Banks' Securitization Indicator

International Research Journal of Finance and Economics ISSN 1450-2887 Issue 24 (2009) © EuroJournals Publishing, Inc. 2009 http://www.eurojournals.com/finance.htm Banks’ Securitization Indicator Roza Hazli Zakaria Lecturer, Faculty of Economics and Administration University of Malaya Abdul Ghaffar Ismail Professor of Islamic Economics and Finance, Faculty of Economic and Business Universiti Kebangsaan Malaysia Abstract This paper proposes an alternative measure to quantify banks’ off-balance sheet securitization activity. We argue that the conventional method of simply taking the notional value of each securitization instrument to represent banks’ securitization activity provides bias estimates since it ignores other factors such differing bank sizes and the variety aspect of securitization, while using dummy failed to gauge the extent of banks’ involvement in securitization. Hence, our proposed indicator is novel in the sense that it takes two aspects of securitization into accounts; which is the amount of securitization activities as well as the different combinations of instruments that constitute banks’ securitization activity. Besides, measure to prevent biases from banks’ size is also taken into account. Application to the Malaysian setting demonstrates that the indicator is able to distinguish between active and passive players and it provides significantly different measure as opposed to the conventional methods. At the same time, the indicators and its relationship with banks’ performance variables are consistent with theoretical prediction. Keywords: Off-balance sheet securitization, banking, Malaysia JEL Classification Codes: G20, G21 1. Introduction Activities banks engage in have altered significantly. Instead of concentrating on their traditional activities such as taking deposits and extending loans, banks are now increasingly involve in nontraditional activities such as asset securitization, loan sales, underwritings and credit derivatives. Banks are also shifting from interest-based revenues towards fee-based activities, including lines of credit and many types of credit guarantees. This structural shift is prompted by the changes in financial environment they operate in. Interest rate liberalization that results in stiff competition on deposit and lending rate, hence squeezing banks’ profit margin; competition from non-bank institutions that offer traditional banking services; the rapid growth of commercial paper market that took away highly rated, corporate clients; as well as the more structured and stringent capital requirement, dictate that it is no longer optimal for value-maximizing banks to concentrate on their traditional activity of making loans and accepting deposits. Thus, banks are aggressively restructuring their activities. Naturally, this banking revolution calls for researches to judge the viability and profitability to justify this change, both at the micro and macro level. For instance, policy makers are concern on the impact these changes have on the availability of bank lending as well as the stability of banks International Research Journal of Finance and Economics - Issue 24 (2009) 224 following the theoretical prediction that banks tend to be more aggressive in risk taking given the ability to diversify their portfolios. Empirically validating this requires an indicator that could reflect the extent of banks’ securitization activities and at the same time includes every securitization instruments in a bank portfolio. Tendency of existing literature to confine their study to asset-backed securitizations or a few specific instruments explain why conventional measures of measuring banks’ securitization activities failed to depict the extent of banks’ securitization involvement. Nonetheless, choosing a measure to represent a bank’s securitization activity is indeed a complex task due to the broad range of instruments yet limited data available. The variety of instruments, thereby generating the difference specifications of each, has given a rise to “identification and quantification” issues. Identification refers to defining the scope of transactions that could be included in securitization portfolio. Quantification refers to the issue of choosing the appropriate measure of banks’ securitization activity that will allow us to differentiate between an active and passive player in the aggregate securitization market. This motivates us to propose an indicator to quantify banks’ off-balance sheet activities. The novel aspect of our indicator is it allows the inclusion of every types of banks’ off-balance sheet activities, as permitted by the data. At the same time, it takes into consideration factors like the difference of banks and the difference of each banks’ involvement in one particular types of offbalance sheet transactions. Hence, it gives an indication that is not bias to the size of banks or any dominance in one particular instrument. This results in an indicator that could be used to measure the degree of banks’ involvement in their off-balance sheet activities. Next section will discuss the identification issue followed by the quantification issue. In section 3, the construction of a proposed securitization indicator will be outlined. A calculated real data will be presented subsequently, in section 4, followed by the verification of the indicator’s ability to reflect a bank’s degree of involvement in securitization activities in section 5. A comparison of the measurement produced by the proposed indicator and the conventional measure is presented in section 6. Section 7 concludes. 