Fair Value in Banks and Volatility in its Stock Returns

An Expert's View about Banking in India

Posted on: 20 Jul 2010

Relationship between fair valuation in banks' investment portfolio and volatility in Indian Banks' Stock Returns

BOTOH KES R AE NLADT IVOONLSAHTIPIL IBTEYT WINE E SNCH O AFNRAT EIRE X P TVRA ICLEU ERSE TINU R BNASN IKNS ? T THREA DININDGIA N DDRe.U p TnaA B y irvStNmeeUrnsePiinoAtytr C OOLHef AKRABORTY f c CCtuoarmer l merce cutta ADDR ESS : FLAT NO. ? 12 2JA3D, A CVEPNUTRRAL ROAD P H O N E : ((MR)) 928431021-47957635 3 E-MAIL: tanupachakraborty@gmail.com KINODLIKAA TA ? 700 032 tanupa_cu@hotmail.com ATPcohrcpaiksycn rtiroicgeweshselet?adrcghe mpaepnet r is originally published in ?The IUP o ,f tVhoisl .a rItXic,l e Nlieoss .w 1it h& I U2P , PJuabnluicaartyi o&n. A pril 201J0o u[Rrneafl. #o f 0A9c1c-2o0u1n0ti-n0g1 /R04es-0ea4r-0ch1 ] &an Ad utdhiet * he T facts and f organized by Iinnddingsi toyf this paper have been presented at 9th International Accounting Conference Delloitte at Scienciaen C Ac,c Kouonlktiantga iAn sJsaoncuiaartiyo, n2 0 R0e9s aenardc hth eF cooupnydraitgihotn r (eIsAtsA wRiFth) t hine aasustohcoira.t i on with pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! BTohokes Aend Volatilitye tIwne eSnhCa 1 R lationship B o rFnea i rP rVicael u eRse t uInrn sB a Innk s T? h Ter aIdndinigan A tBSTR ext ?lFair ACT value? of an asset or a liability refers to the amount at whicgwiraobwiliinty h such asset could be exchanged, or the g settled, between knowledgeable, willing parties in an arm?s length transaction. Although the irrelevance of historical cost based accounting numbers in the financial statedeabkae nd ments, in the of developments in financial markets a advancements in technology, has triggered off the afpplitcea otino nfair value accounting a decade and a half ago, some issues still stand in the way of extensive offi1inrman?csial fair valiuned uaccecdounting framework. One such issue is the excessive level of volatility in the statements by fair valuations and its resultant impact on flight of capital from the equity. In accordance withb9aan9k5 the series of guidelines issued by the Reserve Bank of India between s ?? 2000, fair value accounting has been applied only on the ?held for trading? securities in the investment portfolio in India till today. Accordingibnttempt ly, this researcho upt aper makes a modest to examine whether fair valuations in banks? trading books bring ab an increased volatility banks? stock returns over the time period 1994 ? 1995 to 2007 ? 2008 using a sample teacnhkns of Indian i quanesd bank index i.e. BSE BANKEX, and autoregressive and multiple linear regression . pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! Fn n 2 I troductio sgt i I o anvk a ee n rh c no imaldl eesnrtstatements, the primary output iitny an accounting information system, enables a firm?s , whether acting in the capac of investors, creditors, lenders, suppliers, customers, or members of the society in general to focus the lens on the state of affairs of a bu bt siness. smhy is widely acknowledged in both academic literature and regulatory debate that the discipline enforced generally accepted accounting principles t a onuald in modeling, preparing and presenting financial statements ge mheelp in reducing the information asymmetry between the firm?s decision makers i.e. she nt and the stakeholders with a view to facilitating a timely and proper decision making by stakeholders. But in today?s business environment marked by in ceao rnvsiicdees novations in financial products and r eodffered by financial institutions, development of secondary markets for instruments that were illiqui as d and non-tradable until recently and increasing use of risk transfer mechanisms such securitisation and derivatives contracts, historical cost accounting bci ri rdogss model that currently prevails in gthe world is often criticized on the ground that it is increasingly becoming inadequate in the information gap cs between the firm?s management and its stakeholders. The primary reason that as assets, in historical cost based financial statements, are usually carried fto ornsideration at the amount of paid for its acquisition while liabilities are recorded at the amount received in exchange the obligation, they fail to r ro eflect the current values of assets and liabilities and hence are irrelevant its users i.e. the firm?s stakeholders. Further, historical cost based accoun its ecognize alized ting system does not appreciation in the value of an asset unless re whereas deterioration in the asset quality normally taken into accoun ahsoseutgh t by way of creating provisions and charge-offs against ientecroimorea tsitoantement, such charges are not aligned with the year-to-year change in the degree of d of the as perceived by the market. Besides, historical cost financial statements keep innovative financial pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! p sfip ro na d arnk u ce c id ts such as derivatives off-balance sheet. Such criticisms of historical cost based statements hav3e a loff debate on ?fair value accounting? with a view to improving the information content of statements. Atwhs defined by International Accounting Standards Board (IASB), ?fairsbei e value? of an asset or a liability is el kl aisn mgount feosr which such asset could be exchanged, or the liability settled, between knowledgeable, parti in an arm?s length transaction. Implicit in this definition is the notion that fair value to capture the value of an asset or a liability to an aetween y entity by way of estimating its exchange price two unrelated parties even in the absence of an actual transaction or an active market for thatrvme s af s le e ur t ee nocr eliability. Accordingly, quoted markeatl uperice for an identical asset or a liability in an active market is the best evidence of fair v and hence is used as the basis of valuation under fair accounting. However, where active markets for assets or liabilities are unavailable, q bewe atrwkeeet uoted nprice of similar or related asset or liability, adjusted objectively to account for the differences the traded instrument and the v msatimate alued item?s specific characteristics, is regarded as a fair value . If the fair value of an asset or a liability cannot be determined in any on vmao y lud s e of the above two ee ,l sthe firm may make use of an appertocpriate valuation model, such as binomial option pricing or discounted cash flow method ., to arrive at an estimate of fair value. Thus, under fair accounting, all assets and liabilities are to be carried on the balance sheet at th vaelauseured scussed eir fair values in accordance with any one of the three bases as di above, and any change in fair between beginning and end period to be recorded in income statement, whether