Measurement of Liquidity Risk in Keshavarzi Bank (With Value at Risk Approach)

One of the most critical competitive fields of companies and financial institutions and banks to provide optimal financial sources to survive in the turbulent business environment is financial decision making. It`s safe to say any decision should be taken to consider the risks and dangers associated with it. Of course specialized risk management is also one of the most important and fundamental fields that need financial institutions adaptation with new conditions to understand increasing complexity of the rules, technology and customers and it can act better. To make this issue happen, the risk measurement would be the banks and institutions today challenge. In this study the liquidity risk of one second rate branch of Keshavarzi Bank during 2008 to 2012years and with usage of the bank's financial statements, including a variety of deposit accounts, savings, revenue, cost and convenience, have been calculated. Writer hypothesis that significant of trends in the data research (liquidity risk) has been over these years. Should mention to demonstrate this method the value at risk has been used. The reasons for using this method are low cost and fast pace. The results show that this hypothesis along 2008, 2009, 2011 and 2012 years will be rejected and over the years 2008and 2009 are accepted. Generally liquidity risk trend towards an equilibrium (to reach a certain point). Finally, it is proposed that the risk would be tested for other branches as well in varying degrees.


Introduction:
Financial institutions are very important in the economic structure of society.One of the requirements of these institutions to meet the challenges successfully is flexibility in facing different changes in the monetary and financial markets.Professional management of the various risks is the main areas that require adaptation of financial institutions to new conditions in order to understand increasing complexity of the rules and regulations, customers and the technology and manage their activities effectively and efficiently.Any institution that fails to adapt to these changes and complexities would face with a situation that will ultimately threaten their life.Nature of trade and investment activities is in such a way that business efficiency needs risk tolerance.Risk management is a process in which managers identify, measure, decision make and monitor entity risks that posed by the firms.Liquidity problems for a bank besides it has a huge loss for bank customers, it can quickly spread to other credit institutions and cause the financial system collapse of a country.Considering the importance and effect of this issue, tools such as deposit insurance ,legal reserve with the Central Bank and access to the Central Banks liquidity has been created to manage bank liquidity.Methods of liquidity management faced major changes over time.In the past, liquidity management usually interpreted as asset management, but with the development of financial markets and particularly interbank markets, liquidity management spread under obligations management.This means that banks to provide liquidity from market borrows from banking market or capital markets and create its commitments.Usage of obligations management should be considered in liquidity management according to the limitations that in Islam exist on borrowing transactions and in the Islamic banking system.In the present study has been reviewed one of the method of calculating liquidity risk.
• First question: is it possible to measure liquidity risk using value at risk theory with internal bank data (last few days liquidity)?
• Second question :how are liquidity risk changes in the branch?Is this a significant trend?
The importance of research can be expressed as follows: A) Risk management in banks is important, and supervising regulations specially provisions of the Central Bank and Basel guidance's should be taken into consideration.B) Liquidity risk produced because of lack and uncertainty of bank liquidity.Sufficient cash holdings makes payment obligations and the liquidity needs of depositors would be answered in a good time and the sanctity of the inadequate funds provide impaired banking system that may even lead to bankruptcy.C) Recent changes in financial markets caused that banking and payment system would be very interested in short-term predictions, so financial flow control system should be taken that be able to measure the performance of liquidity risk properly and constantly.D) Providing fast and correct information of liquidity risk timely leads risk managers and financial managers to be able of best performance and optimize the use of capital and maximize the value of shareholders' assets as its main target raised by using appropriate measures and strategies in a timely manner(Sina Bank Economic Journal,2008)The main objective of the present study is measuring the liquidity risk in a second rate branches of Keshavarzi Bank and generalizes this result to other branches.Also reviewing the significant and justifiable data (liquidity risk) and its estimation on liquidity risk based on one, two and three period came before is the more detailed objectives of this research.
The key variable of this research is bank liquidity rate that involves physical inventory, ATM system inventory, transport documents, paralleled documents, and people deposits, granted facilities, revenue and...The main variable is the rate of bank liquidity that involves physical inventory, ATM system inventory, transport documents, paralleled documents, and people deposits, granted facilities, revenue and... Risk: In a banker`s view the risk means uncertainty in the relationship with an event.