2. What Constitutes Banks’ Securitization Activity? Given that existing literature are divided where the definition and scope of securitization are concerned, this review begins by offering reconciliation on the definition, hence the scope of banks’ securitization activity. Securitization is one of the financial innovations that have a significant impact on the function performed by the banking sector since it implies a shift of credit flows from bank lending to debt securities. Cummings (1987) describes securitization as a process of matching up borrowers and savers wholly or partly by way of financial markets. Thus, it represents the process of disintermediation in which borrowers shift from indirect borrowing from financial institutions such as banks to direct borrowing in the financial market through the issuance of securities such as bonds and notes. Consistent with this development, banks too start shifting their financing mode, from onbalance sheet lending to off-balance sheet activities such as back-up credit facilities and contingent banking. Hence, where banks are concern, securitization was originally applied to the process of disintermediation, or the substitution of security issues for on-balance sheet bank lending. Within this context, securitization includes standby letters of credit, guarantees, underwritings and loan commitments. Earlier researches on securitization, such as Greenbaum and Thakor (1987), Benveniste and Berger (1987), Berger and Udell (1993) refer to these instruments to represent banks’ securitization activity. Then, further innovations are developed towards refinancing of existing banks’ loans by packaging them into a tradable form through the issuance of securities; better known as loan sales in cases where a single loan is sold or asset securitization in cases where homogenous loans are repackaged to be sold to investors. These activities and the earlier securitization instruments described 225 International Research Journal of Finance and Economics - Issue 24 (2009) above, shared the common feature of separating many of the services associated with lending, such as credit risk evaluation and underwriting, from the funding of loan, (James 1987). However, as opposed to the broad measure of securitization, asset securitization refers only to the process of pooling and repackaging loans into securities that are then sold to investors, (Ergungor 2003). Under this definition, banks’ securitization activity is confined to loan sales, asset-backed securitization as well as mortgage-backed securitization. The development of this new securitization instruments has been dominating the scope of banks’ securitization activities. Recent literature on banks’ securitization activity tends to restrict their studies on securitization specifically to asset securitization, for instance Wolfe (2000), Murray (2001), Roever et al. (2003), Ergungor (2003), Ambrose et al. (2003), Minton et al. (2004), Calomiris and Wilson (2003). This could be attributed to the complexity of the process involved in asset-backed securitization and the possibility of continuously including new types of banks’ assets to be packaged and sold in the secondary loan market. For instance, asset-backed securitization which started with real estate loans now comprise of commercial and industrial loans, study loans and credit card receivables. Parallel to asset-backed securitization, the increasingly volatile financial environment has further spurred banks to be more innovative in risk-management. Banks start involving in synthetic securitization, which are derivatives trading, such as swaps, options, swaptions, and futures and forward. Again, the relatively complex process and since it is a more recent securitization instruments, it claims a great deal of scholarly research attention. Nevertheless, more often than not, research on banks’ derivatives involvement is being carried out independently, not as a portfolio of securitization activity. By understanding the development of banks’ securitization involvement, it implies that banks’ securitization involvement is not restricted to asset securitization, as widely perceived. Instead, it involves more comprehensive instruments, from off-balance sheet back-up credit facilities, contingent banking, commitment, derivatives and asset-backed securitization. It is the research preference that has either disregarded the more conservative modes of securitization such as contingent banking and commitments, or purposely chose to focus only on either derivative or asset-backed securitization, and not on securitization portfolio. Essentially, securitization should be defined in a manner that reflects its content or instruments and subsequently determined the scope of banks’ securitization activities. Perhaps the best way to define securitization is to refer to the regulatory perspectives. Ernst and Young (2002) claims that securitization includes any transaction under which a securitization vehicle directly or indirectly acquires receivables or bears risk associated with commitments taken or activities carried out by third parties and issues in exchange securities whose return is directly linked to the risks borne. Within this definition, securitization describes a broad range of disintermediation and off-balance sheet activities of commercial banks including issuing standby letters of credit, extending loan commitments, selling loans with and without recourse and manipulating derivatives instruments. This definition and the scope of securitization that follows is also consistent with the Basel Consultative Documents 2003, which dictates that “(banks’) securitization exposures can include but are not restricted to the following: asset-backed securities, mortgage backed-securities, credit enhancements, liquidity facilities, interest rate or currency swaps, and credit derivatives.” (The Basel Consultative Document 2003, para 504, page 100). Reliable and consistent data on specific bank’s securitization activity is not available. Following Fung and Cheng (2004), this study would assume that the transaction volume in broad categories, such as direct credit substitute, trade-related contingencies, transaction-related contingencies, exchange rate and interest rate contracts, can to a certain extent reflect banks’ securitization activity. International Research Journal of Finance and Economics - Issue 24 (2009) 226 3. Quantification Issue Quantification issue evolves around how to measure the depth of a bank’s involvement in securitization activity. Ideally, the indicator should be able to reflect the extent of a bank’s securitization activity or could set an active and a passive bank in securitization market apart and at the same time, permits a cross-section comparison. Previous studies on securitization has approached this issue using a few different proxies, namely ratio of securitization to total assets (Georges and Harchoui 