Delivery squeeze

[Arka Bose]

In the event of delivery squeeze, The value of the CDS declines. This is because the price of the underlying bond increases and hence payoff from settlement is lower. So how does this delivery squeeze induce negative CDS-Bond basis? The bond spread should be now lower/cds spread should be higher as value declines ( it says value not price) and hence should it not tend to make the basis more positive?

 

[QuantMan2318]

Hi @arkabose

Gregory has been known to be dubious with the CDS-Bond basis. Look at his error with the Funding cost example, as you know, increase in the Funding cost of a Bond, increases the spread demanded by the Bond holders and hence the Bond spread rises, CDS being unfunded remains the same causing the CDS-Bond basis to be negative (Gregory errs as positive), in the same vein, he says that the delivery squeeze causes an increase in the Bond price that in turn causes the Bond spread to drop and CDS values being suppressed should theoretically have higher spreads resulting in more positive CDS-Bond basis. I am again unsure of the distinction between CDS value and price here, sure enough the market would price the CDS low and hence the spreads would be high?

I don't know what he was thinking to state that it becomes negative, suppose, he is referring to the increase in funding cost as a result of the delivery squeeze?
then the Bond spread would rise, but I am not too sure.

Perhaps @ami44 or @David Harper CFA FRM can throw more light into this?

 

[Arka Bose]

I know, when i was reading that section for the first time, i scratched my head over and over again for the funding cost section until i realized that I am not alone

 

[ami44]

Arkabose,

as you said, the payoff from setttlement will be lower in case of a delivery squeeze. That means for a protection buyer the cds is less valuable, which means he will want to pay less for it. This means the cds spread will be lower with the possibility of a delivery squeeze, than without it.

If a delivery squeeze also affects the cds value for the protection seller is not clear to me. On the one hand the delivered bonds are worth more, but are they able to take advantage of that, since the squeeze will be over after the setzlement period? Maybe somebody here has experience with cds settlement?

 

[QuantMan2318]

Beautiful @ami44 , Thanks that cleared my confusion!

So, what I get is that the CDS spread is not the same concept as Bond spreads, therefore, if I get it correctly, CDS spread is the premium paid by protection buyer to protection seller and since the protection buyer is willing to pay less of this premium, the CDS spread falls; So, the CDS being less valuable in the event of a default squeeze, reduces the CDS spread. One must again assume that this fall is probably more than the reduction in the Bond spread causing the Bond value to increase and therefore, Gregory's conclusion that the basis is negative, right?

EDIT: I was under the impression that the reduction in CDS value increases the spread akin to its effect on Bonds, sorry

 

[ami44]

Beautiful @ami44 , Thanks that cleared my confusion!

So, what I get is that the CDS spread is not the same concept as Bond spreads, therefore, if I get it correctly, CDS spread is the premium paid by protection buyer to protection seller and since the protection buyer is willing to pay less of this premium, the CDS spread falls; So, the CDS being less valuable in the event of a default squeeze, reduces the CDS spread.

Hi Quantman2318,

I agree with you until here.
I'm not sure I understand the following:

One must again assume that this fall is probably more than the reduction in the Bond spread causing the Bond value to increase and therefore, Gregory's conclusion that the basis is negative, right?

Maybe the following is the misunderstanding, i'm not sure:
At the moment of the delivery aqueeze the credit event already happened. At this point nobody will sell credit protection on that bond. So the fact that we have a delivery squeeze has no effect on the cds spread.
What does affect the cds spread is the fact that there is a probability greater than zero, that a squeeze might happen in the future. The protection buyer factors that probability into the price he is willing to pay. that happens at the time of contracting the cds, when the cds spread is fixed, which is before the credit event.

The possibility of a delivery squeeze does not affect the bond spread before that squeeze actually happening. At least thats what I assume, I might be wrong.

Now thinking about it, do you mean, that the possibility of a squeeze increases the value of a bond, because people hope that in the case of a credit event they will loose less money because of the squeeze? You might have a point, I don't know.

 

[Arka Bose]

CDS spread being less valuable reduces the spread. I thought there is a difference between a value of a CDS and a price of a CDS? I was under the impression that price of a cds when low, the cds spread is low and vice versa. I thought value to be same as we treat the bonds?

 

[ami44]

CDS spread being less valuable reduces the spread. I thought there is a difference between a value of a CDS and a price of a CDS? I was under the impression that price of a cds when low, the cds spread is low and vice versa. I thought value to be same as we treat the bonds?

As I see it, it's not the same for cds and bonds:

For both a spread is contracted at issue date. For a cds it is contracted at inception how much the protection buyer has to pay for the protection. For a bond it is defined at issue date how much spread above the riskless interest rate the bond issuer has to pay for the money.

