The period of fast development in the mortgage lending system in several transitional countries including Russia during the first years of the 21st century coincided with the period when securitization of mortgage loans was the most fashionable method of housing finance. The method was based on converting loans into securities which included: bundling mortgage loans into pools, selling these pools to SPVs (Special Purpose Vehicles), issuing securities based on the mortgage loans, making them rated by rating agencies and selling the securities to investors. This method of housing finance has become extremely popular among Russian bankers because securitization enabled them to tap cheap resources from abroad and provide mortgage loans at attractive terms competing with mortgage programs of state - owned institutions (banks and a secondary mortgage market agency).
Through the first years of the century securitization was extremely popular among mortgage lenders, not only in Russia but all over the world. In developed countries numerous well-established financial institutions gradually started to substitute other methods of finance by the securitization method. The share of the mortgages financed by issuance of securities has grown in many banks in developed countries. For example in the notorious Northern Rock, the share of mortgages financed by issuing securities was close to 50%.
One of the specifics of Russia was that in that country rather a big number of banks used securitization as the only method of housing finance. So nearly 100% of their mortgage portfolios were financed by securitization. Even among the top 10 Russian mortgage lenders, there were banks that relied exclusively on securitization (for example Moskommertsbank or City Mortgage Bank).
More than that. These banks were initiated as specialized mortgage institutions whose business model was based on securitization. They had no other sources of finance and never planned to have any. The whole business process of these banks was dedicated to the goal of preparing securities that would satisfy the requirements of potential investors. Since it was clear that the investors in their turn would rely on rating agencies' assessments of the loans, the methodology of assessment developed by the agencies became the major driver for the mortgage lending business process (underwriting, processing, and servicing) for these banks. The banks tried to create a portfolio that would satisfy the rating agencies rather than their own risk management.
It is clear that when investors shunned from securities based on mortgage loans, these specialized banks suffered even more than such banks as Northern Rock whose financial sources were diversified. (Only from one point of view they suffered less. There was no run on these banks because they had no deposits). The most prudent specialized banks have secured in time credit lines from foreign financial institutions (mostly from shareholders of these banks). Others were forced to radically reduce their mortgage lending activities. In most cases they achieved thise result by increasing mortgage interest rates to a level above the average on the market. For example, in September 2007 Moskommertsbank increased its mortgage interest rate by 1 % while Ursa Bank increased its by 2% in spite of the fact that all the major players kept their rates intact.2
Currently several of these banks are in a dangling position. They should as soon as possible clarify for themselves why the business model they have selected turned out to be so vulnerable to liquidity risk and whether the model can be amended and used in the future or whether it should be substituted by another financial model. The same problem concerns regulators.
This paper will argue that a housing finance system based on securitization of mortgage loans has imminent shortcomings, which inevitably make the system vulnerable to liquidity risk.3 The shortcomings are associated with the way the fundamental risks of mortgage lending are distributed between participants of the mortgage lending process if securitization is chosen as a housing finance system.
It is well known that fundamental risks of mortgage lending cannot be made to disappear but can only be distributed and redistributed between various participants involved. For every risk, a participant can be identified that is better equipped than others equipped to manage the risk. We will call the participant a "risk-relevant" participant. Different housing finance systems are characterized by different schemes of distribution of risks between participants and (unfortunately) by the risks they appoint to various non-risk-relevant participants.
It is very important (but unfortunately rarely noticed) that under various housing finance systems some of the risks are allocated to non-risk-relevant participants with the following negative effects:
- the risk margin added by the participant to hedge the risk tends to be higher than it would be if the risk were managed by a risk-relevant participant. So mortgage interest rates for final borrowers will increase.
the value of other risks often increases dramatically so that mortgage lending becomes a riskier business.
The purpose of this paper is to demonstrate that the unprecedented growth of liquidity risk under the housing finance system based on securitization of mortgage loans is just one of the examples when one of the risks (liquidity risk) has grown because another risk (credit risk) was transferred under the system to the non-risk-relevant participants.
Major housing finance risks
There are several fundamental risks inof a housing finance system. The most important of them are credit risk, interest rate risk and liquidity risk.
Credit risk in housing finance systems usually means the risk that the borrower will default and the loan owner (or lender) will not be able to cover its losses by means of foreclosure. Below we will refer to thise risk as to the borrower credit risk.
Intermediary credit risk is the risk of default of the financial intermediary that attracts financing for mortgage loans from the market (capital market or deposits markets) financing for mortgage loans. Since in our case the intermediary attracting investments is the mortgage lender we will name the risk - the lender credit risk.
Interest rate risk is the risk that interest rates will rise. Liquidity risk refers to inability to get access to the cash when necessary.