2003), the notional value of each securitization instruments (Hugh 1998; Fung and Cheng 2004) and dummy variable (Demsetz 2000; Cebenoyan and Strahan 2004). In most cases, these studies were referring specifically to a single securitization instruments, thus the usage of the mentioned variable is deemed enough to capture the extent of a bank’s involvement in securitization activities. On the contrary, this study argues that to learn the impact of securitization on bank behavior requires the inclusion of a comprehensive list of securitization instruments or the entire items that made up a securitization portfolio. This is so because there are a myriad of securitization instruments available and a bank’s choice of securitization instruments varied according to their need. For example, a bank might adopt a particular instrument for hedging purposes while at the same time, adopting another to arbitrage capital. Thus, each securitization instruments will have their own bearing on a bank behavior. Hence, the ultimate impact depends on the combination of the securitization instruments adopted. For that reason, this study argues a study that intends to observe the impact of securitization on bank behavior must not restrict their study to just a few selected instruments. Instead, it should include every single instrument on securitization portfolio. Therefore, the measure should be able to account for the variety of the securitization instruments that make up a single securitization portfolio. However, as mentioned above, specific, detail and consistent data on banks’ transactions in each securitization instruments is not available, at least in the Malaysian setting. Hence, the best that can be done is to include each broad section of securitization instruments, as reported in respective banks’ financial statement. Nonetheless, in an economy where the data on securitization transactions is reported in a more specific manner, then a study on securitization should attempt to include every securitization instruments available, as permitted by the data. Conventional analysis has resorted to employing either dummy or the notional value of securitization activity. Assigning a dummy to differentiate between a bank that adopts a particular securitization instrument and one that does not, cannot depict the extent and effect of securitization by banks. The results could only demonstrate the difference between banks that adopt securitization and those that do not, rather than those which are actively versus passively involved. On the other hand, using total notional value of securitization activities to reflect the extent of a bank’s involvement in securitization activities would be bias since small banks, might have a small notional value. However, this does not mean that it is a passive player. The more appropriate measure would be to look at the proportion of securitization to total assets to determine whether a bank is an active player or vice versa. Another commonly used indicator, percentage of notional value of securitization from total assets does not take into account the “variety” of securitization instruments. A bank might have a large value of transactions in a single securitization instrument, but its involvement is purely concentrated in that instruments. Therefore, it is bias to define this bank as an active securitization player. Thus, this study proposes a securitization indicator that is able to quantify the extent of a bank’s securitization activity, given the wide variety of instruments available and accounts for differing bank sizes. 227 International Research Journal of Finance and Economics - Issue 24 (2009) 4. Securitization Indicator: The Construction An indicator is supposed to track the total value of a basket of commodities, (Belter et al. 2005). In this case, basket of commodities is a portfolio of securitization instruments with fixed or time-varying weights. In addition, Fung and Cheng (2004) point out two important dimensions of financial innovations, (securitizations) that are volume of transactions and “variety”, such as the number of creative combinations of securitization instruments. As such, the indicator should be able to capture both volume and variety aspects of banks’ securitization activity. The limitation in constructing the indicator is limited data availability. The only data or detail that is available is the volume or the notional amount of securitization transactions, in broad general categories. Therefore, that is the only variable to be manipulated in constructing the indicator. As noted above, the first criterion to judge whether a bank is actively involve in securitization activity is to compare its notional volume of transactions in each securitization instruments to its total assets, or its size. This is to overcome the possibility that a small bank, which is an active bank, is deemed as passive, since its notional value of securitization transactions is much smaller compared to larger banks. Accordingly, the higher the ratio of its securitization activity to total assets, the more active a bank is in its securitization activity. Secondly, a bank’s contribution in each securitization instruments’ market should also be noted in determining its degree of market participation. Holding other factor constant, a bank’s securitization instruments’ notional value that constitutes a large amount of the aggregate value is a more active player as compared to its counterpart. Therefore, the degree of a bank’s involvement in securitization activity is jointly determined by its notional value as a percentage of total assets and its value as a percentage of the aggregate market value. Another important dimension to consider in selecting the indicator, is the diversity of the instruments available under securitization. Therefore, the measure should take into account the variety of the securitization instruments. The more securitization instruments that are adopted by a particular bank should represent a greater participation as opposed to the banks that choose to concentrate only on a few securitization instruments. Hence, an aggregation method is adopted in constructing the indicator. Besides, this aggregation method allows any new or instruments that are unique to any one bank be