Than things happen and markets move. After some time you can look what spreads are contracted on new cds for the same counterparty and if that spread has risen, than for a protection buyer your cds value becomes positive, because you pay less than market rate, and vica versa.

For a bond holder you can look at the current spread that the bond issuer has to pay on new issued bonds, or you extract a spread from current transaction prices for bonds. If that current market spread is above the one of your bond, than your bond is less valuable than 100, since you receive less interest than current market rate.

Of course these depend on your perspective of being protection seller/buyer and bond holder/issuer. With bonds it seems clear to me what is meant if you talk about the value of a bond. Its from perspective of a bondholder. For a cds this seems less clear to me, but I would guess that the buyer perspective is meant if nothing else is mentioned.

 

[QuantMan2318]

Hi Quantman2318,

Maybe the following is the misunderstanding, i'm not sure:
At the moment of the delivery aqueeze the credit event already happened. At this point nobody will sell credit protection on that bond. So the fact that we have a delivery squeeze has no effect on the cds spread.
What does affect the cds spread is the fact that there is a probability greater than zero, that a squeeze might happen in the future. The protection buyer factors that probability into the price he is willing to pay. that happens at the time of contracting the cds, when the cds spread is fixed, which is before the credit event.

The possibility of a delivery squeeze does not affect the bond spread before that squeeze actually happening. At least thats what I assume, I might be wrong.

Now thinking about it, do you mean, that the possibility of a squeeze increases the value of a bond, because people hope that in the case of a credit event they will loose less money because of the squeeze? You might have a point, I don't know.

Yes, I don't know what the exact mechanics of the market are, but a reading of Gregory seems to concur with your last point above, Gregory states that the possibility of a squeeze like how it reduces the CDS spread and thus its value, also reduces the Bond Spread because the scarcity of supply of the Bond makes it go up in value. These issues probably crop up just before the settlement. Again, I am not sure as to its veracity

 

[ami44]

...
Gregory states that the possibility of a squeeze like how it reduces the CDS spread and thus its value, also reduces the Bond Spread because the scarcity of supply of the Bond makes it go up in value.
...

Quantman2318,

I don't understand the logic of here. Why would a possible future scarcity influence the current bond spread?

I'm not saying that it's not true, I just don't understand the reasoning.
Maybe we can argue like this: the bond spread is the compensation to the bond holder for his risk. He can eliminate this risk by buying a cds. Thus the bond spread should equal the cds spread, except, that the bond spread contains also some more components, like a liquidity premium. That means, if the possibility of a delivery squeeze reduces the value of a cds and with it the cds premium, but does not diminush the usefulnes of the cds as a hedge for the credit risk of the bond, than the bond spread should be reduced by the possible squeeze too. The reason is simply, if a bond holder needs less premium to insure himself against the credit risk, than he will also demand less bond spread as compensation.

Following this line of thought, the basis will not be affected by the possibility of a delivery squeeze, because cds and bond spread will be affected equally. Is it that what you are trying to say the whole time Quantman2318 and Arkabose? I just did not understand.
Of course these are all theoretical considerations, I have no Idea what the markets do in reality.

After some websearch I found this paper, with a totally different idea:
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.483.6430&rep=rep1&type=pdf

If I understand correctly, than the cheapest to deliver option increases the basis. A delivery squeeze amends these effect, because the prices of the cheaper bonds will rise and the differences between the options will be leveled. So in some way the squeeze has a depressing effect on the basis through neutralising the positive effect of the cheapest to deliver option (provided there is one).

How about that?

 

[QuantMan2318]

Quantman2318,

Maybe we can argue like this: the bond spread is the compensation to the bond holder for his risk. He can eliminate this risk by buying a cds. Thus the bond spread should equal the cds spread, except, that the bond spread contains also some more components, like a liquidity premium. That means, if the possibility of a delivery squeeze reduces the value of a cds and with it the cds premium, but does not diminush the usefulnes of the cds as a hedge for the credit risk of the bond, than the bond spread should be reduced by the possible squeeze too. The reason is simply, if a bond holder needs less premium to insure himself against the credit risk, than he will also demand less bond spread as compensation.

Following this line of thought, the basis will not be affected by the possibility of a delivery squeeze, because cds and bond spread will be affected equally. Is it that what you are trying to say the whole time Quantman2318 and Arkabose? I just did not understand.
Of course these are all theoretical considerations, I have no Idea what the markets do in reality.

After some websearch I found this paper, with a totally different idea:
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.483.6430&rep=rep1&type=pdf

If I understand correctly, than the cheapest to deliver option increases the basis. A delivery squeeze amends these effect, because the prices of the cheaper bonds will rise and the differences between the options will be leveled. So in some way the squeeze has a depressing effect on the basis through neutralising the positive effect of the cheapest to deliver option (provided there is one).