realized or not. Ath emajor advantage with fair value accounting is that it permits a tdimeteelryio rreactioognnition of any change in quality of an asset or a liability, whether enhancement or , by bringing about a pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! r tah ev cc ealuation of the item whenever there is a change in its market price or, in the absence of a market, i4n c opurenstienngt value of the stream of benefits expected to be generated by the item. In contrast, the current shows little or no impact on the provisions when the asset value is revised upward. Even in tph aaste t of depreciation in asset value, current accounting may encourage managemen in creating buffers would shield the balance sheet from revealing such det too rovisions account erioration because a significant increase in and charge-offs to take of a substantial asset devaluation may come as a big surprise the stak t ff eholders and drive them to overreact to such news. As for example, a bank may decide to sell its investments, when the value of its investment portfolio has vht a e appreciated significantly, and utilize l uperofits, so earned, to set off and limit the losses caused by deterioration in its loan portfolio. In fair accou rhe nting, such a strategy cannot be adopted because the profits resulting from appreciation of investment portfolio are recognized in income statement much earl isn atshtreurm ier at the time of its occurrence theannts at the time of disposal of investments. Furthermore, the problem of valuation of derivatives under hist sheet orical cost based accounting system has kept such instruments off the balance . An increasing reliance on fair value measurement is expected to pave the way otpaateqmueennetss for financial s properly accounting for derivatives transactions, thereby amending a major element of in the present accounting and reporting system. Dof fespite the advantage of providing more relevant information to the stakeholders in comparison to that current historical cost based accounting, thlmair ry ere are still some issues which have made the debate on value accounting ve contentious. When market values are not observable, fair values of assets or miaebetiahlsoitudieresms are closely approximated by the use of valuation models. Since a wide variety of valuation aernet available, with each having itso nown set off assumptions, estimates, data inputs and methodology, and as the valuati done for a particular purpose at a particular date may pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! n hodiae gtteh b rlmye appropriate for other purposes or for other dates, all non-market based fair values tend to b5e isnuebjective and hence unreliable. Not only are fair value measures of illiquid assets difficult to , verify and audit, an indiscriminate use of valuation models across balance-sheet items and vcaAla r tl ouses subjectivity firms and countries may also make them less comparable. Moreover, such in fair measurements provides the management with an opportu s g ear nity to manipulate the bottom-line and i nthe various financial ratios in order to paint a rosy picture of the state of affairs of the business. , mar statement king down a firm?s financial liabilities to fair values and recording it as a gain in the income in the event of the firm?s own credit deteriorationooawlvnent sues may result in an insolvent firm appear . Even if the above-mentioned is are resolved by disallowing the fair valuation of the firm?s outstandi fnd ng liabilities, and reaching a consensus as regards the use of valuation models asset-wise country-wise, the apprehension that fair value will bring an ex irn in epv a oe nscrttio ia nr l cessive level of volatility to the gs statemenptrse atnd hence make the statements more difficult for the stakeholders, especially the , to inter may still stand in the way of full endorsement of the fair value accounting and . Itt is primarily argued that today?s business environment, maheoecnhcneology ions rked by rapid price changes and shorter cycles, requires fair valuat to change frequently in line with such developments and this may ind f f uce considerable volatility in income and capital measures. The reason why the issue volatility in fair value measurement approach is of utmost concern to fifwi r on mans regulators, standard setters and c, iaanl d warrants a separate empirical examination is the impact that such increased volatility of statements mig oruumlds ht have on the existence of the firm altogether. It is often debated in academic that such increased volatility in accounting magnreitcutdlyes brought about by fair value accounting not necessarily be a problem if shareholders cor interpret such volatility and its possible pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! s poaa uprecres and do not panic or get ecstatic and trade in the stock market. So, the research question that thi6s intends to empirically examine is whether v vbout s olatility in fair value based financial statements bring volatility in firm? share price returns as well. There are two schools of thought on this issue of ssh oalaretility. Those who argue that volatility in fair value accounting numbers also cause volatility in price returns contend that shareholders may not properly otop aitneimonents nce interpret the volatility in financial induced by fair valuations and he trade frequently owing to violent fluctuations of their thereby resulti immpinion ng into volatility in share price returns. However, scholars holding a contrary argue that the increased volatility in fair value based financial statements tha p erirk acet would have no t on shareholders? perceptions as insftiatiurtional investors and individual investors, with the help of analysts, always try to extrapolate valuations from a variety of information sources to base investment decisions even in the absence of fair value based numbers in the financial statements. Tiasislul eddate, the areas where fair value disclosure, measurement and recognition standards have been by the standard setting bodies internationally primarilypmnro ei d relate to financial instruments like debt n acesiqupuraielmty securities, derivatives contracts, loans, deposits and employees stock options, which are the ceonmt ponents of a bank?s balance sheet. Besides, the concept of fair value has been used in the of goodwill, intangible assets and other non current assets held Ian f for sale within the realm vn s ?dt iimtupteairment of assets?. Indiaan itss not an exception in this regard. Although it is only recently that the of Chartered Account of India (ICAI) has come up with two new accounting standards 30 31 dealing with recognition, measurement and presentation of financial instruments based on baine l te u en es rve fair (effective for all forms of companies from 1st April, 2009), the Rese Bank of India (RBI) has rn daetimonaanlding fair valuation of Indeieann banks? investment portfolio since 2000-2001. Like in most countries, the RBI has b requiring Indian banks to hold securities in their investment pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! p ?