included in the measurement of the level of securitization activity. This feature is deemed important as banks’ securitization instruments keep increasing and developing through time. In order to aggregate every securitization instruments in a group, a weight is needed to reflect the relative importance of each instrument in a portfolio. The choice of weight to be used is limited, since the only data available is the volume of securitization activities. This study employs a particular instrument’s market share in the aggregate securitization market as the weight. Nevertheless, market share is one of the commonly used choices for index weights, apart from share prices and equal weights, (Belter et al. 2005). Thus, every agent (bank) is subject to the same securitization instruments’ weight. However, this weight varies annually. Mathematically, the proposed indicator for each bank can be written as: = ∑ (seci , j , / totalassets j )(seci , j , / ∑ seci )t wi N i =1 [ ] Where N = number of securitization instruments i = securitization instrument, i j = bank j w = market share of securitization instrument, i The indicator for a bank’s involvement in each securitization instrument is given by a multiplication of the ratio of notional value of that particular securitization instrument to total assets; and the ratio of the notional value to the aggregate volume of that particular instrument. This is to overcome the bias that would prevail if only notional value of securitization transactions is used in International Research Journal of Finance and Economics - Issue 24 (2009) 228 quantifying a bank’s involvement in securitization activity. The two ratios are multiplied to make the difference among an active and a passive player more discernible. The calculation describe above would yield a different indicator for each securitization instruments, for each banks. To account for “variety”, the indicator for each securitization instruments above, are aggregated, for each bank. Therefore, banks that adopt more securitization instruments should have a higher value. The indicator is calculated on annual basis to gauge the extent of a bank’s involvement in securitization activity. Banks that are not involved in any securitization activity would yield the value of zero. On the other hand, the higher the indicator, the more active a bank is in securitization activity. However, there is no maximum limit for the indicator for two reasons. First, the notional amount is compared to its total asset. Since securitization is an off-balance sheet item, total asset is not a constraint, therefore it is not the maximum amount of securitization that a bank could undertake. Hence, the maximum ratio of notional amount of securitization to total assets is not restricted to 1. Statistics depict that there are banks, especially foreign banks in the case of Malaysia, manage to have a much higher amount of securitization transaction relative to its total assets. Secondly, the aggregation method keeps this indicator flexible. Any new securitization instrument, if available, even if it is only adopted by a single bank in a certain studied location, could be included. Thus, the maximum amount is not limited to existing numbers of securitization instruments. Only each bank’s contribution to each securitization instrument has a maximum value of 1, and the maximum market share of each securitization instrument is limited to 100. Since the indicator involves the interaction of all the above mentioned items, the fact that there are two items have no maximum values results in no maximum limit for the value of the indicator. 5. Applications Using the method described above, the calculated indicators for Malaysian commercial banks are summarized in Figure 1. The calculation of the index is done with 39 sample banks, since we segregate the Islamic banking sector in each bank, as an individual entity. A sample calculation of the index and the complete index table is given in Appendix A. However, the graph is plotted only for the 25 main commercial banks, including Bank Islam Malaysia Berhad and Bank Mualamat Malaysia Berhad. Nevertheless, except Affin Islamic Bank, other Islamic banks are relatively passive securitization players, with indicator lesser than 1. Hence, the omission of those banks from the graph does not distort the overall picture of the relative level of securitization involvement among banks. There are three observations that stand out from the graph. Firstly, the securitization indicator value for most banks is concentrated within the low region value, which is lesser than the mean (2.0236) value of the securitization indicator, over the studied period. This demonstrates that most of the Malaysian commercial banks are relatively passive players in the securitization market. This is supported by the descriptive statistics of the indicator. The indicator constructed to represent commercial banks’ level of involvement in securitization activity yields the following statistical properties, which is displayed in Table 1. The statistical properties too demonstrate the gap between the levels of involvement in securitization activity among commercial banks in Malaysia is large. The median is just 0.7808 but the mean is 2.0236. It shows that more than half of the banks are relatively passive players. 229 International Research Journal of Finance and Economics - Issue 24 (2009) Figure 1: Malaysian commercial banks: Securitization indicator, 1994-2004 index 20 EON Bank Affin Bank 18 Hong Leong Bank Maybank Public Bank 16 RHB Bank Bank of Commerc 14 Alliance Bank HSBC Bank 12 Citibank Ambank 10 UOB Bank of China Southern Bank 8 Standard Chartere OCBC 6 Deutsche Bank ABN Ambro 4 Bank of America Bank of Nove Sco Bangkok Bank 2 Bank of Tokyo Chase Manhattan 0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Bank Muamalat Bank Islam year Table 1: Securitization indicator: Statistical properties Value 2.0236 0.7808 19.3498 0.0000 3.2050 2.9237 12.6070 270 Description Mean Median Maximum Minimum Std. Dev. Skewness Kurtosis Observations Secondly, the graph depicts that there is no one definitive trend either for each bank, where their securitization activity is concern. Instead, each bank indicators are volatile over the years. This is due to two