Definitions of words are discussed below:
Risk Management: is the process by which an organization or investors react to an optimal manner against a variety of risks.
Liquidity Risk: is the risk arising from the lack of necessary liquidity to cover short-term liabilities and fund unexpected outputs.
Value at Risk: In finance and economics literature, value at risk is the maximum loss over a period of time at a specified confidence level with a probability.(Bandeh, 2012).
Expressions are presented in this section are the main definitions of this part VaR (Value at Risk): -Value at Risk (VaR) is the method to assess the risk that the standard statistical techniques routinely used in other technical fields.In contract, valued at risk measure most loss expected in a specified horizon time at given confidence level.For example, a bank may announce the daily purchase and sale of portfolio value at risk bank confidence level of 99%, as 35 million dollars.In other words, only one out of 100 daily trading losses may occur more than 35 million dollars.This single number (the order of Value at Risk) briefly show the bank's exposure to risk market.As well as value at risk, risk is measured in terms of dollars.
Unlike traditional value at risk measurement, risk exponentially comprehensive overall measure of portfolio risk that the assets and related liabilities and current status are used.As a result, the value at risk is really risk forward looking assessment.Value at risk not only of all banks branches but also is effective for all financial instruments.In addition, the value at risk methodology risk, market risk can be generalized to other types of financial risks.
Worth the risk, that can measure tangible and accessible overview of the set risks of future development.Limitation of this measure is its cause of static, dynamic benchmark for researchers pursuing it.(MohsenRoosta,2011) -Value at risk measurement and risk analysis is a framework for the types of assets that can be equally applied.As a result, the portfolio consists of bonds with a portfolio composed of stocks would be comparable.Well worth the risk to investors offers about the nature and types of risk insights.Thus, using the concept of risk management, asset allocation allows different methods to find efficient evaluated portfolios.The overall risk of the portfolio components fixes analysis of control.Risk allocation and risk budgeting leads recent cases eventually.Allocation causes the portfolio managers of risk capital and to the sectors that have the potential for greater efficiency to conduct activities and VaR assets using standard according to diversify risk reduction strategies.-Willing and Kalkberner aimed liquidity risk management and risk-free rate fluctuations debt maturities:

International Journal of Academic Research in Business and
Based on this model, to measure liquidity risk-free debt maturities from cash flows have used the concept of temporal structure.Using the concept of using time series, liquidity can be predicted for the future.-Maryam Shabani Motlagh research liquidity risk in the banking industry with using an Imperious Landa Index: This research aim introduces an approach to measure liquidity risk using Ianda Index parameters as well as the best conditions for accurate predictions of daily cash.Case is considered a branch of the National Bank.Objective measurement is Landa and the cumulative distribution function of variables such as cash, average and net standard deviation of branch operation is used.Research shows that the most economical mode of historical information relating to past3 or 4 days and 5 and 4 days anticipated future liquidity situation of banks with 100% confidence.Values of cut are according 2.4 to 2.6.Therefore reduction of Landa to less than 2.4 of branch it would face with a shortage of cash and it needs to get money.

Research Method:
Using the information of this data (net liquidity) deals with to measure liquidity risk that the fluctuations of risk are shown in 2008 to 2012 years.Following explain the formula is applied.Since the value at risk expected for worst loss, the most accurate method is the use of portfolios distribution.Assume that FΔP is the probability density of (Pdf) function from ΔP and C is confidence interval, so in this case the horizontal value at risk is calculated by the following formula: For a given portfolio, if the return on the portfolio normally to be distributed with μ mean and standard deviation σ the value at risk would obtain from the mentioned procedure.From the normal standard table for a given portfolio, the α number are corresponding with C confidence level.For example if C is 95% the corresponding α would be 1.65 and if C is 99%, α would be 2.33.As the left trail is the corresponding value at risk, the actual cut is -α, as you see in 1-4 diagrams.