How about that?

Dear @ami44
Thanks for the above document. If I get it correctly, you have stated throughout this thread that the possibility of the Delivery Squeeze affects the CDS alone, reducing its value by reducing the spread. It has no effect on the Bond spread, however, this in itself depresses the basis.

The actual delivery squeeze itself has no effect on the CDS spread, however, this has an effect on the Bond spread, ironically, it has been observed that the CTD option makes the value of the Bond to be more valuable because of scarcity of supply and hence has a negative effect on basis as above

No, I didn't mean the same thing as what you meant in the second para, what I meant was, a reading of Gregory (it's his opinion, not mine) seems to imply that basis becomes negative as a result of the delivery squeeze, he states that the scarcity in supply of the Bonds makes the Bond go up in value, he is not clear when, its quite possible that he means the actual event of the squeeze or just prior to settlement as I mentioned above and hence it corresponds closely to the CTD option in your paper; According to that logic, the CTD option on the CDS makes the CDS more valuable and hence increases its premium/spread, however, the huge demand for cash Bonds at the time of default, reduces the value of the option and at the same time increases the price of the Bonds and hence reduces the Bond spread. He also states that the CDS reduces in value as well, though he is not clear as to when; so, what I meant was the combination of these two effects, makes the Basis to become negative as he has stated.

Thanks

 

[Dr. Jayanthi Sankaran]

Hi @arkabose,

In answer to your question let's take the following example:

An investor wants to hedge his exposure to a corporate bond by buying a CDS. He buys a 5-year corporate bond yielding 7% per year for face value. He also buys a 5-year CDS, with the CDS spread = 2% per annum (200 bps). The outstanding notional = $100,000,000

If the bond issuer does not default the investor earns 5% per year since the CDS spread of 2% is netted against the corporate bond yield of 7%. If however, the bond does default, the investor earns 5% up to the time of default. This can be expressed as:

n-year bond yield - yield on the n-year risk-free bond is approximately = n-year CDS spread
7% - 5% = 2%

This shows that the excess of an n-year corporate bond yield over the risk-free rate should be approximately equal to the n-year CDS spread. If it is significantly more than this, the investor can earn more than the risk-free rate by buying the corporate bond and buying the CDS protection. If it is significantly less than this, the investor can borrow at less than the risk-free rate by shorting the corporate bond and selling the CDS protection.

Under physical settlement of a CDS, a delivery squeeze is caused by a lack of deliverable obligations in the market and causes bond prices to inflate due to strong demand.

Going back to our example and defining CDS-bond basis to be:

CDS-bond basis = CDS spread - Excess of bond yield over risk-free rate

as bond price increases, the bond yield decreases causing the CDS spread to be lower. Also, the excess of bond-yield over risk-free rate (credit spread) declines. As a result, there is every chance that the CDS-bond basis is negative. Prior to the 2007 credit crisis, it was on average slightly positive. During the crisis, it tended to be negative and became highly negative for a short period of time in January 2009.

Hope that helps!

 

[Arka Bose]

@DR Sankaran, thanks for taking your time and writting such a detailed explanation.
I am just confused about the usage of the word 'value' in terms of CDS. Since the author used 'value of CDS leg falls' when there is default, I am confused that whether the value here means higher spread or value here means the 'price' of the spread.
After reading your post, I get the tone of what author tries to say. Thanks.

 

[ami44]

Going back to our example and defining CDS-bond basis to be:

CDS-bond basis = CDS spread - Excess of bond yield over risk-free rate

as bond price increases, the bond yield decreases causing the CDS spread to be lower. Also, the excess of bond-yield over risk-free rate (credit spread) declines. As a result, there is every chance that the CDS-bond basis is negative. Prior to the 2007 credit crisis, it was on average slightly positive. During the crisis, it tended to be negative and became highly negative for a short period of time in January 2009.

Hope that helps!

Hallo Jayanthi,

I agree with your great explanation above, except that I'm not clear about the last paragraph. Sorry for asking so stupidly, but do you mean that the possibility of a delivery squeeze increase/decreases the basis or has no effect at all?

Following your line of thought, the possibility of a delivery squeeze should depress the cds spread, which in turn would also decrease the bond spread to avoid arbitrage opportunities. Which means the basis will not be affected.
I might be wrong though.

 

[Dr. Jayanthi Sankaran]

Hi @ami44,

You are right on spot. Yes, in general the excess of the bond yield over the risk-free rate will be approximately equal to the CDS spread. In which case, due to your arbitrage argument and mine (see above), the CDS-Bond basis would be equal to zero.

Scenarios where the CDS-bond basis is negative or positive could probably be due to the fact that the bond yield spread is calculated using the LIBOR/swap rate as the risk-free rate. Usually the bond yield spread is set equal to the asset swap spread.