Before turning to the analyses of the system based on mortgage loans securitization we will have a brief look at the traditional housing finance system Particularly we will have a look at how under this system allocation of one of the risks to non - risk-relevant participants causes increase of another risk.
Traditional housing finance system
Within a traditional housing finance system, a lender (bank, building society, credit union, thrift, etc.) is typically responsible for origination, servicing, funding and portfolio management of mortgage loans The sources of funds for the mortgage loans under the system are debt obligations of the lender. These obligations are in most cases deposits but also may be in the form of mortgage (or non mortgage) bonds, dedicated savings, loans from other financial institutions or from special liquidity facilities, etc.
If the mortgage loans provided by the lender are fixed rate loans, the risk distribution under the system is the following:
- borrower credit risk - the lender,
- interest rate risk - the lender,
- liquidity risk - the lender,
If we start analyzing the risks from the point of view of whether they are managed by risk-relevant or non-risk-relevant participants we will see that the borrower credit risk is managed by a risk-relevant participant - the lender. The lender is the best equipped among all the actors to bear the risk since it knows the borrower, the property and in many cases the specifics of the local market and local community.
On the contrary the interest rate risk is borne by the lender, which is poorly equipped to mange the risk (it is a non-risk-relevant participant). The lender borrows short (mostly deposits) and lends long (long-term mortgage loans). It means that each mortgage loan through its life is financed by a chain of several short-term deposits. Each of the deposits is attracted by the bank on a particular day at a market interest rate for deposits of that day. Often, the loan on another hand was issued by the bank at the market interest rate for mortgage loans of the day of the issuance at a rate fixed for whole life time of the loan.
If, through the period of the mortgage loan, life the market interest rate for deposits rises, the margin of the bank (the difference between the mortgage interest rate and deposits interest rate) decreases and may even become negative, causing instability or even the bankruptcy of the bank.
By changing the form of lending from fixed rate loans into ARMs (adjustable rate loans) interest rate risk is transferred from lenders to borrowers. But since the borrowers are also not equipped to mange the risk, the transfer causes another misplacement of the risk from one non-risk-relevant participant to another. This misplacement can also result in growth of another risk and another risk's growth actually happens. In this case, the growing risk is not a liquidity risk but the borrower credit risk.
Even with stable macroeconomic environment, delinquency rates of ARMs are approximately three times higher than for fixed rate mortgages. In periods of fast growth of interest rates borrower credit risk grows to a level unbearable for lenders. The most notorious samples are mass defaults in South Africa and less dramatic events that took place several years ago in Great Britain.
In South Africa where ARMs dominate the mortgage market, the growth of interest rates during the late 1980s caused numerous defaults. As a result, most financial institutions withdrew from mortgage loan originations at many territories initiating practice of geographic discrimination - the so called redlining.4
In order to attract banks back into the mortgage market, the government established (on parity terms with the Banking Council) a special company named Servcon, the mandate of which was to purchase more than 30,000 thousand defaulted loans from the banks. Nevertheless the redlining - the major negative effect of the burst of the credit risk caused by the interest rate risk misplacement - has not been completely eliminated till now.
The sample demonstrates how interest rate risk can result in negative effects both on pricing and on availability of mortgage loans. The rise of credit risk was due to a shift of interest rate risk to a non-risk-relevant participant (the borrower).
Securitization as a housing finance system
If we talk about plain vanilla securitization5, the risk distribution under the securitisation system is the following:
- borrower credit risk - investors,
- interest rate risk - investors (a portion of the risk always remains with the lenders as a pipe-line risk),
- liquidity risk - lenders,
The situation is totally different from the traditional system. While under the traditional system practically all risks are borne by lenders and / or borrowers, the bulk of the risks under the securitization system are transferred to investors. Lenders keep only the liquidity risk and this risk, as we all know now, becomes their Achilles' heel.
Why does it happen? The major reason is in transferring of borrower credit risk to investors. Investors cannot measure borrower credit risk. Their risk management branches (if there are any) have neither knowledge nor ability to measure the final borrowers' credit risk. They do not know the specifics of local borrowers, cannot assess the quality of underwriting, the reliability of independent appraisers, etc. It means that for that risk investors are non-risk-relevant participants.
Being non-risk-relevant participants investors have no choice but to rely on the assessment of borrower credit risk conducted by a third party. The only third party they can rely on is a rating agency hired by the lender.
As soon as investors come to the conclusion that they cannot rely completely on the assessment of these particular third parties (there may be plenty of reasons for that) they have two opportunities: either to rely on the assessment of another third party or to avoid the investments bearing the credit risk of the final mortgage borrowers altogether.