factors, which are the nature of bank’s securitization activity and the method of calculating the index. Banks engage in securitization activity for different reasons. For instance, banks that are short of capital might securitize to arbitrage capital in certain year. The same bank might also securitize to off load their risk to the market, in other years. Hence, each bank securitizes for different reasons, in every different time frame. Therefore, the amount of notional value of securitization itself varies greatly from year to year. For example, Table 2 to 3 below present the data for notional amount of securitization and the index for two banks, to show how the amount of each securitization instruments undertaken by banks vary greatly from year to year. For instance, Bank of America was not involved at all in interest rate derivatives activity in 1995, while the notional value of activity in that particular instruments was RM793.10 billion in 1994 and RM1 087.77 billion in 1996. Similarly, its involvement in foreign exchange derivatives products increase tremendously from RM3 612.62 billion in 1996 to RM8 153.67 billion in 1997. Similarly, notice the movement of notional amount of trade related contingencies International Research Journal of Finance and Economics - Issue 24 (2009) 230 product, for JP Morgan Chase Bank (Table 3.3), from 1997 to 2000. The value was RM26 525 million in 1997, in 1998 and 1999 this bank did not trade at all in this securitization instruments and recorded RM8 661 million of transactions in that particular instruments in 2000. Table 2: Bank of America: Securitization portfolio, 1994-2002 Direct Credit Substitute (RM million) 1994 1995 1996 1997 1998 1999 2000 2001 2002 Source: 78,152 64,101 283,196 272,750 225,621 213,807 219,563 128,176 118,080 Transaction Related Contigencies (RM million) 264,989 350,812 119,483 94,939 81,703 396,557 15,623 16,613 17,061 Trade Related Contigencies (RM million) 55,648 24,351 0 194,586 22,381 17,092 3,993 2,851 3,518 Commitments (RM million) 527,004 983,120 2,195,957 1,156,888 1,262,033 7,251,319 896,305 443,694 268,310 Foreign Exchange Derivatives (RM million) 2,371,625 5,455,045 3,612,644 8,173,656 4,937,629 9,534,173 5,353,152 999,616 1,507,384 Interest Rate Related Derivatives (RM million) 793,104 0 1,087,777 638,986 164,600 184,292 133,000 0 0 Indicator 3.7443 10.6684 5.3444 15.1803 8.7597 4.5866 5.0033 0.2559 0.6284 Bank of America Financial Statements, Various Issues Table 3: JP Morgan Chase Bank: Securitization portfolio, 1994-2003 Transaction Related Contigencies (RM million) 1,799,101 3,156,179 4,165,568 2,920,235 121,951 96,675 90,921 10,949 2,796 Trade Related Contigencies (RM million) 23,453 13,166 54,455 26,525 0 0 8,661 6,784 6,171 Foreign Exchange Derivatives (RM million) 93199 46,600 440,448 521,313 2,124,441 3,336,690 4,091,806 3,911,786 1,044,728 Direct Credit Substitute (RM million) 1994 1995 1996 1997 1998 1999 2000 2001 2002 Source: Commitments (RM million) Interest Rate Related Derivatives (RM million) 298,510 292,300 413,516 619,264 369,360 412,460 1,263,000 1,407,000 3,584,600 Indicat or 507,674 526,203 412,403 601,489 163,688 54,207 47,728 34,564 40,355 2,097 2,219 450,295 408,711 27,618 1,348 66,688 61,194 89,136 8.0630 10.7576 11.3307 3.9209 1.7605 2.7505 4.2440 3.6472 3.3198 JP Morgan Chase Bank Financial Statements, Various Issues Secondly, in constructing this indicator, the composition of overall securitization market is also taken into account. Since each bank’s choice of securitization portfolio allocation depends on various factors and varies greatly over the years, the effect on securitization market is more pronounced since the composition of market portfolio is formed by each commercial bank’s portfolio choice. Therefore, each bank’s market share of each securitization instruments, as well as each securitization instrument’s market share as compared to the aggregate securitization market varies accordingly over the years. Hence, when the final indicator value is calculated, the indicator is volatile even for each bank from year to year. Only banks that possess a securitization portfolio that is similar to the market securitization portfolio would have index readings that are more stable. In fact, given the uncertainty of the securitization market, the inclusion of market share and banks’ contribution in constructing the index allows it to reflect the market trend in each year. For instance, a bank that accords it securitization portfolio more towards instruments that make up a larger part of the securitization market would be considered more active. This is more so if its contribution to that particular securitization instruments is bigger. For this reason, we cannot expect a smooth and definitive trend for any one bank’s securitization activity. For a descriptive example, look at the value of the indicator for the Bank of America. As stated above, its involvement in foreign exchange derivatives products increase tremendously from RM3 231 International Research Journal of Finance and Economics - Issue 24 (2009) 612.62 billion in 1996 to RM8 153.67 billion in 1997. Given that foreign exchange derivatives claims a large market share (hence a higher w), this surge in the notional amount of securitization in derivatives, explains the volatile readings in the value of securitization indicator for the bank within that two years. Moreover, within this period, other banks involvement in derivatives is very limited. Therefore, this surge, too increases Bank of America’s contribution in derivatives transactions, thus the value of the ratio of Bank of America’s derivatives instruments to total derivatives market was also high. Collectively, the value of Bank of America’s securitization indicator increased sharply from 5.34 in 1996 to 15.18 in 1997. Since 2001, Bank of America has decreased their securitization activity. This is evidenced from the notional amount of securitization activity, in every securitization instruments. On the other hand, other banks are gearing up their securitization activities. This makes the market contribution of Bank of America gets smaller. Therefore, the value of securitization indicator for the Bank of America gets smaller and smaller at the end of the studied period. However, the