-I onrvtfeosltimo enutn?der two or three broad categories ? ?Trading? ora il?aCbluerrent? investment category7, or ?Held to maturity? or ?Permanent? category, and ?Av for sale? category (optional) d ai dsecploesnudrieng on the intendeodr period and purpose of holding the securities. The RBI mandates a monthly of fair values f securities held for trading purposes as these are to be marked to market on s monthly basis and any change in their fair values should be adjusted as profit or loshtaoaltdeimngent y s in the income . The ?permanent? category comprises securities acquired b the banks with the intention of them till maturity and hence are not to b acsae mqcouurirtsiiitztieaiostin e marked to market but should be valued year end at o ncost less any amount written off in income statement to reflect diminution in value or of premium amount paid over face value. ?Available for sale? category includes those which are neither traded frequently nor held till maturity but can be so tqh netiicr ipation ould ld in response to of changes in interest rates etc. Such securities sh be carried in the balance sheet at fair values estimated by drawing refe (sRo u tan o tl t ey am)tions ive rence to the best available source of current market or other data relat to current values issued by regulatory agencies, and any depreciation einnt their fair values are to be recognized year-end by way of adjustments between income and investment fluctuation reserve account. However, unlike in interna ocf BI f tional countries, the has prescribed the percentage holdings in each category o a bank?s investment portfolio. By way issuing a series of guidelines on classi ?thraandged fication and valuation of investments, the RBI has gradually the percentage pattern of holdings from 70% in ?permanent? category and 30% in ?current? or irnves itnmg?e nctsategory in 1992-1993 to limiting ?held to maturity? i.e. permanent portion of total to only 25% in 2000-2001. Thus, the above discussion suggests th tih es ne earch at at present, the on volatility in fair value accounting can be done with reference to banking sector alone, given te rnstaatniodnaardllsy or regulations otino nefadir value accounting and reporting prevailing nationally and . The above-men accounting treatment indicates that the change in valuation of pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! s cesvi ucgr u nre r ifn it it ies held in the Indian banks? trading book (i.e. ?current? or ?trading? investment category) from8 c aanctcloyunting to fair value accounting must have increased volatility in banks? income and capital and this, in turn, casts a doubt on the shareholders abil fmooeclauatssiuelirsty equity ity to correctly interpret such and make no impact on banks? return volatility. Accordingly, this research paper e mone nItnsdian banks? trading books to examine empirically the volatility effect of fair value on banks? share price returns. Thhe paper revf ypothesis iews the available literature on the research subject and outlines the objective and of the study. It also presents the sample selecLindings owed tion and research methodology, enumerates the of the study foll by conclusion. Imnitera ture review ternationally, several researc r easurement h studies have argued the case for conceptual desirability of fair value , notably on the ?value relevance? dimension, and explored eswetalenvdaanrdce closed mpirically the value of investment securities? fair values dis by the banks after issuance of an accounting mandating s e ith uch recognition by way of examining the incremental association of such a change banks? share prices or share returns after controlling fo vxalpulaoratory stified r other market information. In his study, Hitz (2007) has ju the decision relevance of fair value measurement from both tihtn e tion and information perspectives, but mhaesn tcontended that the theoretical reasoning that underlies standard setters? and regulators? endorse of fair value measurement is inadequate and restricted its validity and applicability. According to Reuters (2008), proponents of fair v ahragt alue accounting assert u eit reveals ecoantoilmityic realities which were hidden by previous accounting methods while its critics that the vol in financial information induced by fair value accounting reduces stock prices pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! aE Dnxdc htaankgees the investors by surprise. In the light of comments made by the chiefs of Securities an9d Commission (SEC), Federal Deposit Insurancei(nui vfef Corporation, and financial advisory firm s t&m ePnhtelps in a oroundtable discussion organized by SEC on mark-to-market bookkeeping of firms? portfoli , Barr (2008) has opined that, though some changes are needed, mark-to-market .e. fair value) accounting does a fair job of assessing a company?s financial health. Afiemwong the empirical research on book versus fair value accounting of banks? investment portfolio, a noted ones are mentioned below. Using a samplfneovi sr estment re e of US banks, Barth (1994) has found that securities? fair values a incrementally associated with bank share prices after controlling s einntviaelsltyment securities? book values. On the contrary, Barth, Landsman, and Wahlen (1995), using the same data base, have found that such incremental volati rs lity in fair value-based numbers se ae lnceuv o ra t itn r ic e ee flected in bank share prices. Again, using similar approaches to assess the incremental value s of fair values of principal categories of US banks? assets and liabilities such as investment , loans, deposits and long term debt, Barth, Beaver, and Landsman ( ian n an cd 1996), Eccher, Ramesh, dr eTmheinatgaallryajan (1996) and Nelson (1i9r 96) have found that investment securities fair values are informative relative to the book values in explaining bank share prices. Aboody, Barth, Kasznik (1999) have examined the association between asset revaluations for financial, tangi anfn d ble, d intangible assets and share prices based on a sample of Australian firms between 1991-1995 period obtained evidence of a consistently significant oieb rmses association for financial as well as non-financial r.v eInd an interesting study conducted on Danish banks, Bernard, Merton, and Palepu (1995) have t faqiurity hat fair value based net equity book values are more reliable estimates of market values of than historical cost based book valutmese notf equity. In another study on the effect of book versus value accounting on a bank?s (re)inves behaviour and investment value, Burkhardt and pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! S aastrossy a emutsmsze (t2ri0c06) have shown that, as fair value accounting enhances disclosure and reduces the degree 1o0f information between the management and stakeholders, it increases the liquidity of bank?s and hence raises the bank?s investment opportunities leading to riskier tih v In eerall ducted investments and a lower value of the bank. By far, the study con by the European Central Bank (2004) to examine d iicmesp?act of fair value accounting on the volatility of individual banks? equity returns and bank equity returns selected from five countries comprising th iGntaly gdom e European Union (EU) ? France, Germany, , Spain and the United Kin (UK) ? is most significant. The study has shown that the change i e rvmolaantyility of returns is not statistically significant in any case excepting for one bank each in (out of five banks), Spain (out of three index banks), and UK (out of four banks) and for UK bank out of five different nations? bank indices. Thus, a review of international research stu ancdciocuantetis dies n gthat mixed results have beefonl ioobtained while testing the statistical significance of fair value in banks? investment port on bank?s stock returns. Iinnv Iensdtoiars,? however, no published research work relating to the impact of fair value accounting on behaviour appear as yet thereby creating a gap in Indian literature on the said subject. OInbjective, Hypothesis and Significance of the Study view of the research gap identified in the preceding section, the analysis conducted in this article rmaesaeya rbche seen as a first attempt to examine the complex issue of ?volatility? that requires significant and careful consideration before replacing fully the current historical cost-based asstnhud ccounting ad rey reporting system with a fair value accounting and reporting in India. Thus, the present research h aoslsduemrse?s significance by making us understand the impact of a change in accounting paradigm on perceptions. pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! Aosoencluy ued oduc 11 s arg in the intr tion, fair value accounting has found application in the banks? trading book riitni eIsndia till today. Since it is unknown whether volatility in fair values of ?trading? category of and hence in banks? tbo f income will leadu dtoy a substantial or a negligible change in the volatility banks? equity returns, the objective of the st is to ascertain whether the shift from historical cost fair value accounting for the Indian banks? trading book has a panks? ffected the return volatility of the own listed shares and of bank index. Although the trading category of a bank?s i tih o nr rotfwoliome of nvestment constitutes a relatively small percentage its total assets, the said analysis is expected to so light on the issue of volatility in fair value measurements which has not been documented Indian literature as yet. INnf u tlhle light of the above-mentioned objective, the following hypotheses may be framed ? Hypothesis (Ho) : Change in acior counting regulattyions from historical cost basis to fair value the banks? trading book does not induce any volatili in returns on banks? stocks or bank index [ c.hea. nsgtaetistically speaking, Population Value of the Coefficient of the Variable associated with such in accounting regulations is equal to zero]. AtVrlatderinngative Hypothesis (Ha) : Change from historical cost accounting to fair value for the banks? book induces volatility in returns on banks? stocks or bank indeqauluae ex [i.e. statistically, Population l of the Coefficient of the Variable representing such change in accounting regulations is not to zero]. pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! T eeshrqt e uim tatiaeotsetnds of the above hypotheses are only for statistical significance (and not correct sign) of t1h2e regression coefficient of the variable, representing a change in accounting regulations, in the process generating volatility of banks? stock returns or bank index retur sectno atuetfi rsnfitci s ns as variance of c. ieanl Ilfty the coefficient of thhee raecbcyounting variable in the regression equation is found to be significant meaning t that the hypothesized (Ha) population value of such regression is not equal to zero, then the null hypothesis (Ho) that the shift t voert o fair value accounting does s ainduce any volatility in stock returns gets rejected in favour of alternative hypothesis and vice-. SOample Seleosrf ction and Research MethTohde ology the two major stock exchanges in India _ Stock Exchange, Mumbai (popularly known as ?BSE? Bombay Stock ExchanIAntarted ge) and The National Stock Exchange (NSE) __ it is the BSE which has computing a separate index, based on ?free-float market capitalization? mfBacdirciaon ethodology, for the rd iBngalnyking Sector, in the name of n??BSE BANKEX?, with effect from 1st January, 2002. , the study on bank index retur and individual banks? stock returns volatility induced by value accounting concentrates respectively on BSE BANKEX and a few banks comprisingef iAghNteKeEnX sector the , considered as barometer of performance for the Indian banking stocks. Among the (18) banks that comprised the BANKEX as o1loat n 11th February, 2008 (i.e. the date on which free adjustment factors of index constituents have been first revised by BSE since its compilation on p.bBe1ar.i2o0d02), the following nine (9) banks are selected in the sample for whom sufficient data for the prior to the adoption of fair value in annkks the bank?s trading book are available. However, the sample , i.e. ICICI Bank Ltd., State Bank of India, HDFC Bank Ltd., AXtIaSl Bank Ltd., Kotak Mahindra Ltd., Bank of Baroda, Bank of India, Federal Bank Ltd. and Orien Bank of Commerce, include pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! bw o vt ae t ri h ig ah p nt u cs blic (4) and private (5) sector banks and are a good representation of higher as well as low1e3r e in total free float market capitalization of BANKEX on 11.2.2008 [refer Annexure ?I]. As of individual stocks? or index returns are affected by the variability in overall m thcmeos e arket returns, an t en rtkr dull hypothesis (Ho) for zero population value of the coefficient of the accounting variable is first o etlwithout controlling for market variance and then by using volatility in pmlearket returns as the variable. For the said purpose, BSE SENSEX is also included in the sam as the indicator of movement. Ahass noted earlier in the introductory section, the RBI, vide its guidelines issued on 9th November, 2000, 7t 5% required a complete marking of an Indian bank?s trading book, which should account for atleast of the bank?s total investment portfolio, to fair phe value. As the