If the reason for losing trust in a rating agency is just misbehavior (fraud) or a mistake conducted by one employee of the institution the investors will probably just refrain from buying securities assessed by this particular agency. But if the reason is different and investors have a reason to mistrust all the assessments of the final borrowers' credit risk conducted by all the agencies they will refuse to buy any securities bearing the risk.
For example when it was discovered that all the rating agencies were a bit too optimistic in assessing securities based on sub-prime mortgage loans the investors preferred to avoid buying securities based on any types of mortgage loans. They did it because they had a reason to believe that the assessment of these securities was also too optimistic.
Since the borrowers' credit risk was transferred to non-risk-relevant participants the liquidity risk borne by lenders turned out to be extremely high. When the nonrisk-relevant participants (investors) lost trust in rating agencies they refused to acquire securities altogether because being non-risk-relevant participants they had no means to manage the risk themselves. If they were risk-relevant participants at their place (having the ability to measure and manage the risk themselves), they would probably just have reduced their purchasing activity or added an additional risk margin.
Since all investors refused to buy securities completely, all the institutions (mostly primary lenders) that bear liquidity risk, were hit much stronger than they would have been hit if the borrowers' credit risk were met by risk-relevant participants.
What can be done to revive securitization as a housing finance system?
It is clear from the above that securitization as a housing finance method could be improved if the opportunity is found to transfer borrower credit risk from non-risk-relevant participants (investors). There are several ways to do it. The one most fully tested is the US secondary mortgage system.
The system is based on two Government Sponsored Enterprises (GSE): such as Fannie May (FNMA - Federal National Mortgage Association) and Freddie Mac (FHLMC - Federal Home Loan Mortgage Corporation). Under the system, lenders underwrite mortgage loans in accordance with GSEs' standards, issue the loans, sell the loans to GSEs and service them (sometimes transferring servicing functions to specialized servicing institutions).
The GSEs, in their turn, bundle mortgages into pools and issue securities backed by the underlying collateral of these loans. The securities are sold to investors together with a full GSE guarantee against borrower credit risk.
The major risks are distributed in the following way:
- borrower credit risk - GSEs,
- interest rate risk - investors,
- liquidity risk - GSEs.
If we consider the system as an improved securitization system (historically it is not so because the secondary mortgage market system was developed earlier) we can see that the major difference between the systems is that under the secondary mortgage system the borrower credit risk is transferred from one non-risk-relevant participant of the process (investor) to another non-risk-relevant participant (the state). At first glance, it does not seem reasonable because the state is not better equipped to meet the borrower credit risk than the investors.
Nevertheless the secondary mortgage market system works much more smoothly than the securitization one. It happens because the liquidity risk under the system turns out to be very low. Since investors under the secondary mortgage system do not bear credit risk they cannot change their perception of it and hence are unlikely to shun from purchasing mortgage-based securities.
Another advantage of the system is low cost mortgage loans for the final borrowers. GSEs do not add a credit risk margin because they have implicit and explicit state support and can rely on it even under adverse economic conditions. Other participants add a minimum risk margin because they are risk-relevant participants for the risks they bear. The mortgage interest rate for the final borrowers turns out to be the lowest possible.
At the same time the system has one serious shortcoming that hinders its development. If the system is used and borrowers credit risk is transferred to the GSE (actually to the state) the state exposure becomes extremely high.
Experience shows that that there are practically no countries besides the US that can afford and are willing to reduce the liquidity risk of mortgage lenders at the expense of making the government responsible for the credibility of vast number of mortgage borrowers.
From here it follows that if the securitization system were transformed so the way that the borrower credit risk is kept by the lender (risk-relevant participant) rather than investors or government (nonrisk-relevant participants) and at the same time state exposure is not increased too much, the risk distribution would become close to the ideal one.
The result could be achieved in various alternative ways.
Alternative 1. The lender sells the loans to an SPV. In addition the following is done:
- the lender provides a guarantee to the SPV against the borrowers' credit risk (for example a guarantee to repurchase delinquent loans),
- a back-up guarantor becomes a participant of the housing finance system. The back-up guarantor makes a pledge that it will substitute the lender as a guarantor in case of default of the lender.
Alternative 2. The lender sells the loans to a conduit which is (or is related to) a strong institution (we will also name the institution a back-up guarantor). In this case the following is done:
Bearers of major risks under the system (under both alternatives) will be:
- borrower credit risk - lender,
- interest rate risk - investors,
- liquidity risk - lender,
- lender credit risk (default of the lender) back-up guarantor,
- Guarantor risk (default of the back-up guarantor) - investors.