Diagram1-4:normal standard curve with 99% accuracy
To obtain the value at risk, it is a standard conversion can be done as follows: Then we can write: VaR = σα -μ (Formula 4-1) Hence if FΔP(x) is cumulative distribution function (cdf) for ΔP the equation will be written as follows: So, as seen in this study, the formula (4.1) is used to measure the value at risk in the following sections that will describe each of these formulas in below.(Dai Bo,2001) In order to obtain the liquidity risk of the end of each month the amount of liquidity that is needed here is the information collected from the bank daily that the sum of these data and eventually 61 data is available to obtain a figure for each month .σindicate the standard deviation.Then 12 standard deviation first data will obtain in the next phase standard deviation of 12 other data is obtained by excluding the first data and so data from 12 to 12 and minus one standard deviation from the baseline values can be calculated.α is considered the confidence level, where here considered 99% that the corresponding Z of this confidence level is 2.33.The μ indicator is data average.Average data is computed by exactly the same standard deviation, means it would takes first 12 data then the average the next 12 data minus first dada is calculated and the same way until the end of the 61 data "Is computed similarly to the end of the 61 are calculated.Then, as can be seen in the formula, the obtained individual deviation, multiplied the level of confidence that is 2.33.Finally to obtain the value at risk should minus each average of obtained numbers from multiplied standard deviation to 2.33.The value at risk or liquidity risk shows for each month.Then after calculation of net amount of liquidity can be provided to the first hypothesis that was originally intended the trend of this risk will estimate in second rate Keshavarzi Bank risk.Then the H0 hypothesis defines as bellow:

H1= trend is statistically significant in the data
Before we review the model reliability or stability must be tested.In fact the aim of this is evaluation of false or true estimation.In time series data, the unit root test should be carried out to determine the reliability or stability time series variables.Following two assumptions are considered:

H0 = Considered time series, is steady H1 = Considered time series, is not steady
As mentioned, the unit root tests on liquidity risk are considered that the results are in table (1)(2)(3)(4): model.In other words, these figures show the considered number is static, so there is no need to take the first-order difference.Actually the Dickey Fuller test, tells the regression is fake or real.So in the test H0 assumption be accepted.
As mentioned, to get the amount of liquidity risk, the liquidity of one of the Keshavarzi Bank branches has been used.Then standard deviation and mean liquidity has been calculated and with 99% of confidence in (4-1) formula has been placed.Accordingly, liquidity risk was obtained for2008 to2012years.After obtaining the amount of risk, the statistical significance of the data evaluation is discussed and the trend of risk in Keshavarzi Bank will be considered.To reach above aim the (2-4) formula was used.
Formula (2-4) In formula (2-4) autoregressive process the third order can be seen, in fact the whole formula is given as a lump sum, which can be divided into smaller parts and thus the results were analyzed separately for each phase.In fact, it can be stated that the level of liquidity risk over a period of itself (The same as a month earlier) and in next steps and in the two or three previous period it's estimated.It means at first, the estimation of the first order, second order approximation in the second step, the third-order approximation in the third step and finally estimation of the first and the second and third will be done at the same time that further each of these steps will be explained.

First Sate:
In the first state, the liquidity risk is estimated that over a previous period that for a first-order estimate of the formula the Eviews software will be used: VaR = α + β VaR(−1) (Formula 3-4) As you seen in table (2)(3)(4), this equation can be justified in terms of significance.Because its possibility is less than 5%, no matter how small the amount is likely to be close to zero, it indicates that the model fit is correct and the explanation is for a correct model.Because here, the likelihood is zero, then it is clear that the model fit well and is reliable.
The coefficient for liquidity risk has obtained is a positive acceptable value that shows 89% of liquidity risk at the present time follows the amount of the risk in an earlier time and previous period affects effects this period results.Indeed, the monthly data on an annual trend is observed.
The obtained t (13.7) is also acceptable, in fact as long as this stat amount is more than 2 the hypothesis will be considered and acceptable.
The R (coefficient of determination) obtained also shows that 80% of the variability (liquidity risk at the present time) by the independent variable (liquidity risk in a prior period) is described and the accuracy of the model is sufficient to justify this percentage.
Here the Watson camera statistic is 2.192 that shows a disturbing statements are not correlated to each other and doesn`t have correlation .The amount of F statistics also shows an acceptable value.Level of accuracy statistic shows whole model correction.

Second State:
The liquidity risk is estimated in two periods that it came before it, that the formula (4-4) is shown: (Formula 4-4) Again ,the variables according to the formulas (4-4) were placed on the software and the results can be seen in Table 4-3: Test indicates that the probability is zero, the model is fitted correctly.Than the other 82% of the value of risk changes the course of his two terms is related to changes in risk.
Greater precision can be seen that the value of R2 (the coefficient of determination) is 68%, this amount is not enough to justify the accuracy of the model.So, this model is rejected by the bank's point of view and it`s not reliable, because the bank shall determine the coefficient is above 70 to be acceptable.Well as the camera parameters Watson should be about 2 to acceptable results as you can see the value of this statistic is 1.26 which returns the desired value does not justify.
The results show that the probability that liquidity risk is to follow the course of his two terms as before is weak; In fact this result is obtained in the second stage to estimate the data statistically justified are not significant.