fact that the readings of the indicator is volatile even among each bank over the years, does not distort the indicator’s ability in measuring banks’ securitization level involvement. Instead, the volatility of the index reflects the nature of banks’ securitization activity. It includes a broad range of instruments and it serves many purposes for banks. Therefore, banks’ securitization portfolio allocation differs in each time period, according to their needs. But given that this indicator includes market trend as well as the notional value of bank’s securitization activity deflated by total assets, it still allows for a cross-section comparison on the relative level of securitization involvement among studied group of banks. The third observation that stands out from the graph is banks that have relatively high readings in the earlier period seem to experience a sudden drop in the index readings starting from 2002. Most banks in these groups are foreign banks from developed nations, such as Deustche Bank, Citibank and ABN Ambro. On the other hand, local banks whose securitization indicator is more moderate before, in turn, experience an increase in their index readings. The explanation behind this trend is prior to 2002, local banks involvement in synthetic securitization, or derivative markets is very limited, or even in some banks, non-existence. However, the amount involved in synthetic securitization is large. Therefore, these instruments carry a higher weight. Since only big foreign banks are involved, their contribution is also very high. As a result, their indicator’s values are high. However, starting from 2002, the Malaysian banks have been encouraged to participate actively in derivatives market. With this, the share of each large foreign bank in derivatives instruments decreased tremendously. On the other hand, indicator for local banks which just started to participate in these instruments market also is subjected to the higher weight attributed to these instruments. Thus, this explains why foreign big banks experience a decrease in their indicator while some banks with relatively moderate readings before experiencing a surge in their securitization indicator. One of the objectives of creating this indicator is to distinguish between the active and passive players in the securitization market. In this case, the interest is to make a distinction between commercial banks that are active in securitization activity and those which are not. This study employs the mean as the benchmark to distinguish between active and passive banks. The indicators are constructed using annual data. Hence, each bank possesses different readings each year. The mean of the indicator of each bank is compared to the mean of the overall index, to dictate whether a bank is active or vice versa. Using this method, we find that there are 11 commercial banks in Malaysia that could be classified as actively involved in securitization activity (see Table 4), while the remaining 28 are relatively passive. The indicator claims commercial banks that are active, mostly are foreign-own banks from developed countries, such as Bank of America, Deustche Bank, ABN-Ambro, Citibank and Chase Manhattan; while commercial banks from developing economies such as Bangkok Bank and Bank of China is relatively passive. This is not surprising since securitization activity is already developed and is widely adopted by banks in the developed nations. This might explain the differences in securitization activity among those two groups of foreign-own commercial banks. Where local banks International Research Journal of Finance and Economics - Issue 24 (2009) 232 are concerned, the indicator points out local commercial banks that are relatively active are the big banks (total assets) which are Maybank, Affin and RHB. What this indicator depicts on securitization activity among the Malaysian commercial banks is consistent with Hugh (1999) observations that participation in securitization activities has its barrier of entry; therefore the extent of each bank’s participation in securitization market varies greatly. Dealing and trading securitization instruments as a bank profit centre requires a substantial investment in financial, intellectual, reputational and to a lesser extent, physical capital. Hence, securitization activities are restricted to a unique set of most efficient banks, and other banks would have little opportunity to get involved in this risk-management technique. Therefore, this study argues that this indicator is able to represent the level of a commercial bank’s securitization activity. Table 4: List of active and passive banks Passive Banks EON Bank Berhad Hong Leong Bank Berhad Public Bank Berhad Bank of Commerce Bhd Arab-Malaysian Bank Bhd Alliance Bank Malaysia Bhd United Overseas Bank (M) Bhd Bank of China (M) Bhd Southern Bank Berhad Overseas Chinese Banking Corp Ltd The Bank of Nova Scotia Bhd Bangkok Bank Berhad Bank Islam Malaysia Berhad Bank Muamalat Active Banks Perwira Affin Bank Malayan Banking Bhd RHB Bank Berhad HSBC Bank Malaysia Bhd Citibank Bhd Standard Chartered Bank Deutsche Bank (Malaysia) Bhd ABN AMRO Bank Berhad Bank of Amerika (M) Bhd Bank of Tokyo-Mitsubishi (M) Berhad The Chase Manhattan Bank Apart from that, statistical correlation between the indicator and several banks’ specific characteristics are conducted to check its consistency with theoretical predictions, (see Table 5). The difficulties in choosing bank characteristics is due to the fact that securitization activity could register a contrasting impact on banks, for example securitization is risk-reducing but following that, banks might be encourage to undertake riskier activities, resulting in an increase in risk. Similarly, banks that are managing their risk via securitization activities should be able to operate with lesser capital (Froot and Stein 1998), but there are banks that securitized to arbitrage capital, (regulatory capital arbitrage hypothesis, Jones 2000), hence will increase their capital with securitization. One possible choice would be to observe the correlation between securitization and earnings. Assuming other factors constant, a bank involvement in securitization should increase their earnings. Hence, the indicator should be