or long the underlying reference) except they fund the credit risk. In brief, as the CDS is unfunded, the CLN's funding (which virtually eliminates counterparty risk for the credit protection buyer) is the key difference: a CLN is like a funded CDS.
• Question 608.3 (copied below for reference) models Crouhy and you are correct that I parroted Crouhy why he writes "In this structure there are no margin calls, and the maximum downside for the investor is the initial investment of [$20 million.; in my question]. If the fall in the value of the loan portfolio is greater than [$20 million], then the investor defaults and the bank absorbs any additional loss beyond that limit. For the investor, this is the equivalent of being long a credit default swap written by the bank." However, I believe what he means is that the investors are essentially long a senior (basket or portfolio) CDS. My example is analogous to a portfolio with senior 100 XS $20 and subordinate 20 XS 0, where I am using the convention [width of layer] XS [attachment point of layer]. Analogy-wise, the investors here are short a CDS on the subordinate $20 and--I interpret Crouhy to mean--long a CDS on the senior $100.
• Given that, I have tasked myself to re-write 608.3.d because it is too simple for the situation. I hope that's interesting. Thank you!

608.3. Mainway Bancorp arranges a credit-lined note (CLN) with the following structure (source question is here):

• The bank buys $120.0 million of non-investment-grade loans that are assumed to yield Libor + 400 bps, and transfers them to a trust
• The trust issues an asset-backed credit-linked note (CLN) to investors (aka, CLN buyers) for $20.0 million, investing those proceeds into U.S. government securities with a 1.0% coupon
• The bank finances the $120.0 million loans at Libor, and receives from the trust Libor + 100 bps; i.e., net cash flow to the bank is 100 bps which is compensation to cover its default risk beyond the $20.0 million
• The investors receive 1.0% on the collateral of $20.0 million plus 3.0% (300 bps) on the notional amount of $120.00, in addition to any change in value of the loan portfolio

 

[Arka Bose]

I am thinking Crouch's example like this:
The tranche is divided into Senior of $100 and subordinate of $20.
Now, sine the subordinate has to first absorb the losses, they are long on the $20 tranche which i term as subordinate. Thus, essentially, they are short CDS for this $20.
The bank has to absorb the rest, thus making them long the $100 tranche. This is economically similar to writing a CDS by the bank that Crouchy referred above.
IMO, this also explains why high amount of seniority means high counterparty credit risk (since the bank is economically a protection seller in the senior tranche)

Plz let me know whether I am getting it correct or not?

 

[David Harper CFA FRM]

Hi @arkabose Yes, thank you, that looks perfectly captured to me (and this is confusing); thank you for clarifying my mistaken reversal (subordinate <--> senior) from yesterday. Because you basically described my 608.3, I will confirm using Crouhy' example (picture below). Here is how I see it:

• Notional of $105.0 mm parses into subordinated $15, which is exposed to the first losses, and the senior $90, which only absorbs losses above the $15
• With respect to the subordinate $15 XS 0 [i.e., layer width XS attachment], the investors are long the layer (tranche) which is to say they buy credit risk or sell credit protection and therefore are in a position similar to short CDS on this subordinate layer
• With respect to the senior $90 XS 15, the investors are effectively short the layer which is why Crouhy writes "this is the equivalent of being long a credit default swap [i.e., on the senior 95 layer] written by the bank."
• And I agree the investor has counterparty risk. Unlike an investors who sells a naked CDS (where the CDS buyer incurs virtually all of the counterparty risk), the key difference in the CLN is the funding: by funding the CLN, the Investor incurs most of the counterparty risk. Thanks!

 

[berrymucho]

David, arkabose,

Thanks for the insight (and for responding on weekends), that makes much better sense now. I realize that in the Crouhy reading:

1- "In this structure there are no margin calls, and the maximum downside for the investor is the initial investment of [$20 million.; in my question].
2- If the fall in the value of the loan portfolio is greater than [$20 million], then the investor defaults and the bank absorbs any additional loss beyond that limit.
3 - For the investor, this is the equivalent of being long a credit default swap written by the bank."

my mind was set on the "CDS embedded in the CLN" (=subordinate piece, funded by the investor) so I was trying to tie sentence 3 to sentence 1, seeing one CDS in isolation without considering a greater structure. Instead, sentence 3 refers directly to sentence 2, but from a senior tranche (and so, structured CDS) perspective.

I'm glad I asked. Sentence 3 had been bugging me for a while. David, your re-worked question 608.3 turns out to be a great illustration now of CLNs, single-name CDS and basket CDS at the same time! (I assume you meant $105 instead of $95 in the text, the figure is correct). Thanks!

 

[David Harper CFA FRM]

Hi @berrymucho Yes! Thank you ... and I needed to fix my text above (again). I hope i have it right finally In regard to Crouhy's Figure 12-7, as the total notional is $105.0 mm with investors buying a CLN for $15.0, the structure is:

• Subordinate 15 XS 0; Investors exposes to first (ie, attaching at zero) $15 mm loss
• Senior 90 XS 15; bank absorbs any loss above the $15 mm "attachment point" so effectively per Crouhy has written (sold) a CDS for this layer. Thank you for your help!!

 

[Arka Bose]

@David Harper CFA FRM , no probs! With the amount of hard work and support you are giving to this forum at this moment, little mistakes are not unexpected!