date(s) on which individual banks in sample have started applying fair valuation in their trading books is/are not readily availableilnutbrolidsuhceed in d sources, and there usually being a distinction between de jure (i.e. when a new regulation is in the country?s legal frameworaaw) k) and de facto (i.e. when the banks actually adopt the new change in accounting regulation, it is assumed for the purpose of the study that the arfse storical banks which included in the BANKEX must have adopted the change from hi cost to fair value as early at the start of the next financial 2wo0r year i.e. from 2001-2002 onwards. Thus, the time period considered conducting tests of volatility in BANKEX returns is a fourteen (14) year periodh i0th7-2008 adoption from 1994-1995 to , comprising 7 years before and 7 years after of fair value in bank?s trading book, April 2001 being the moadeneqce nth of adoption of fair value. But the beginning period of research and u atthee period of study varies for some individupalle mbaennktast isoenlected in the sample because of lack of data for the entire 7 year period before im of fair value. pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! S oonince higher frequency data such as monthly or weekly observations are needed to carry out the te1st4s omMf volatility of stock returns, and as Indian banks are required to report fair values of trading category b securities on a monthly basis, monthly index values of BSE BANKEX and BSE SENSEX mo tanitnhelyd are from the Securities and Exchange Board of India (SEBI) website (www.sebi.gov.in) while stock price data for each of the sampled b i e onntiitoonriendg database anks are procured from CMIE (Centre for Indian Economy) Prowess over the study period. In this context, it may be that di rgnored vidends paid by the banks annually or semi-annually, being low frequency data, are in return calculations on the ground that the objective of hepa asction e analysis is to examine the share price to volatility induced in th financial statements by fair value. It may also be noted that as BSE berio sdtarted computing BANKEX values from 1st January, 2002, monthly BANKEX values for the prior to December 2001 (whose BANKEX value is taken as 10 e?a y 00 as base value) are calculated c phrocuring information on month-wise market capitalization and percent of promoters? holdings for index constit ppercentage uents as those on 1.1.2002 from CMIE Prowess database. At first, monthly free-float? is determined for each index constituent as 1 minus t nr? o ormmoalters? never hat month?s proportion of shareholdings in the bank which would come in the open market for trading in the course. Thereafter, tBhpiDnA eerNceKnEtaXge ?free float adjustment factor? as specified by BSE for that particular free float? is tracked and multiplieadr kweitth the specific month?s total market capitalization of bank to arrive at month-end free-float m capitalization of the bank. Then, monthly free float values are edceexm determined by aggregating month-specific free float market capitalization of each cboenrstituent (calculated in the akbeotve manner) relative to 1000 points for the base month?s (i.e. end , 2002) total free float mar capitalization of BANKEX. pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! B sBtaastiesdti coanl the above database, the analysis in this study is conducted using Microsoft Excel and SP1S5S softwares. In the first step, the serinoaAtuNraKlEX es of continuously compounded monthly returns for , SENSEX and for each of the sample banks over the study period are determined as the logarithcr m of price relative {ln (Pt / Pt-1} i.e. ending or current month?s price (Pt) over beginning preceding month?s price (Pt-1) since stock return volatility is eopxnamtininueously usually measured as its variance of compounded returns. Thereafter, a test is run on BANKEX and SENSEX series to whether the return process oarticular can be explained by any autoregressive (AR) time series model of a order, as documented in many previous research studies concerning the stock marketdf . A plot ae vreelgorpeesdsion residuals, Durbin-Watson (D-W) statistic, LBQ statistic (i.e. Box-Pierce Q statistic by Ljung and Box) and Correlogram, and ploogtenainst that t of actual and autoregressive estimated returns lag 1 returns indicate an autoregressive model of order one i.e. AR(1) model can be applied s tthe sample and that AR(1) model errors are heteroskedastic pienr )nature (the detailed findings of the run have not been presented for the sake of volume of the pa . Imnt o asct cordance with the methodology used in European Central Bank (2004) study, considered as the important study in the context of ?volatility? in fho air valuations, the following model is developed fDoa rs generatee dvolatility of individual banks? stock returns, as well as BANKEX returns. As AR(1) model emerg as the correct order in the test run, the following autoregressive equation is first estimated each sampled bank and BANKEX by regressing the time serie at c ecsopuinte s of returns on its own lag 1 series. t itnhge fact that variation in a bank?s profitability and asset structure influence its stock returns, an va ahne riable reflecting such a change in bank?s operational position could not be factored into d following autoregressive equation primarily because Indian banks usually disclose their earnings financial position on a yearly (or quarterly) basis whereas monthly data are needed to conduct tests pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! o ecv n fh e an st nt ock return volatility. However, such an exclusion will not obscure the analysis because, in t1h6e g ethsat a bank?s earning is not observable on a monthly basis, common investors tend to follow the in the bank?s current month?s stock price to form expectations and base their trading de ooafv fr cisions the next month. Since the movement of stock prices is influenced mainly by opinion and decisions aainlaabllyests, who possess superior knowledge and skills to extract additional information from those in public domain, investors rely on recent changes in stock prices to devise trading str iSnowo vr ategies etshteo rssucceeding period. Thus, such ?feedbarcnks trading? or ?return-chasing behaviour? exhibited by suggests that one month lagged retu should adequately explain current month?s returns. , here ri,t = ai,o + ai,1 . ri, t ? 1 + ei,t ?