Third state:
Liquidity risk in the next state over three rounds before his time is estimated that in the formula (4-5) is evident (Formula 5-4)

Fourth State:
In this section we estimate the effect on liquidity risk is shown three times and in fact the liquidity risks over a period of two or three rounds before his term are estimated.Then (Formula 6-4) According to Table 4-5 to estimate the probability of the first order is zero and that indicates that the estimated model is the first order of fit goodness and the model explanation is correct.However, estimates of the probability of the second and third values are 0.98 and 0.34, both values are greater than 0.05 and the assumptions there, if you are having a good fit to estimate the probability that it is less and from 0.05 to zero as possible.
Risk ratio for a prior period has been obtained, indicating that the risk of each period, 77 percent of the risk is related to its previous period.At Second case, the second order approximation to a negative value(-0.0030)indicates that this is indicative estimates are negatively estimated time before is related.This suggests that these risks cannot be calculated using two before cycles.In continue is seen in triple estimation that 14% of time the liquidity risk explains at present time.T base for first-order estimation is 5.04 and for the second order approximation, is -0.015 for the third-order approximation, is 0.95, while the base t amount should be more than 2 to the considered hypothesis be acceptable.Due to the amount of three times the estimate, the total numbers that were obtained for the first stage estimates were more than justified.That is, assuming H0 is rejected in a first-order estimate, but the estimate of the second and third hypothesis H1, which represents a significant trend in the data, is rejected.Because only a first order estimate plausible explanation is that liquidity is going on and could just days before his Finally, "months before its to follow.Follows the plot of the monthly data is shown as well as the liquidity risk.Source: research findings As can be seen in Figure 2-4, the monthly changes of liquidity, suggests that liquidity over time, almost a natural process but the trends are sinusoidal shape.Because they may one day liquidity to banks for various reasons, including an increase in deposits, checks are divesting.Increases other reasons, such as depositing and withdrawing subsidies and reduced liquidity shortage faced by bank customers.Required to process the data in three months has been considered.
Social Sciences February 2014, Vol. 4, No. 2 ISSN: 2222-6990 77 IJARBSS -Impact Factor: 0.305 (Allocated by Global Impact Factor, Australia) www.hrmars.comInfollowing research is summarized that is done inside and outside Iran:-Abrahasion and Abbott research on the bank's balance sheet value at risk (2000):Hu,Abrahasion and Abbott research about value in the balance sheet is one of America's banks.These values are exposed in his article titled risk analysis and determination of sample items in the balance sheet of a bank's value calculated in banks are involved, and how to manage the information on value at risk calculation are described in the decision.Classification on the balance sheet value at risk based on a variety of rates available from banks in the banking industry is done.(Berkotiz,2000) model is not sufficient to justify the amount again.The other side of the camera base Watson is not even close to that shows disturbing sentences are interdependent and this causes the error.

Figure 2 - 4 :
Figure 2-4: the cash monthly data over 2008 years from 2012Source: research findings As can be seen in Figure2-4, the monthly changes of liquidity, suggests that liquidity over time, almost a natural process but the trends are sinusoidal shape.Because they may one day liquidity to banks for various reasons, including an increase in deposits, checks are divesting.Increases other reasons, such as depositing and withdrawing subsidies and reduced liquidity shortage faced by bank customers.Required to process the data in three months has been considered.

Table 1 -
Watson camera statistic level that has the value of 1.94 is more than the critical values of the test that it means -3.58 and -2.92and -2.60 and this represents a disturbing statement is dynamic or static.On the other hand, the absolute value of the number -5.14 is higher than the critical values for the test that his demonstrates the reliability of the

Table 4 -
4 : estimate the liquidity risk resulting third orderAs you see in 4-4 table, in this case, as in the previous case, the risk is estimated over three periods before himself.Test indicates that the probability is zero, the model is fitted correctly.On the other hand, 78% of the value of risk changes the risk of change is dependent on previous periods.However, it is seen that the coefficient of determination R2 is 61%.The accuracy of the

Table 5 -
4 : thus estimate the first and the second and third liquidity risk