I need to let you know that your question is not stupid at all. This problem threw me off and it took me a while to understand what was going on....I find your answers to be very intuitive!

Hope that helps!

 

[ami44]

Dear @ami44
Thanks for the above document. If I get it correctly, you have stated throughout this thread that the possibility of the Delivery Squeeze affects the CDS alone, reducing its value by reducing the spread. It has no effect on the Bond spread, however, this in itself depresses the basis.

Actually I changed my opinion in the course of the thread (learning is great, isn't it). Initially I thought the possibility of the delivery squeeze only affects the cds spread. Now I think, and Jayanthie just confirmed it above, that the bond spread is affected as well, so that the effects on the basis neutralize each other. Otherwise we would create an arbitrage opportunity.

But that said, their seems to be a different effect on the basis, if a cheapest to deliver option exists. The positive effect of this option on the basis will be partly neutralized by the possibility of a squeeze. So you can say, that under certain circumstances, the possibility of a delivery squeeze has a depressing effect on the basis.

The actual delivery squeeze itself has no effect on the CDS spread, however, this has an effect on the Bond spread,

I think it's always about the future possibility of the squeeze. After the credit event happened, the terms 'basis' and 'current cds spread' are meaningless.
I do not have Gregorys text available at the moment, but I strongly suspect, that he only talks about bond and cds spreads before a credit event happened.

 

[Dr. Jayanthi Sankaran]

Hi @ami44,

Just adding a little more to the CDS-bond basis = CDS spread - Asset swap spread

An asset swap exchanges the coupon of a corporate bond for LIBOR plus a spread. Reproducing an example from Hull:

Let's take a situation where the asset swap spread on a particular bond = 150 bps. There are three likely scenarios:

1) Bond Price = Par value of $100. The swap involves Company A paying coupons on the bond and Company B paying LIBOR +150 bps
2)Bond Price < Par value, say $95. The swap is structured such that, in addition to the coupon Company A pays $5 per $100 of notional principal at the outset. Company B pays LIBOR + 150 bps
3) Bond Price > Par value, say $108. The swap is then structured so that, in addition to LIBOR + 150 bps, Company B pays $8 per $100 of principal at the outset. Company A pays the coupons

The effect of all this is that the PV of the asset swap spread = Price of risk-free bond (risk free rate = LIBOR/swap rate) - Price of Corporate Bond.

I do not know if you can say:

PV of CDS-Bond basis = PV of CDS spread - PV of Asset Swap spread

Jayanthi

 

[[email protected]]

All, this is an excellent thread (thanks!)

Delphi looks like an interesting case (2009) around short squeeze

https://ftalphaville.ft.com/2012/01/05/779501/why-do-they-exist/

I understand CTD to give a benefit to the CDS holder (insured) in that on default s/he can deliver a variety of bonds, some of which will be cheaper for various reasons. For me this helps explain the additional CDS spread (i.e. positive basis)

The delivery squeeze confuses me more intuitively.

A shortage of deliverable bonds suggest bond prices rise (reducing the yield) and therefore a positive basis effect. The notes say this is negative overall. Conceptually, perhaps the CDS spread reduces due to the increased recovery value on the more valuable bond?

I am also struggling to understand the negative basis caused by accrued interest at default. This is accrued in the CDS price/ spread. If it were removed, the spread would be lower (suggesting negative basis), but it is not.

 

[QuantMan2318]

All, this is an excellent thread (thanks!)

The delivery squeeze confuses me more intuitively.

A shortage of deliverable bonds suggest bond prices rise (reducing the yield) and therefore a positive basis effect. The notes say this is negative overall. Conceptually, perhaps the CDS spread reduces due to the increased recovery value on the more valuable bond?

I am also struggling to understand the negative basis caused by accrued interest at default. This is accrued in the CDS price/ spread. If it were removed, the spread would be lower (suggesting negative basis), but it is not.

Hi @[email protected] , Thanks for bringing this thread up again, yes, it was one of the best discussions that we had in these forums and brings back fond memories, in fact I had been wanting to answer your question a bit earlier, but paucity of time stopped me from doing so

A shortage of deliverable bonds would increase the price of the Bond and reduce the Bond spread, theoretically this factor viewed in isolation would make the basis positive, however, when there is increase in price of the bond, the funding costs to finance the purchase of the Bond increases (which lowers the CDS spread) as well as the CDS premium by itself would fall contributing to a negative basis (For a stronger Bond)

When there is an accrued interest at default (which is not paid on a Bond), the Bond spread would increase for the increased riskiness of not being able to collect the AI, meanwhile the CDS has the option of paying the accrued premium which contributes to the CDS spread remaining constant which leads to the negative basis

Hope this helps

Thanks