(1) ri,t = Continuously compounded returns on ith bank for month ?t?, r i , t- 1 = iLmamg e1d cioatnetliyn uporeucselyd icnogm mpoonutnhd?esd c orenttuinrnuso ounsl yit hc boamnpko fuonrd medo nrethtu ?rtn? si,. e. ai,o = Intercept of AR(1) model for ith bank, ai,1 = Slope coefficient of lag 1 return series for ith bank, ei,t = Residual or error term in autoregression of ith bank for month ?t?, and m i = Specific bank in the sample or BANKEX. Awsf he aticn h hi en gtest run has shown that thtieo nseries of residuals (ei,t) over the study period is heteroskedastic thereby that the distribu of ei,t?s is normal with zero mean and a time varying variance is conditional on how large the squared error is in the immediately preceding pe -ollowing ank riod (e2i,t-1), the variance process is adopted for each sampled b and BANKEX on the ground that the auto c ornesgtrreusesdive vreasriiadnucaels, which reflect the differences between actual and estimated values, can be as i.e. volatility of returns ? pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! w here a r e d hi,t = ?i + ?i . e2i, t ? 1 + ui, t ?(2) 17 hi,t = Squ residuals of ith bank?s or BANKEX? AR(1) model for month ?t? (i.e. e2i,t), e 2 i, t- 1 = Lag 1 squared residuals of ith bank?s or BANKEX? AR(1) model for month ?t? ? i.e. squared residuals in the immediately preceding month, i = Intercept in above linear regression for ith bank or BANKEX, ?u i = Slope coefficient of lag 1 squared residuals series for ith bank or BANKEX, and i, t = Error term in above regression of ith bank or BANKEX for month ?t?. Ha owever, equity return volatility tends to be asymmetric in the sense that it is likely to increase mtshfteeor ore r ineesgative return shocks than after positive return shocks of the same magnitude. Two possible ? leverage effect hypothesis and volatility feedback puecq ricuc h effect ? are usually put forward to explain ie ts asymmetries in volatility. The leverage effect states that after an unexpected decrease in equity ie, sthe debt ? equity ratio of a firm tends to rise thereby causing the riskiness i.e. volatility of to surge as well. According to volatility feedback effect, the incre rae o rc smcuolputesnsate quity ase in expected return to the investor for increased volatility (i.e. e variance) following a negative return shock n tinto further movement of equity prices and a consequent increase in volatility. In order to take of such asymmetric volatility phenomenon, equation (2) is enriched with an additiona f?u epnrcetisoennting as l term the asymmetric volatility variable, whose values are obtained the product of indicator [I (ei,t-1 < 0)] and lag 1 squared residuals ft(o ? of ith bank?s or BANKEX? AR(1) model for month imr [e2i, t-1]. The value of the indicator function is conditional on the sign of the AR(1) model?s residual pthlyei nigmmediately preceding month (i.e. ei,t-1) and accordingly trake the value ?1? when ei,t-1 < 0 negative return shock as actual return < estimated return) o else the value ?0?. Thus, hi,t = ?i + ?i . e2i, t-1 + ?i . [I(ei, t-1 < 0) . e2i, t-1] + ui, t ?(3) pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! w? here oeff 18 ?i = slope c icient of asymmetric volatility variable series for ith bank or BANKEX. (di o mwany) turn out to be positive (negative) in regression meaning thereby that volatility actually goes up when ei, t-1 is negative [i.e. I(ei, t-1 < 0) takes the value ?1?]. NvwaosAtuh lwuice, to test whether equity return volatility has increased significantly following the adoption of fair h in bank?s trading book, a dummy accounting variable (di,t) is introduced in the above equation (3) takes the value ?0? for the entire period before adoptio wp driy n of fair value (i.e. for each month in the l period prior to April 2001) and the value ?1? for the whole period after such adoption (i.e. from 2001 to March 2008). Accordingly, hi, t = ?1,i + ?2,i . di,t + ?i . e2i,t -1 + ?i . [I (ei, t-1 < 0 aTnh(ihd ere n ) . e2i,t-1]+ui,t ?(4) ?1, i = Iintercept i above multiple linear regression for ith bank or BANKEX, and e ?2, i = Slope coefficient of dummy accounting variable series for ith bank or BANKEX. idea is to check whether the intercept of the equation generating volatility of stock returns changes .e. shifts parallelly) with the introduction of fair value accounting by looking at the statistics relating to the coefficient of the dummy accounting variable (?2, i). Ac lso, to assess the impact of a change in accounting regulataowss notrcoialltiendg he ions on banks? return volatility after for volatility of t whole market, another variable denoting the time varying variance with the whole market (i.e. hm,t) is introduced in above equation (4) as follows - hi,t = ?1, i + ?2, i.di,t + ?i here . e2i,t -1 + ?i . [I (ei, t-1 < 0).e2i,t-1] + ?i .hm,t + ui,t ?(5) ?i = Slope coefficient of market volatility var BANKEX. iable series corresponding to ith bank or pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! Im t aTno mdaeyl be noted that hm,t series is obtained using a procedure similar to that used for hi,t. At first, AR(119) is run on continuously compounded monthly re vh d turns of SENSEX in line with equation (1) above e nits residuals are extracted for the same time period as that coinciding with ith bank or BANKEX. , such residuals doerliavtielity are modeled in line with equation (3) above after considering asymmetric phenomenon and estimated values of hm,t are recovered and plugged into equation (4) to equation (5). Tmhuultsi,p ltehe following two equations are estimated for each sampled bank and BANKEX using the linear regression technique in SPSS: ?i,t = ?1,i + ?2,i .di,t + ?i . e2i, t ? 1 + ?i .[I(ei, t ? 1 < 0) .e2i, t ? 1] ?(6) (without controlling for market variance) T ( a f t e r c ?oin,t t r=o l ?lin1,gi +fo ?r 2m,i a.drki,te +t v?air .i aen2ic, et ? t 1o +c h?ei c.[kI (feoi,r t ?a n1 y< i0m) p.ero2iv, te ?m 1]e n+t ?i . hm,t ?(7) d( uhmenm in significance of ?2,i ) , ystandard error (SE), t-statistic and significance of t-statistic (i.e. p-value) of the coefficient of accounting variable (?2,i) ifT7a) n each of the above two multiple linear regression equations (6) and are tracked and by looking at the p-value of ?2,i, inference is drawn as to whether the adoption of s hire value accounting in bank?s trading book has changed volatility in bank?s equity or index returns. p-value of estimated regression coefficient of an ind(lieiuch t ependent variable helps to determine whether independent variable is significan in regression at a conventional significance level such as 5% .vee. l0.05) or any other standard that may be considered appropriate. The p-value denotes the smallest of significance at which the null hypothesis (Ho) that the population value of t tnah d lte ependent ded he coefficient of variable is equal to zero can be rejected, in a two-si test. Thus, the lower the p-value, e rsntraotnivgeer the evidesnce against Ho (i.e. Ho can be rejected at a higher confidence interval) in favour of hypothesi (Ha) and hence the independent variable is considered highly significant in pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! r egress ion. Since regression is run with 95% confidence interval i.e. at 5% or 0.05 level