positively related to measure of earnings, return on equity (ROE) and return on assets (ROA). This study observes that the measures are positively related to the indicator, where the correlations are 0.032313 and 0.125171, respectively. The correlation between the indicator and ROA is statistically significant but it is not so with ROE. Table 5: Securitization indicator and bank specific characteristics Pearson correlation test Statistical Correlation 0.1251 0.0323 -0.1979 -0.2742 0.1104 (95% confidence level) Variables Indicator-ROA Indicator-ROE Indicator-Ratio of Loans to Assets Indicator-Risk Pearson Statistics Test Next, this study examines the correlation between securitization indicator and risk. Accordingly, securitization should decrease risk inherent in a bank’s portfolio. The increase in risk 233 International Research Journal of Finance and Economics - Issue 24 (2009) following securitization is only possible if there is an increase in commercial banks’ risky activity such as lending. Hence, this study expects the indicator to be negatively related with measures of risk if it is negatively correlated with measure of lending activity, total loans over total assets. We find that the correlation between the indicator and measures of lending is -0.19791 while with measure of risk (nonperforming loan/total assets) is -0.274255. Both correlations are statistically significant. Based on the fact active banks are foreign-own from developed economies where securitization market is more established and local banks which are considered as big banks, couple with the correlation between the indicator and banks’ specific characteristics that are consistent with theories, this study argues that the indicator could represent the level of a bank’s securitization activity. Hence, this study proposes this indicator as an alternative measure of a bank’s securitization activity to be adopted in empirical research. 6. Securitization Indicator and Notional Value of Securitization: A Comparison This study argues that since banks’ securitization activity involves a myriad of instruments, taking only the notional amount of the securitization activity could not give a good indication on the relative level of banks’ involvement in securitization activity. Instead, the measurement should take into consideration relative bank size, bank’s contribution in each securitization instruments, the variety of securitization instruments in a bank securitization portfolio and the securitization market trend, as captured by the market share of each securitization instruments. Ideally, a smaller bank should not be considered passive with its small notional amount of securitization involvement. Similarly, a bank that has a securitization portfolio with a very high notional amount, yet highly concentrated on specific instruments should not be labeled as more active than its counterpart that has a more diversified portfolio. This section presents a comparison analysis undertaken between the two methods of measuring banks’ level of securitization involvement. We randomly selected the statistics for the year 2003 to carry out the analysis. The level of banks involvement in each year is ranked, according to the indicator value. The higher the indicator value, the more active a bank is in its securitization activity. The rank is arrange in descending order, with the most active bank, the one with the highest reading assigned the first position. As for the notional value, the bank with the highest notional value is ranked first. Then, the two methods of assessing banks’ level of involvement are compared. Table 6 below presents the ranking of banks’ level of involvement in securitization activity using both methods. International Research Journal of Finance and Economics - Issue 24 (2009) Table 6: 234 Comparison between the conventional and proposed method of quantifying banks’ securitization involvement Alternative Method (Securitization Indicator) 16.04759 13.90676 10.61649 5.072053 4.484124 3.124584 2.592993 2.449063 1.845094 1.685174 1.575695 1.525553 0.702565 0.598013 0.535623 0.388189 0.307516 0.241633 0.170946 0.096251 0.045273 0.034321 0.032366 0.031859 0.027105 0.019929 0.017559 0.016991 0.009676 0.007887 0.002634 5.86E-06 Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Conventional Method (Total Notional Value) (RM million) 73597894 18616115 22870498 35542180 7429859 33108764 71055263 7895016 43775018 28329265 31446936 22930127 10179716 10631296 12249078 9153553 12569863 4122108 5221848 3147474 575039 2919689 624212 461307 1151646 393574 392356 1665919 339337 260982 203790 2297 Rank 1 10 9 4 17 5 2 16 3 7 6 8 14 13 12 15 11 19 18 20 25 21 24 26 23 27 28 22 29 30 31 32 Bank Citibank ABN Amro Deustche Affin ChaseManhattan Standard Chartered Maybank Affin (Islamic) BCB HSBC RHB UOB Southern Bank OCBC Hong Leong EON Bank Berhad Public Bank Ambank Alliance Maybank (Islamic) Southern (Islamic) Bank of Nova Scotia EON (Islamic) Bangkok Bank Bank Muamalat OCBC (Islamic) HSBC (Islamic) Bank Islam Ambank (Islamic) Bank Of China Public (Islamic) StdCtd (Islamic) It is obvious that the two measurements used give different indications on the level of banks’ involvement in securitization activity. If notional amount is the only factor that matters in measuring the extent of a bank securitization activity, the five banks that are considered most active would be Citibank, Maybank, Bank of Commerce, Affin Bank and Standard Chartered, while ABN Ambro, Deutsche Bank and Chase Manhattan is considered as passive players, due to the relatively lower notional amount of securitization transactions undertaken. Nonetheless, these three banks are much smaller in size (as measured by total assets). In addition, their securitization portfolios are more diverse (variety) and is more reflective of the market trend, in the sense the distribution of their notional amount of securitization follows the relative importance of securitization instruments in the aggregate securitization market. In other words, a larger proportion of their securitization transactions are allocated in the instruments with higher market share, (which is w, the weight in calculating the indicator). Thus, they should not be considered as passive relative to big banks with a high notional value of securitization activity, without referring to factors such as variety, market contribution and the distribution of the notional amount of securitization transactions. For instance, consider the case between Maybank and ABN Ambro. Why is Maybank, with the second largest amount in securitization transactions, is not considered as the second largest player by the indicator, while ABN Ambro, with a more modest total amount of value (ranked 10th place with the conventional method), ranked second? 