CHAT ON SECURITIZATION OF NON-PERFORMING LOANS

(SPECIAL REFERENCE TO KAMCO TRANSACTION)

Anchor persons: Amit Agarwala, Fitch IBCA, Hong Kong and Don Tang, Deutsche Bank Securities, Hong Kong

guest_Amit:
I think we can start now. The topic as Vinod has mentioned is on NPLs with focus on Kamco transaction. I hope the participants have an idea of the Kamco transaction . Any questions or discussion on this transaction are welcome. I suppose Don can handle the transaction structure issues giving the arranger perspective while I can focus on the rating analysis issues. Thanks

guest_bama:
We have been following with keen interest the KAMCO transaction. In particular, it would be interesting to understand the investor response to the transaction. 

guest_Don Tang:
The investors response has been great. Finally, most of the investors are from Asia and US.

guest_Don Tang:
I think the main reason is due to i) good yield and ii) the investors are very comfortable with KDB risk.

guest_bama:
Could you give us a feel for the final economics of the transaction? We understand that the average cost of funds for KDB is LIBOR plus 90 while the deal was priced at LIBOR 200. 

guest_Don Tang:
I think the economics work out fine for everyone. There is a premium for the structuring risk.

vinodk:
Amit, would you like to elaborate how is the Kamco deal different from other non-performing deals. Is it only because the exposure is mainly on the credit of KDC?

guest_Amit:
Sure. While Fitch has worked on a number of "true" NPL deals across US, Italy and Japan, this is not actually a true NPL transaction. the reason being the focus of analysis was not on underlying assets but on put backs or recourse to Korean banks

guest_Amit:
only marginal benefit was given to direct payment from the loans and main source of payment was assumed to be the put option payments. 

I agree. What I am pointing out is that unlike other transaction where the asset analysis was important and cash flows were assumed from the NPLs with no recourse, in this transaction such analysis was not carried out.

guest_bama:
Why was the two-SPV structure adopted? 

guest_Don Tang:
The two-SPV structure is mainly due to the usage of the ABS law and get away from the Security Bond Trust Act in Korea.

vinodk:
Don, I read a statement of K V Prabhakar that the response was only due to credit of KDC. Do you agree? My question to both Amit and Don: the originating banks continue to have full recourse due to the put option; then what is the role of KDC in the transaction?

guest_Amit:
KDB is one of the put option banks (60%) and also provides credit facility on top of it.

guest_bama:
In India, sale of loans and transfer of instruments backed by mortgages attract a prohibitive duties. What is the scene in Korea? 

vinodk:
To think of it, the originating banks continue to carry full recourse on the loans, right?

guest_Don Tang:
To my knowledge, the Act will limit the secondary trading activity of the bond in theory. We just want to make sure we are comfortable with that legally. 

guest_Amit:
KDB is rated as BBB+ at Korea's sovereign FC rating level and as KDB accounts for bulk of the put options and credit facility, the transaction rating is closely linked to KDB.

guest_bama:
How does the two-SPV structure address the issues under the ABS Law?

guest_Don Tang:
As I mention, there could be structuring risk. Also, prepayment risk of the bond should also take into account. 

In terms of ABS law, our legal counsel advises us to use ABS law to transfer asset under two SPV structure.

vinodk:
To Amit and Don, will Korean supervisory norms allow a loan to be off the balance sheet though sold with full recourse?

guest_Amit:
It will continue to appear as contingent liability, for the put options provided

guest_bama:
Amit, does this mean that the banks would continue to provide capital against the loans?

guest_Amit:
Yes, I suppose so. I am not sure about the exact capital requirement. But these put options rank pari passu with their other unsecured obligations.

guest_bama:
Could you give us some feel for the net cost of the transaction to KDB given the payment of fees for the credit facility.

guest_Don Tang:
I haven't run the numbers. Please remember, KAMCO has the asset, not KDB.

vinodk:
Amit, back on the question of providing capital, here is something I read: "The sale therefore leaves the banks with the risk of the loans, but enables them to take the assets off balance sheet. The banks can also put the expected collections to productive use immediately, rather than waiting several years for the money to come back"

guest_Amit:
That's right. The assets go off balance sheet but a contingent liability arises in relation to loans sold off

guest_bama:
On the issue of investor response, was there a book building exercise to fix the yield on the transaction? What was the extent of over-subscription?

guest_Don Tang:
Sorry, I can't disclose the details, but it is very satisfactory.

vinodk:
Re Bama, a report says 1.25 billion – Don, I am sorry it was there in the Press

guest_bama:
Were the obligors of the underlying loans notified of the transaction? 

guest_Don Tang:
Yes

vinodk:
Don and Amit, to what extent is this methodology relevant for other markets say India? Amit, in particular, could you think of some agency in India similar to Kamco?

guest_Amit:
Well the structure can be applied in India. Kamco is basically like RTC in US..

guest_Amit:
I do not think there is any agency in India similar to Kamco, as the disclosed NPL position in India is not as bad as in other E Asian countries. Well if some banks in India want to clean up balance sheet, they could possibly try to use this kind of structure.

vinodk:
Do you think the situation in Korea is comparable to India?

guest_Amit:
The current situation is not comparable to India, given the magnitude of NPL build up in Korea after 1997.

vinodk:
Amit, if banks in India were to want to do it, who could be similar to KDC, for credit enhancing?

guest_Amit:
KDB is at sovereign level, so i presume SBI, IDBI

guest_Don Tang:
How about the legal framework in India? Can it support this sort of structure? vinodk:
Amit, re the legal structure, being a common law country, it should not be a problem, with some amendments to the Debt Recovery law.

guest_Amit:
Vinod, for onshore fine, what about offshore structure

vinodk:
Amit, offshore scenario is not clear – may with a tax haven with double tax avoidance treaty.

guest_canberra:
what is the smallest securitisation that can be achieved??

guest_Don Tang:
To be economical justified. At least USD 150 MM minimum.

guest_canberra:
Don, Just in passing, what about creating an SPV which largely took on A rated paper and added some mortgages to enhance the rating. Any arbitrage?? Is the SPV an opportunity for arbitrage trading in bonds??

guest_Don Tang:
Are you referring something like a CDO plus mortgages?

guest_canberra:
yes, 

guest_canberra:
The mortgages being from prime jurisdiction such as Australia and bonds from India 

guest_Amit:
In this transaction it will be difficult to rate above BBB , even if better quality mortgages are added, if the balance risk is on KDB.

vinodk:
Amit, canberra is probably not referring to Kamco.

guest_Amit:
You can possibly try to create a CBO type structure, wherein tranching cud take it to A

guest_canberra:
This may lead to an arbitrage type of yield if the opening is right in the market. Are there investment banks already bringing deals of this type of structure to market??

vinodk:
Don and Amit, would you say tranching is the answer to Canberra's question?

guest_Amit:
if CDO type structure is being referred to, then tranching is possibly right. We have seen CDO structure with ABS assets.

guest_canberra:
Ys, I do see the complete picture analysis being relevant. I was wondering if the tranching of different assets and the enhancement given by the mortgages may capitalize on inefficiencies of the different markets??? It is certainly a profit opportunity if it is viable. 

guest_Don Tang:
I would not say no. It all comes down to the final sizing and pricing.