of significanc2e0, ihfiHsy pp-ovalue of ?2,I < 0.05 i.e. ?t 2,i is statistically significant in above multiple linear regressions, then null thesis (Ho) tha the population value of dummy accounting variable?s coefficient is equal to zero n o rejected meaning thereby that fair value accounting has induced volatility in stock returns; otherwise is accepted. Thus, acceptance of Ho amounts to supporti tle o es t ng the belief that fair value accounting does vt ei i ln n sg duce volatility in equity or index returns and vice-versa. Such a p-value approach to hypothesies is particularly useful when testing the same hypothesis using a 0.05 and a 0.01 significanc lead to incomparable results. FIrnindings veg lrienses iownith the methodology described in the preceding section, the results of autoregression and estimating the two different return volatility equations - one without controlling for mcsoao omnlattpriollelity e ther arket [i.e. equation (A) in Section IV] while the o containing the whole market variance as a variabl [i.e. equation (B) in Section IV] - for BANKEX and for each individual banks in the are presented in Annexure ? II df and III respectively. In order to draw an inference on the impact t u mfamiry valuation in banks? trading book on its equity or index return volatility, the coefficients of accounting variable, their standard errors, t-statistics and p-values in sougmemthaerrized / regression equations, with the remark on acceptance rejection of null hypothesis (Ho) based on p-value are in Table 1. pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! NiInnamdeex Results of relationship between fair vTalaubel eaccIoun ung 21 ? ti and volatility of stock returns o /f B Baannkk PeSrtiuoddy o f C Dummy Accounting Variable e [ c( -ot-whwneiit tShha oamuntpdle (oS Eetafrfnricdoirae)rndt / t-Statistic p-val HHina0 d:: uFFcavmartar saeiti orrvc ovvklaaaltu Rielie tamyc cianor ukstn oting does not rlueetu arcnc ovuonlattciinklig tr yien]turns duces 1 ira oknl ecftoe r . IBCaInCkI MNao1rv9ce9hm7 ber ?cm Ltd. 2 ?0 08 voa wa ithout (rnritkarneotcl efor [-0)0 .0. 0117 1 1] (-) . 567 0 .1 20 Apd s p-valuc zTeourpmou.mlayti oanc c eo vuanlnuoteitn gole fv sast rhieat bhlaceno eisff0 i.ec0qi5eu,n atl th otoef Thus, Ho accepted. -fv ooaarfter rn imtarnoa clrlekinegt (- ofh uds-,uv Hmalomue ayc cnevopat trleieadsbs. l et haisn 0n.o0t5 , sciogenfiffiicciaenntt. [0) .00.1011]8 (-) 1.729 0.086 As p 2 . Somfta Itned Biaa nk MMaaryc 1h9 29040 ?8 -cwoanirttkhreootlu (-) t [0 .00.0030]3 (-) 1.074 0.285 AvHaser niapcb-evlea?lsu ec oneoftf ilceisesn tt hiasn n o0t .0s5ig, ndiufimcamnty. , Ho accepted. -cwonitthro l the for market M ([-0)0.0.00034] (-) 1.147 0.253 Aacsc eppt-evda.l u e not less than 0.05, Ho 3. HBaDnFkC L td. Jualyrc 1h9 29050 ?8 -mcwoanirttkhreootlu t (-) 0 [0.0.0020]5 (-) 2.287 0.024 Apdouspm umlpa-ytvialue is less than 0 oanc covuanlutien go fv arthiaeb lcoeffi.c0i5e,n t thoef -c m tco zero. Thus, Ho rej e is not equal thwoenit a ected. thro lr kfeotr (-) 0.006 [0.002] (-) 2.434 0.016 Areosje efcfpti-ecvdiealue < 0.05, dummy varia. n t is significant. Thus, bleH?so p d f M a c h in e Table ? I (continued) A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! N iInnamdeex o /f B Baannk PeSrtiuoddy o f C(oSEetafrfnricdoDiraeummy Accounting Variable H 22 c( -ot )rndt / t-Statistic p-value [H a n0o :t Finadiru vceaRlueem aacrcko u nting does v-mwhwneit Sha amnpdlk e artiatr : Firhaoknol eucftoet r satiorc vka rluer evteuorlnatility in stock atuccr onvuos nlattiinligt yin] duces 4 . ABaXnIkS Ltd . J Ma1ar9cn9hu9a 2 r?0y 0 8 -mcwoanirttkhreootlu (-) 0 .005 t [0.009] (-) 0.523 0.602 Ad s p-value not less t( siugmnimficya nvta. riable?s coefficiheannt is0 .n0o5t, c Thus, Ho accepted. - [-0) thwoenit tmhro l for .00.0090]5 (-) 0.605 0.546 Aacsc epp-value not less than 0.05, Ho 5 ted. . KMB arket maoanthakik nLdtrda AMparricl h1 929040 8?. -cwoanirt ([-0) .00.1001]7 (-) 1.767 0.0 79 Adc tkhreootu t - s Aaiucgsm scneimpfpi-cyva anvltaue not less than . rTiahbules,? sH 0c oaecfcfeicpiteendt. is0 .n0o5t, thwoenitth l ro l for ([-0) .00.1001]7 (-) 1.715 0.0 88 p-tveadl.ue not less than 0.05, Ho 6 B market ma Ma . Banrokd oaf Aprricl h1 929070 8? -cwoanirttkhreootu t (-) 0 .003 accepp-value not less than 0.05, Ho l [0.005] (-) 0.578 0.5 64 As ted. -cthwoenit tmhro alr kfoert ([-0) .00.0050]3 (-) 0.585 0.560 Aacsc ep-value not less than 0.05, Ho 7 pted. . BInadnika of July 1-m March9 29070 ?8 cwoanirt tkhreootu t 0.00 0.05. l [0.0053 ] 0.629 0.531 Ho accepted as p-value not less than pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! I Nn P eS r t iuo dd y o 23 thdaewme e of Bank f C(oSEetafrfnriDcdoirauemmy Accounting Variable Table ? I (continued) xSi a/ mBapnlek ( i-n co-mwnttiahtrrho kaoln )rndt / t-Statistic p-value [H Hin0a d:: uFFcair valuRe eamccaorukn ting does not var ufdot r saeti orvc ovklaa rltueietliu tarycn cin ov uostnlaottciinklig tr yien]t duurnces s 7 ianecte . BI - 8 c n thwoaednintikah of tmroalr kfoert 0 005] 0.576 0.565 p-value not less. FB [0..003 accepted. than 0.05; therefore Ho meadnekr aLl td. - MaA1r9cu9hg4u 2 s?0t 0 8 cw ( pa-cvcaelputee dnot less than 0.05; thus Ho [-0) .00.00705 ] - 9 c oanirtkheotu t (-) 0 .799 4 0 . 25 . thwoenittrol tmhro alr kfeor ([-0) t .00.0070]6 (-) 0.861 0.3 91 AvTahsr uipas-bv, lHael?oue not less than 0.05, dummy sa ccocefficient is not significant. epted. . O BC -moarinekn toafl MFa1er9bc9hru5a ?ry 2008 [00 A0..000 acs p-value not less than 0.05, Ho cwaimrtkhmeotue trce 043 ] 0.6 87 0 4 . 93 cepted. -control .0 0 [0.0043 ] 0.710 0.4 79 10 t . Bh woenit tmhro al for B AM pp a-value not less than 0.05; hence Ho ( S rket ar ccepted ricl h 1 I 929040 8? mBAa ENn kK EX -cn vwoditehxo) ut ([-0) .00.0020]1 (-) 0 .401 0 .6 89 A pcaosp upl-avtiaolune vanlot less than 0.05, the ca arnritkarneotl for - fvaooafrt ce ( rne imtarrno aclrlekin egt [-0) .00.0020]1 (-) 0 .471 0 6 . 