235 International Research Journal of Finance and Economics - Issue 24 (2009) The statistics reveal that Maybank is a large bank (in size). Therefore, after discounting total assets to make a fair comparison in measuring banks’ involvement in securitization per se, the amount of its securitization activity is not so much. Furthermore, Maybank’s contribution in each securitization instruments is less than 20%. Its biggest contribution, 35% is in the obligations sectors (revolving underwriting facilities and underwriting agreement), and the market weight of this sector is very small, which is only 0.4. On the other hand, ABN Amro, is a small bank. Nevertheless, the notional value for its securitization activity by far exceeded its size. Moreover, a large amount of its securitization activity is concentrated in the derivatives sectors, which carry higher weight; while activities in other instruments is much less. In other words, the allocation of its total securitization amount to each securitization instruments follows the market share of each securitization instruments in the aggregate market. Therefore, with its relatively smaller size (total assets), yet a more market like distribution of securitization portfolio, ABN Ambro was ranked second according to the securitization indicator. 7. Conclusions The measurement of banks’ involvement in securitization activity should reflect both the quantity aspect as well as the variety aspects of securitization instruments. However, none of the conventional measure satisfies the criterions. To fill in the gap, this study proposes an alternative indicator, which is a composite measure that combines securitization instruments into a single index. This indicator is structured in such a way to accommodate both the quantity and variety aspects of securitization. The comparison analysis done between the indicator value and the conventional measure depicts that discounting total assets, taking into account factors such as each bank’s contribution in each securitization instruments and highlighting the varieties of securitization instruments, do yield a different indication on the level of each bank’s securitization involvement. As such, this indicator could be employed as a measure for banks’ securitization involvement in the subsequent study concerning how securitization affects bank lending and stability. International Research Journal of Finance and Economics - Issue 24 (2009) 236 References [1] [2] [3] Ambrose, B., Little, M. & Sanders, A. 2003. Does regulatory capital arbitrage or asymmetric information drives securitization? Financial Management Association Working Paper. BIS. 2003. The Basel consultative document 2003. Switzerland. Bank of International Settlement. (online) http: www.bis.com Belter, K., Engsted, T. & Tanggard, C. 2005. A new daily dividend-adjusted index for the Danish stock market, 1985-2002: Construction, statistical properties and return predictability. Research in international business and finance 19: 53-70. Benveniste, L.M. & Berger, A.N. 1987. Securitization with recourse: An instrument that offers uninsured bank depositors sequential claims. Journal of Banking and Finance 11: 403-424. Berger, A.N & Udell, G.F. 1993. Securitisation, risk, and the liquidity problem in banking. Chapter 4 in M Klausner and L White (eds.) Structural Change in Banking, pp 227-288. Homewood Il: Business One Irwin. Calomiris, C.W. & Wilson, B. 2004. Credit card securitization and regulatory arbitrage. Journal of Financial Services Research 26 (1): 5-27 Cebenoyan, A.S. & Strahan, P.E. 2004. Risk management, capital structure and lending at bank. Journal of Banking and Finance 28: 19-43. Cummings, C. 1987. The economics of securitization. Federal Reserve Bank of New York Quarterly Review. 12(3): 11-23. Demsetz, R. S. 2000. Bank loan sales: A new look at the motivations for secondary market activity. The Journal of Financial Research 23 (2): 197-222. Ergungor, M.O. 2003. Securitization. Economic Commentary - Federal Reserve Bank of Cleveland. Ernst & Young. 2002. Securitization transactions covered by the law (online) http://www.ey.com/global/download.nsf Froot, A. K. & Stein, C.J. 1998. Risk management, capital budgeting, and capital structure policy for financial institutions: an integrated approach. Journal of Financial Economics 47: 55-82. Fung, M.K. & Cheng, A.C.S. 2004. Diffusion of off-balance sheet financial innovations, information complementarity and market competition. Pacific Basin Finance Journal 12: 525540. Georges, D. & Harchaoui, T.M. 2003. Banks’ capital, securitization and credit risk: An empirical evidence for Canada. Bank of Canada Working Paper. 03-11. Greenbaum, S. I. & Thakor, A. V. 1987. Bank funding modes: Securitization versus deposits. Journal of Banking and Finance 11: 379-401. Greenbaum, S.I., Kanatas, G. & Venezia, I. 1991. Loan commitments and the management of uncertain demand. Journal of Real Estate Finance and Economics (4): 351-366. Hugh, T. 1999. A preliminary look at gains from asset securitization. Journal of International Financial Markets, Institutions and Money 9(3): 321-333. James, C. 1987. Off-balance sheet banking. Economic Review-Federal Reserve Bank of San Francisco 4: 21-36. Jones, D. 2000. Emerging problems with the Basel Capital Accord: Regulatory capital arbitrage and related issues. Journal of Banking and Finance 24: 35-58. Murray, A. P. 2001. Has securitization increased risk to the financial system? Business Economics 63. Roever, W. A. & Fabozzi, F. J. 2003. A primer on securitization. Journal of Structured and Project Finance 9 (2): 5-20 Wolfe, S. 2000. Structural effects of asset-backed securitization. The European Journal of Finance 6: 353-369. (online) [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22]