The major difference between two alternatives is that under Alternative 2 investors in the case of the default of the conduit and the back-up guarantor will inevitably lose money while under Alternative 1 if the lender and the borrowers remain solvent the investors will not encounter any problems.
Major characteristics of the system (both alternatives):
- both borrower credit risk and interest rate risks are met by risk-relevant participants
- the institutions best equipped to manage the risk (relative risk margins are minimal),
- the lender provides a guarantee to the conduit against the borrower credit risk (for example a guarantee to repurchase delinquent loans),
- a new risk - the risk of the back-up guarantor failing - emerges and is met by investors (the relative risk margin depends on the creditworthiness of the back-up guarantor),
- liquidity risk becomes dependent on the probability of changes in investors' assessment of the back-up guarantor's creditworthiness (the relative risk margin depends on the creditworthiness of the back-up guarantor).
From here it follows that lower interest rates (a reduction of risk margins to the lowest possible level) could be achieved by the proper selection of the back-up guarantor. The role of the back-up guarantor may be played by:
1. The state (in developed countries with a high credit rating of the country).
Positive impact of the selection:
- the state is well equipped to meet the lender's credit risk since the level of the risk is managed by the banking regulations and by banking supervision conducted by the respective Central Bank or another specialized state entity.
- in case of default of the lender the backup guarantor takes the responsibility as an owner of the loans (or as a guarantor of the Conduit).
- for investors the credit risk of back-up guarantor will become equal to the risk on government debt,
Negative impact of the selection:
- the state exposure will grow (though to a lesser extent than under the system based on the secondary mortgage market).
2. International organizations such as World Bank, UN Habitat, OPIC, etc (for transitional or developing countries).
Positive impact of the selection:
- corresponds with the mission of the international organization (promotes private investments into transitional countries, supports development of financial markets in the countries, increases housing affordability, stimulates housing construction, reduces poverty, etc).
- from the investor's perspective, the credit risk of back-up guarantor will become equal to zero,
Negative impact of the selection:
- international organizations' support can be provided only for a limited period of time. Substitution of the international organizations back-up guarantee for another type of back-up guarantee may cause a shock for the finance system of the country.
3. Association (partnership) of lenders (in countries with a strong banking system). Lenders can create a back-up guarantor in the form of a mutual guarantee-fund, a coowned specialized insurance company, etc.
Positive impacts of the selection:
- there is no state exposure,
- association of lenders is well equipped to meet the lender credit risk of the members of the association (provided that it has the knowledge necessary to measure the risk level),
Negative impacts of the selection:
- back-up guarantor risk and hence liquidity risk are not eliminated (in the case of systemic problems in the financial sector the fund may become unable to fulfill its obligations).
The problem of liquidity risk could be solved in the case by the limited state involvement. The state may be involved either as a co-founder of the back-up guarantor or as a liquidity provider to the back-up guarantor.
If any of the above described amendments to the securitization system are made the lenders will not be able to obtain capital relief while transferring credit risk to investors. The credit risk will be kept with the lenders. As a result, they will face a worse credit risk vs. capital ratio. Probably it is paradoxical but in spite of that the financial system as a whole will be more stable. It is explained by the fact that credit risk (which cannot be made to disappear) will be kept by the risk-relevant participants.
It seems that very close to the described above system is the MPF (Mortgage Partnership Finance) system developed several years ago in Chicago6. The system is based on the scheme as outlined in Alternative 2. Lenders sell their mortgage loans on a recourse base to a strong institution that fulfills conduit functions - the Federal Home Loan Bank (FHLB) of Chicago. FHLB is a statebacked institution so the functions of a back-up guarantor are fulfilled by the government. The securities issued by the FHLB are considered by the market as credit risk free which helps both to reduce mortgage interest rates and to eliminate liquidity risk.
Unfortunately in the US where the Government willingly accepts the borrowers' default risk under the secondary mortgage system the MPF system could not demonstrate its advantages. Other countries also have not been interested to promote the system since the securitization system has been developed rapidly. Nowadays the situation has changed and it seems a right time to revive and promote the system for the usage in various countries.
2 See Interface news agency report http://www.rusipoteka.ru/research/interfax-.htm.
3 Liquidity risk will be defined below.
By Victor Mints
4 See Mary R/ Tomlinson, "South Africa's Financial Sector Charter: Where From, Where To?" Housing Finance International, December 2005
5 Most of securitization deals include special mechanisms actually transferring some of the major risks (very often interest rate risk) to lenders
1 Housing Finance Expert, at Financial Corporation Uralsib, Russian Federation.
Article source by :Mints, Victor
0 komentar:
Post a Comment