39 Hocecfficient is e uqeu oalf tdou mzemroy. vTahriuasb, leH?so Avas epted. er nipac-bevla eHl?uos eac conceoeftpf iltceeisdes than 0.05, dummy .n t is not significant. pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! A rle proerltaetdively small standard error of the coefficient of dummy accounting variable across banks 2a4s in the Table 1 indicates that such coefficiennow) s t is more or less stable (i.e. its standard deviation is across observation . A compilation of the empirical evidences regarding acceptance / rejection of cvEo u al lul mhynpothesis based on the p-value of dummy accounting variable?s coefficient (as shown in the last of above table) suggests that the change in volatilit i v leune g y of stock returns due to adoption of fair accounting in the bank?s tradin book is almost never significant except for HDFC Bank Ltd. with the bank i tndex ndex (i.e. BSE BANKEX), fair value accounting has not induced any volatility in returns (as Ho is accepted for BANKEX as well). It can also be observed fr dBhuamt m om the above table coyntrolling for volatility in market returns does not improve significance (i.e. p-value) of the accounting variable in generation cter aAanNKEX ypothesis of stock return volatility for any of the sampled banks or . As the null h (Ho) is accepted for 8 out of 9 sampled banks and the BANKEX, it d ibneg inferred that the change from historical cost to fair value basis of accounting for the banks? book has not induced volatility in returns on banks? stoc nwxoceTi teption ks or bank index. Even for the only (i.e. HDFC Bank Ltd.) in the above empirical findings, the rejection of null hypothesis may d ebley conclusive because its p-values (for both ?with? and ?without? market control) do not differ from the specified 0.05 level of significance. As a result, it would be bett mxhcuesptional ts er if such an outcome is recorded as a piece of information that cas some doubt on the null hypothesis. , it may be said that the above evidenc ablr oevaidnyg ng es of negligible change in stock return volatility when from one accounti regime to another corroborate the contention that financial markets ba deenciesfiiot se their valuation of bank?s stocks on fair values so that the reform would only have the nosf aligning the balance sheet treatment with the market valuations that already guide investment . Such an argument, in turn, helps in allaying the fears of re-allocation of capital away from pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! tahcec obuanntkin?sg equity following the additional volatility in financial statements brought about by fair val2u5e . CTvr ah o le nclusion u eesmpirical investigation of the relationship between stock return volatility and the application of fair in the banks? investment portfolio over the time ismevrtup e da a iec le st d n air period April 1994 ? March 2008 has thus that the introduction of f value accounting in the banks? trading books had no significant o the volatility of banks? stock returns. Such a result is consistent with the previous research conducted by Barth, Landsman, and Wahlen (1995) ttHh ef mh eerred As , and European Central Bank (2004) to in the ?literature review?. explained in the latter part of findings, the investors could be in oe habit of basing their investment decisions on fair valuation and hence have not moved as such by additional volatility in the values of banks? invest ewnteiovneer ds ment portfolio induced by fair value accounting. , such a result nee to be interpreted with caution for the following reasons already goeneralizidn gin the ?Methodology? and ?Objective? sections and calls for further research before the inference for other items of balance sheet and fosbtf r other types of concerns as well. First naa nratlkelsd, the cihnogice of the month (i.e. April 2001) from which the sampled banks are deemed to have valu their investment portfolio on fair value basis may not match with the month when such have effectively implemented fair valuations in their trtioottett ading books, the information of which is ma lreadily available in published sources. Secondly, the relative weight of the bank?s trading book to sassets being modest, the results obtained with the trading book need not necessarily apply to other of assets in the bank?s balance sheet. These c lhike aveats imply that further investigation is needed on e lsyaid subject before resolving the issue of volatility confronting the fair value accounting and this is to be facilitated by the adoption of AS 30 and AS 31 in all forms of Indian companies and not pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! jsuoipf gsnt ibfiacnaknst in the sneear future. Also, if the subsequent research studies point toward a permanent a2n6d increa in stock return volatility after the adoption of fair value accounting, to the contrary the present research study, it wou primr oicpienrgly ld become necessary to examine whether the investors are able to interpret the relevant developments, disentangling the different sources of volatility and the fair-value led incremental volatility in the context of an asset pricing model. msseuapatpyh es Besides, the o drselating to the extent of model use for determining fair values, range and type of estimation to be employed, likely accuracy of reported fair values, and methods by which reported values l ibcaet vioenrified need to be thoroughly explored before imposition of heavy-handed rules on widespread and calculation of fair values. Bib - liography p - eArfboormodayn,c eD? ., Barth, M.E., R. Kasznik, (1999), ?Revaluations of fixed assets and future firm , Journal of Accounting and Economics 26, pp. 149-178. i Ahmed, A.S., Takeda, C., (1995), ?Stock market valuation of gains and losses on commercial banks-nvestment ? securities?, Journal of Accounting and Economics 20, pp. 207-225. - Barr, C., (2008), ?Fair-value accounting : Here to stay?, Fortune. v CaBluaarttih, M.E., (1994), ?Fair value accounting : evidence from investment securities and the market - on of banks?, The Accounting Review 69, pp. 1-25. K _o________, (2004), ?Fair values and financial statement volatility?, in Theoustnatsriessatsaronis audio Market Discipline Across and Industries, edited by Cl Borio, William Curt Hunter, George G Kaufman, and T , MIT Press, Cambridge, MA. pdfMachine A pdf writer that produces quality PDF files with ease! Produce quality PDF files in seconds and preserve the integrity of your original documents. Compatible across nearly all Windows platforms, simply open the document you want to convert, click ?print?, select the ?Broadgun pdfMachine printer? and that?s it! Get yours now! - d - is Bcalortshu,r eMs .E., Beaver, 0W7?.H., Landsman, W.R., (1996), ?Value relevance of banks? fair val2u7e under SFAS 1 , The Accounting Review 71, pp. 513-537. a _________, (2001), ?The relevance of the value relevance literature for accou-wnother nting standard setting : view?, Journal of Accounting and Economics 31, pp. 77-104. B 2 ith arth, M.E., Clinch, G., (1998), ?Revalued financieasl?, tangible, and intangible assets : associations share prices and non market-based value estimat , Journal of Accounting Research 36, pp. 1 - 33 99- . f Barth, M.E., Landsman, W.R., (1995), ?F
Posted: 20 July 2010

See more from Banking in India

Expert Views    
Rising Forex: A Potential for Infrastructural Development   By Jina Saha, The Cathedral Vidya School, Pune
Fair Value in Banks and Volatility in its Stock Returns   By Dr. Tanupa Chakraborty, University of Calcutta, Finance Professor
Financial Services Sector   By UK Trade & Investment
Financial Services Sector   By UK Trade & Investment
Financial Sector Opportunities in India   By UK Trade & Investment
Presenting the contributor